Supplementary appendix
This appendix formed part of the original submission and has been peer reviewed.
We post it as supplied by the authors.

Supplement to: Stenberg K, Hanssen O, Tan-Torres Edejer T, et al. Financing
transformative health systems towards achievement of the health Sustainable
Development Goals: a model for projected resource needs in 67 low-income
and middle-income countries. Lancet Glob Health 2017; published online July 17.
http://dx.doi.org/10.1016/S2214-109X(17)30263-2.
Financing transformative health systems towards achievement of the health Sustainable
Development Goals: a model for projected resource needs in 67 low-income and middle-
income countries
Karin Stenberg, Odd Hanssen, Tessa Tan-Torres Edejer, Melanie Bertram, Callum Brindley, Andreia Meshreky,
James E Rosen, John Stover, Paul Verboom, Rachel Sanders, Agnès Soucat




Supplementary Material
This document provides supplementary information to the main paper.1 It has eight sections:

Section 1: List of countries included in the analysis
Section 2: Country groups and pathways towards Universal Health Coverage
Section 3: Start and end points within models
Section 4: Cost and impact projection methods
Section 5: Methods for projecting available financing
Section 6: Review processes
Section 7: Efficiency considerations
Section 8: Additional results, tables and figures




          Section 1. List of countries included in the analysis


While the SDGs concern all countries, our model includes only low and middle income countries, as these are
faced with the greatest challenges in terms of health burden and mobilisation and effective use of resources. We
model estimates for all low-income countries, the 20 most populous lower middle income countries and the 20
most populous upper middle income countries.2 When excluding 4 countries lacking GDP data we are left with
a total of 67 countries, in size representing 95% of the total population in low and middle income countries,
including a set of the most vulnerable conflict-affected and fragile contexts (Table S1).




1
  For readers wishing additional detail to what is outlined within this document, please contact the corresponding author
(stenbergk@who.int; or whochoice@who.int)

2
  The selection of countries was performed in March 2016. At this time, the Russian Federation was classified as a high
income country, which is why it is not included among the 67 countries. It was reclassified to an upper middle income
country on July 1st, 2016. Furthermore, Cambodia and Tunisia were both reclassified as lower middle income on July 2016,
which means that the resulting country list for which results are presented here includes 28 LICs, 21 LMICs and 18 UMICs.



                                                              1
Table S1. List of countries included in the analysis

                                                                                                   Populati
                                                                                                      on
                                                                                                   affected
                             WHO                                                       Human          by      Fragility
                            epidemio                    Resource      Skilled Birth   Resources    conflict    Index
                   Income    logical    Population     Availability    Attendance     for Health     (%)       Score
    Country        Group*    region      (2015)**         ***             ****          *****      ******     *******
 Afghanistan        LIC      EMRD       32,526,562         2,000          45.2           0.72        8.3         47

 Algeria           UMIC      AFRD       39,666,519         13,880         96.6           5.03         -          31

 Angola            UMIC      AFRD       25,021,974         7,227          46.7           1.57         -          35

 Azerbaijan        UMIC      EURB       9,753,968          16,920         97.2           9.60         -          28

 Bangladesh        LMIC     SEARD      160,995,642         3,330          41.7           0.56         -          36

 Benin              LIC      AFRD       10,879,829         2,020          77.2           0.79         -          37

 Brazil            UMIC     AMRB       207,847,528         15,570         99.1           9.15         -          26

 Burkina Faso       LIC      AFRD       18,105,570         1,600          65.9           0.57         -          40

 Burundi            LIC      AFRE       11,178,921          770           60.3           0.48         -          42

 Cambodia          LMIC      WPRB       15,577,899         3,080          89.0           0.93         -          35
 Cameroon          LMIC      AFRD       23,344,179         2,950          64.7           0.49        1.5         38
 Central African
                    LIC      AFRE       4,900,274           600           40.0           0.29        46.9        46
 Republic
 Chad               LIC      AFRD       14,037,472         2,070          24.3           0.48         -          44
 China             UMIC      WPRB      1,376,048,943       13,170         99.9           3.10         -          26
 Colombia          UMIC     AMRB        48,228,704         12,910         98.7           2.09         -          31
 Comoros            LIC      AFRD        788,474           1,430          82.2           0.49         -          39
 Côte d'Ivoire     LMIC      AFRE       22,701,556         3,130          56.4           0.56         -          42
 Democratic
 Republic of the    LIC      AFRE       77,266,814          650           80.1           0.08        2.1         46
 Congo
 Dominican
                   UMIC     AMRB        10,528,391         12,600         97.7           2.79         -          30
 Republic
 Ecuador           UMIC     AMRD        16,144,363         11,190         96.3           3.78         -          30
 Egypt             LMIC      EMRD       91,508,084         10,280         91.5           5.69         -          36
 Eritrea            LIC      AFRE       5,227,791          1,529          34.1           0.64         -          42
 Ethiopia           LIC      AFRE       99,390,750         1,500          15.5           0.26         -          41
 Gambia             LIC      AFRD       1,990,924          1,580          57.2           0.90         -          37
 Ghana             LMIC      AFRD       27,409,893         3,900          70.8           0.97         -          32
 Guinea             LIC      AFRD       12,608,590         1,130          45.3           0.49         -          45
 Guinea-Bissau      LIC      AFRD       1,844,325          1,380          45.0           0.61         -          44
 Haiti              LIC     AMRD        10,711,067         1,730          37.3           0.51         -          45
 India             LMIC     SEARD      1,311,050,527       5,630          74.4           2.30         -          34
 Indonesia         LMIC     SEARB      257,563,815         10,190         87.4           1.54         -          30
 Iran (Islamic
                   UMIC      EMRB       79,109,272         15,688         96.4           4.98         -          32
 Republic of)
 Iraq              UMIC      EMRD       36,423,395         15,100         90.9           3.86        27.5        42
 Kazakhstan        UMIC      EURC       17,625,226         21,710         99.5          11.48         -          26
 Kenya             LMIC      AFRE       46,050,302         2,940          61.8           1.02         -          41




                                                       2
 Liberia              LIC       AFRD          4,503,438            700           61.1           0.27          -          45
 Madagascar           LIC       AFRD         24,235,390           1,400          44.3           0.50          -          40
 Malawi               LIC       AFRE         17,215,232            790           87.4           0.34          -          38
 Malaysia           UMIC        WPRB         30,331,007           24,770         99.0           4.38          -          25
 Mali                 LIC       AFRD         17,599,694           1,510          57.1           0.45         2.7         44
 Mexico             UMIC        AMRB        127,017,224           16,840         98.7           4.45          -          31
 Morocco            LMIC        EMRD         34,377,511           7,290          73.6           1.45          -          28
 Mozambique           LIC       AFRE         27,977,863           1,120          54.3           0.42          -          40
 Myanmar            LMIC       SEARD         53,897,154           4,546          77.9           1.60         0.7         37
 Nepal                LIC      SEARD         28,513,700           2,410          48.2           0.51          -          36
 Niger                LIC       AFRD         19,899,120            910           29.3           0.14          -          44
 Nigeria            LMIC        AFRD        182,201,962           5,710          35.2           1.92         0.8         41
 Pakistan           LMIC        EMRD        188,924,874           5,090          52.1           1.35          -          44
 Peru               UMIC        AMRD         31,376,670           11,440         89.9           2.56          -          27
 Philippines        LMIC        WPRB        100,699,395           8,450          72.8           1.43          -          37
 Romania            UMIC        EURB         19,511,324           19,950         98.7           9.03          -          21
 Rwanda               LIC       AFRE         11,609,666           1,630          90.7           0.76          -          36
 Sierra Leone         LIC       AFRD          6,453,184           1,770          59.7           0.18          -          40
 South Africa       UMIC        AFRE         54,490,406           12,700         94.3           5.75          -          28
 South Sudan          LIC       AFRD         12,339,812           1,800          17.2           1.34        18.6         49
 Sri Lanka          LMIC        SEARB        20,715,010           10,300         98.6           2.30          -          32
 Sudan              LMIC        EMRD         40,234,882           3,920          19.9           1.06         5.5         45
 Tanzania,
 United Republic      LIC       AFRE         53,470,420           2,510          48.9           0.43          -          37
 of
 Thailand           UMIC       SEARD         67,959,359           14,870         99.6           2.44          -          29
 Togo                 LIC       AFRD          7,304,578           1,290          44.6           0.31          -          37
 Tunisia            LMIC        EMRD         11,253,554           11,020         73.6           4.36          -          31
 Turkey             UMIC        EURB         78,665,830           19,560         97.4           3.92          -          28
 Uganda               LIC       AFRE         39,032,383           1,720          58.0           0.51          -          40
 Ukraine            LMIC        EURC         44,823,765           8,560          99.0          11.38         5.1         32
 Uzbekistan         LMIC        EURB         29,893,488           5,830          99.6          14.64          -          31
 Viet Nam           LMIC        WPRB         93,447,601           5,350          93.8           2.33          -          27
 Yemen              LMIC        EMRD         26,832,215           3,586          43.0           0.80        10.1         46
 Zimbabwe             LIC       AFRE         15,602,751           1,650          80.0           1.29          -          41
* Classification as of July 2016, World Bank Atlas method. At the time of country selection (March 1st, 2016), two countries
had another classification: Cambodia-low income; and Tunisia-upper middle income.
** United Nations, Department of Economic and Social Affairs, Population Division, World Population Prospects: The 2015
Revision, New York, 2015
*** Gross National Income per capita, adjusted for purchasing power parity, from the World Bank World Development
Indicators, accessed on June 3rd, 2016. For countries where data on GNI per capita was unavailable, we used Gross Domestic
Product per capita, adjusted for purchasing power parity (Angola, Eritrea, Iran, Myanmar, and Yemen).
**** From the WHO global health observatory, at the time of country selection, March 1st, 2016.
*****Defined as the combined measure of doctors, nurses and midwives per 1000 population. Data from the WHO global
health observatory, May 24th, 2016.
****** Percentage of the population in a country affected by a conflict, taken from WHO Humanitarian Response Plans and
WHO country Health Resource Availability Mapping Systems.
******* Composite score of five subcomponents of the Fund for Peace Fragility Index, which were demographic pressures,
poverty and economic decline, limits to the provision of public services, inexistence of a security apparatus, and presence of
external intervention. In addition to these final two indicators, we also considered countries that had undergone a serious
shock to their health systems in the past five years (Guinea, Liberia, Mali, and Sierra Leone) to be “vulnerable” countries.



                                                              3
           Section 2. Country groups and pathways towards Universal Health Coverage


2.1 Country groups


Information on the current situation in countries is best known domestically, including current constraints and
opportunities for moving forward. However, to ensure a realistic basis for our analysis, countries were classified
into five groups based on publicly available data. We employed criteria that capture countries’ current risk and
disease burden, their current resource availability, and the effective use of those resources. We consider the
following dimensions: (i) conflict/fragility; (ii) resource availability as measured by gross domestic income
(GNI) per capita and/or gross domestic product (GDP) per capita3; (iii) health system capacity as measured by
the current density of health workers as a proxy for service delivery readiness, and (iv) current health system
performance as measured through skilled birth attendance coverage.

Table S2 provides an overview of the five country categories. The main purpose of the classification is to
inform the modelled timing and duration of strategic investments. Countries belonging to lower level groups (C,
V, HS1) are assumed to be unable to scale-up as rapidly as countries belonging to higher level groups (HS2,
HS3) for many of the investments considered, due to the more limited absorptive capacity in their systems.4



Table S2. Country groups considered for the analysis

Type                 Description                                       Criteria

                                                                       (a) Conflict/fragility

                                                                       (b)   Resource availability5
                                                                                   o      GNI/Capita in PPP
                                                                                   o      GDP/Capita in PPP
                                                                       (c ) Service delivery readiness:
                                                                                   o      HR density
                                                                       (d) Current service delivery performance, as measured by skilled
                                                                             birth attendance coverage (SBA)
Conflict-affected   Countries with an internal or external conflict          •      More than 10% of the population is affected by conflict
states (C )         which considerably limits the state’s ability to                (criteria a).
                    provide health services

Vulnerable          Countries with structural vulnerabilities,         Countries with vulnerable systems that have one or more of the
countries (V)       ranging from localized conflicts, a weak state     following characteristics:
                    apparatus, presence of external actors such as           •    Recent health system crisis (criteria a)
                    international humanitarian response structures,          •    High score on the International Fragility Index6 (criteria a)
                    or recent health crises, which limit the state`s
                    ability to provide health services

Health System       Countries with poor performance across health      Countries have limited resources and low coverage of care.
category 1 (HS1)    system functions. These countries require an           •      GNI (PPP) / GDP (PPP) per capita falls under 2,500 (b),
                    engineering of their health system in order to                AND
                    build the foundations of strong health system          •      Less than 2.28 health workers per 10,000 population (c),
                                                                                  OR
                    institutions, and will thus require significant
                                                                           •      SBA<90% (criteria d)
                    investments across the health system.



3
  GDP data used when GNI data is not available.
4
  One of the main factors for absorptive capacity is the available health workforce which effectively sets the production frontier. Other
criteria include conflict/fragility, governance, and past performance on public expenditure management.
5
  GDP/Capita PPP used when GNI/Capita PPP is unavailable. PPP = Purchasing Power Parity –adjusted dollars.
6
  Countries with a combined score of more than 43.5 out of 50, based on scores for five key components of the Fragility Index developed by
the Fund For Peace. The five components are: demographic pressures, poverty and economic decline, limits to the provision of public
services, inexistence of a security apparatus, and presence of external intervention.



                                                                       4
Health System         Countries have invested in the foundations of    Countries with a combination of criteria:
category 2 (HS2)      health systems but institutional performance is
                      poor and there are challenges related to health      •     Countries that are resource constrained (GNI-PPP per
                      system efficiency and access. There is scope for           capita <2,500) but perform well on a representative
                      rapid health system scale-up to improve                    indicator for complex care (SBA>90%), signalling service
                                                                                 delivery readiness that allows for quick scale up for public
                      performance and move towards greater
                                                                                 service coverage, should resources be made available.
                      domestic financing sustainability.
                                                                            •     Countries that are less resource constrained (GNI-PPP per
                      This includes countries that:                               capita >2,500) but where key health workforce availability
                      •     have limited resources but are performing             is limited (HRH <2.28) , OR countries exceed the health
                            well in terms of SBA coverage                         workforce 2.28 benchmark but are doing less well on
                      •     have fewer limitations on economic                    service coverage and delivery of complex services
                                                                                  (SBA<90 %).
                            resources but face challenges with respect
                            to health worker density
                      •     have fewer limitations on economic
                            resources but are doing less well on
                            service coverage
Health System         Countries with mature health systems but in      This category includes:
category 3 (HS3)      which there is an ongoing need to support health
                      system transformation and reorient models of           •    Countries with relatively high resource availability7
                      care to address emerging challenges and                     defined as a GNI-PPP greater than 5,000, and high levels
                      existing inequities.                                        of delivery of complex care, defined as greater than 90%
                                                                                  coverage of skilled birth attendance (criteria b, d).
                                                                             •    Countries with high resource availability defined as a GNI-
                                                                                  PPP greater than 10,000 per capita (criteria b).



Table S3 presents the average and median values within each country group for economic and health systems
resources, and current service delivery performance, as measured by Skilled Birth Attendance (SBA) and
treatment of acute respiratory infection (ARI) in children.

It should be noted that the scale and scope of investments required are determined within the analytical model
for each health system or service component, based on an account of the current situation as well as the
anticipated system that countries will need to attain by 2030. There is therefore significant variation within each
of the four groups as to what the additional investment requirements are, and what the additional associated
costs would be.




7
    More than 10,000 GNI PPP/capita, or 10,000 GDP PPP/capita when data on GNI PPP/capita unavailable.



                                                                      5
Table S3. Current resource availability and health system performance within the five country groups,
average and median values 8

                                                         Economic          Health system resources               Service coverage
                                                         resources

Type                                    Number        GDP/Capita 9       Health           Number of       Skilled Birth    Pneumonia
                                        of                               workforce        Health          Attendance       treatment
                                        countries                        density per      Centers per     (%)12            (%)13
                                                                         1,000            100,000
                                                                         population 10    people 11

Conflict-affected          4            Average       7,088              2.2              10.5            60.5             62.9
states (C)
                                        Median        3,123              1.1              2.7             41.5             49.2

Vulnerable systems         11           Average       1,637              0.4              10.6            50.9             39.8
(V)
                                        Median        1,626              0.5              10.5            45.2             38.8)

HS1                        15           Average       2,002              0.6              10.6            50.1             24.4

                                        Median        1,529              0.5              9.3             58.0             34.3

HS2                        16           Average       5,892              1.8              6.0             67.8             27.7

                                        Median        4,511              1.2              7.5             64.7             41.5

HS3                        21           Average       13,611             4.3              16.8            98.6             67.2

                                        Median        13,262             4.5              12.1            98.6             59.0




Figure S1 presents the median values for each country group across a pair of service delivery and health system
capacity indicators. We would expect an increasing trend for HS1, HS2 and HS3 countries. Health workforce
density is significantly higher in HS3 countries than in the other categories. With respect to density of available
infrastructure, measured by available health centres and health posts per 100,000 population, the results are
more mixed, which can likely be both attributed to incomplete and unreliable data on infrastructure availability,
as well as different service delivery models in different countries, where a very high number of health posts will
distort this indicator. We therefore did not use facility density as a criterion to classify countries into the five
groups.

In general, results for Conflict countries are more difficult to interpret since this category includes both low and
middle income countries. Moreover, the latest available data may not represent the current situation in Conflict
countries, which has most likely deteriorated since the data was collected.




8
  General note: average values are population weighted averages per group. Average and median values are based on countries for which
data is available.
9
  US$,2014, World Bank.
10
   Doctors, nurses and midwifes per 1,000 population. Latest available data from WHO GHO.
11
   Number of health centers or health posts per 100,000 people. Available data from WHO GHO and national strategic national health sector
plans (http://nationalplanningcycles.org/).
12
   Percentage (%) of births treated by skilled attendants. Latest available data from WHO GHO.
13
   Percentage (%) of children under five with symptoms of pneumonia given antibiotic treatment. Latest available data from WHO Global
Health Observatory.



                                                                     6
Figure S1. Availability of key infrastructure and health workforce per country group (median values)




                  Facility density per 100,000 population. Includes health centers and health posts only.
                  Health workforce density: medical doctors, nurses, and midwifes per 1,000 population.



Meanwhile, data on management of acute respiratory infection in children indicates that on average, population
coverage is higher in HS3 countries than in HS1 and HS2, and similarly coverage is higher in HS2 than in HS1
countries, as would be expected (Figure S2).

Figure S2. Current coverage of essential health services per country group (median values)




           SBA: Skilled birth attendance; ARI trt: treatment coverage of acute respiratory infection in children 14




14
 Percentage (%) of children under five with symptoms of pneumonia given antibiotic treatment; latest available data from
WHO Global Health Observatory. http://www.who.int/gho/en/ Accessed 24 May 2016.



                                                              7
2.2.        Integrated health service delivery

SDG 3 includes a broad health goal, “Ensure healthy lives and promote well-being for all at all ages”, and calls
for achieving universal health coverage (UHC). The package of services to be provided as part of UHC is
country-specific and evolves over time, in response to changes in epidemiology, consumer demand, resource
constraints, and available technology. In terms of setting boundaries for a set of services to model for the
purpose of analysis, we reviewed the disease/programme-specific targets under SDG 3 and other related SDGs
(2 and 6) as well as the proposed tracer indicators for UHC service coverage.15 We reviewed published guidance
on essential health interventions and the list of services included under available disease-specific global
strategies. We also consulted individual technical departments within WHO for each relevant area in order to
obtain a list of recommended essential interventions.

Service delivery platforms: In recognition of the diversity of available guidance and technologies to prevent and
treat health conditions, our analysis considers four service delivery platforms. The delivery platforms represent
different modes for providing patients with information, counselling, essential preventive commodities,
screening, diagnosis, treatment, and follow-up. We discuss the delivery platforms in terms of three
characteristics of health services: discretionary vs standardised services, the level of intensity of the transaction
involved between provider and patient, and asymmetry of information. Before describing the platforms, we
describe these characteristics in brief.

Individualized vs standardized services: the concept of heterogeneous health services refer to those that are
tailored to the individual and not necessarily provided to the population en masse. An example is caesarean
sections, where doctors must exercise significant judgment on the aspects of the individual case in order to
determine what to deliver and how. On the other end of the spectrum, standardized interventions are those that
have great uniformity across patients. The latter tend to be recommended for and provided to a greater share of
the population, without screening or diagnosis. Measles vaccination is a typical standardized intervention which
is recommended for the entire population. Similarly, when a policy is implemented to have plain/standard
packaging and/or large graphic health warnings on all tobacco packages, this is a standardized intervention
which all people will benefit from and from which there is no “opting out”.

Intensity of transaction: transaction-intensive services are those that require a large amount of client-provider
contact.16 This includes services with repeated check-ups, such as management of more or less chronic
conditions like antiretroviral treatment for HIV/AIDS, or management of non-chronic conditions that require
extensive periods of health worker follow-up, such as management of severe acute malnutrition. It also includes
those interventions where significant health worker time is required during a peak time (complex surgery). On
the other hand, several interventions require very limited interaction with a provider, such as deworming.

Asymmetry of information: For many health services, doctors and other caregivers hold considerably more
information with regards to the recommended behaviours or actions to be adopted, than the patient does. For
example, patients contracting a sexually transmitted infection will require medical assistance to diagnose and
treat the infection. Many patients will have asymptomatic infections and will not be aware that they are infected,
which can only be discovered through testing. On the other hand, there are other interventions for which there is
very little asymmetry of information. For example, when bed nets are distributed, there is little doubt regarding
their intended purpose or manner of use. Similarly, immunization campaigns carry a clear message to prevent
illness, which is easily understood by the population.

With the above three characteristics in mind, we define four service delivery platforms, and assign health
interventions accordingly. We fully recognise that the organisation and presentation of interventions by
platforms does not imply that each intervention can only be delivered through one platform or at one service


15
     WHO (2015) Tracking universal health coverage: first global monitoring report.
16
     See World Development Report 2004.



                                                              8
delivery level. There exist a multitude of options for delivering services, and many may be simultaneously
relevant. We also acknowledge that technologies may change over time, and that the characteristics of future
service delivery platforms, or the preferred platform of specific interventions, may differ.

Within our model we use the organisation of services into platforms in order to assess the constraints associated
with the provision of each type of health intervention, and the rate at which those constraints can be overcome.
For presentation purposes, interventions are assigned to the platform where their delivery is considered most
cost-effective.

Platform 1: Policy and population wide interventions. This platform focuses on policies and information
communication that can be delivered to the population en masse at relatively low cost, to support changes in
behaviours among risk groups in the population, whether for preventive purposes (i.e., reduce smoking, promote
physical exercise, sleep under a mosquito net) or to ensure an appropriate response to a health problem (i.e.,
ensure that a child with diarrhoea takes oral rehydration salts and receives an increased intake of fluids).

Interventions included within this category are typically standardized, with low transaction intensity, and limited
information asymmetry. A key characteristic is that they require little or no contact with a health provider, and
thus do not place a burden on health worker time. Some interventions included within this package include
distribution of commodities, i.e., insecticide treated bed nets17.

Not all interventions within this category would follow the same pace of implementation. Many policy
interventions require initial investments in regulatory frameworks to improve implementation capacity. Other
interventions can however be scaled-up more rapidly (i.e., distribution of bed nets, with accompanying
information campaigns).

Many interventions in this category, such as tobacco prevention policies and mass media campaigns for
HIV/AIDS awareness, are highly cost-effective, and can be rapidly expanded at low cost in most countries,
although they will require initial investment in capacity, to design and implement effective programmes, and
institutions, to oversee their implementation.

This platform also includes services mainly funded and delivered by actors working outside the health sector,
particularly those related to water, sanitation, and hygiene, and the reduction of indoor air pollution. The
promotion of healthy behaviours related to environmental conditions is less reliant on health worker time, but
heavily dependent on hardware investments in pipelines and equipment improving households’ access to, and
use of, safe water and clean cooking equipment.

Platform 2: Periodic schedulable and outreach services. This category includes services which are provided
routinely and periodically. They may be provided periodically (for example mass distribution of drugs for
deworming) or provided continuously but accessed at a certain pre-determined period from the perspective of
the patient (such as in antenatal care or iodine supplementation). The key characteristic of these interventions is
that they are standardized and have low levels of information asymmetry. They require user contact with health
workers, but through brief and schedulable interventions. Because of their relatively high level of
standardization, a number of interventions can be delivered through health workers with short training. Some
interventions refer to the provision of counselling to certain patient groups – e.g., counselling parents to ensure
appropriate nutrition for their children, or reaching out to injecting drug users to make sure they exchange
needles safely.

In most settings – including resource constrained systems - services provided through this platform can be
rapidly expanded. There are opportunities for rapid scale-up of preventive care through population-based
approaches including community and outreach services, such as routine immunization campaigns and vector
control for neglected tropical diseases.

17
  Because the actual correct utilisation of bed nets is highly dependent on effective communication around their use, and there is very
limited information asymmetry with respect to the purpose of the nets, we have placed this intervention under platform 1.



                                                                      9
For other interventions in this package, the recommendation remains for delivery at the health center level – e.g.,
delivery of new immunizations based on more recent technology, such as rotavirus and pneumococcal vaccine.

Platform 3: First level clinical services: This platform includes mainly services delivered through primary
level health facilities. Compared to the first platforms, this platform includes individual health-care interventions
that are specific to the patient’s needs. This is the largest category of services within our model, and it covers a
wide range of services with different characteristics. Typically, however, these services require more than a
brief interaction with a health worker. They also require the health worker to have a certain level of skills and
diagnostic tools, and therefore can be described to have a medium level of transaction intensity. Examples
include treatment of sexually transmitted infections, treatment of TB, and treating and managing non-
communicable diseases such as diabetes. One consideration for placing services in this category rather than in
platforms 2 and 4 is the level at which they can be delivered. Many interventions can be delivered through a
primary health care model, and many patients can be seen at health centers for these conditions. Care is more
often than not tailored to specific patient needs (e.g., treatment of high blood pressure and diabetes), and with
medium to high levels of information asymmetry – more often than not at the higher end, following the typical
information asymmetry relationship between patients and health care providers.

Platform 4: Specialized care: Specialized care would typically be delivered by highly skilled health personnel,
and rely on sound diagnostic and referral systems. Examples of interventions include diagnosis and treatment of
cancer, management of obstructed labour, and management of severe acute malnutrition. These are health
services requiring a significant amount of health worker time, and with high transaction intensity. Services in
this category are typically highly individualized – e.g., identification and management of infertility. While
information asymmetry is high, many services are also discretionary in nature, meaning that patients can opt in
or out, and agree on the treatment in a participatory process (e.g., cancer treatment).




                                                         10
Table S4. A summary of the four service delivery platforms

Service delivery platform       Typical service characteristics        Implementation model                 Typical interventions
                                                                                                            (examples)
                                     •    Standardized                 Policies driven jointly with         Increase excise taxes and prices
                                     •    Low transaction              other sectors, such as Ministries    on tobacco products, alcohol
                                          intensity (require           of Finance, for example fiscal       and sugar-sweetened beverages.
                                          little or no contact         policies to make harmful and
                                          with a health                unhealthy products less
                                          provider)                    affordable.
                                     •    Limited information                                               Large graphic health warnings
                                          asymmetry.                   Changing consumer products –         and plain packaging on all
                                                                       e.g., large graphic health           tobacco products.
                                                                       warnings and plain packaging
                                                                       on tobacco products–is a quick-
                                                                       win policy to communicate
                                                                       health messages to large
                                                                       populations, while the
                                                                       production of these packages is
Policy and population wide                                             the obligation of the tobacco
interventions                                                          industry.

                                                                       Mass media campaigns and
                                                                       community mobilization
                                                                       interventions can be rapidly
                                                                       scaled-up since they are less
                                                                       reliant on health system
                                                                       strengthening. These can make
                                                                       standardized products (such as
                                                                       bed nets) universally available.

                                                                       Certain policy interventions         Bed nets for malaria.
                                                                       require a lag time during which      Restrictions on availability of
                                                                       institutions are built up, after     retailed alcohol, and
                                                                       which implementation rapidly         implementation of drunk driving
                                                                       expands.                             laws.
                                     •    Standardized                 Rapidly scaled-up as less reliant    Immunization, Neglected
                                     •    Low transaction              on health system strengthening.      Tropical Disease programmes.
                                          intensity (require           Makes use of outreach into the
Periodic schedulable and                  brief contact with a         community, but also health
outreach services                         health provider)             centre level delivery.
                                     •    Limited information
                                          asymmetry

                                     •    Individualized               Primary health care platform         Treatment of pneumonia in
                                     •    Medium transaction           where service coverage relies on     children,
                                          intensity                    a successively strengthened          Management of sexually
                                                                       health system, with functioning      transmitted infections,
First level clinical services        •    Medium to high
                                                                       and accessible facilities that are   Management of depression.
                                          information
                                                                       adequately staffed with health
                                          asymmetry.
                                                                       workers providing quality
                                                                       outpatient care.
                                     •    Individualized               Expansion of service coverage        Skilled birth attendance,
                                     •    High transaction             only happens after the build-up      Surgery for trauma care and
                                          intensity                    of specialized resources, which      fractures.
Specialized services                                                   may take longer to acquire and
                                     •    High information
                                                                       will rely more heavily on
                                          asymmetry.
                                                                       investments in the health
                                                                       system.




                                                                  11
Table S5. Essential Interventions organized into Service Delivery Platforms

Platform        Interv     Intervention name                                            Programme             Delivery levels          Tool used to
number          ention                                                                                        within modelled          model costs
(1-4)           numb                                                                                          approach18               and impact
                er

Platform 1. Policy and population wide interventions

1               1          Increase excise taxes and prices on tobacco products.        NCD                   National policy          OHT / Excel
                                                                                                                                       (*)

1               2          Implementation of plain/standardized packaging and/or        NCD                   National policy          OHT / Excel
                           large graphic health warnings on all tobacco packages                                                       (*)

1               3          Comprehensive ban of tobacco advertising, promotion          NCD                   National policy          OHT / Excel
                           and sponsorship, including cross-border advertising and                                                     (*)
                           on modern means of communication

1               4          Elimination of exposure to second-hand tobacco smoke         NCD                   National policy          OHT / Excel
                           in all indoor workplaces, public places, public transport,                                                  (*)
                           and in all outdoor mass-gathering places

1               5          Implement effective mass media campaigns that                NCD                   National policy          OHT / Excel
                           educate the public about the harms of smoking/tobacco                                                       (*)
                           use and second hand smoke
1               6          Provision of cost-covered, effective and population-         NCD                   National policy          OHT / Excel
                           wide support (including brief advice, national toll-free                                                    (*)
                           quit line services and mCessation) for tobacco cessation
                           to all those who want to quit

1               7          Hazardous alcohol use: Enforce restrictions on               NCD                   National policy          Excel
                           availability of retailed alcohol (**)

1               8          Hazardous alcohol use: Enforce restrictions on alcohol       NCD                   National policy          Excel
                           advertising (**)

1               9          Hazardous alcohol use: Enforce drunk driving laws            NCD                   National policy          Excel
                           (sobriety checkpoints) (**)

1               10         Hazardous alcohol use: Raise taxes on alcoholic              NCD                   National policy          Excel
                           beverages (**)

1               11         Physical inactivity: Implement public awareness and          NCD                   National policy          OHT / Excel
                           motivational communications for physical activity,                                                          (*)
                           including mass media campaign for physical activity
                           behaviour change

1               12         Sodium: Surveillance                                         NCD                   National policy          OHT / Excel
                                                                                                                                       (*)

1               13         Sodium: Harness industry for reformulation                   NCD                   National policy          OHT / Excel
                                                                                                                                       (*)

1               14         Sodium: Adopt standards: Front of pack labelling             NCD                   National policy          OHT / Excel
                                                                                                                                       (*)

1               15         Sodium: Adopt standards: Strategies to combat                NCD                   National policy          OHT / Excel
                           misleading marketing                                                                                        (*)

1               16         Sodium: Knowledge: Education and communication               NCD                   National policy          OHT / Excel


18
     Non-health indicated for interventions where all or a share of total costs are assumed to fall under other sectors than health.



                                                                         12
                                                                                                                   (*)

1           17       Sodium: Environment: Salt reduction strategies in           NCD        National policy        OHT / Excel
                     community-based eating spaces                                                                 (*)

1           18       Diet: Complete elimination of industrial trans fats         NCD        National policy        OHT / Excel
                     through the development of legislation banning their                                          (*)
                     use in the food chain

1           19       Mass media (HIV/AIDS)                                       HIV/AIDS   National policy        OHT

1           20       Community mobilization (HIV/AIDS)                           HIV/AIDS   Community              OHT

1           21       Distribution of long lasting insecticide treated bed nets   Malaria     Community, first      OHT
                                                                                            level facility

1           22       Management of diarrhoea using Oral Rehydration Salts,       RMNCH       Community, first      OHT
                     zinc and increased intake of fluids                                    level facility

1           23       Use of improved water source within 30 minutes              WASH       Community, non-        OHT/ Excel
                                                                                            health                 (*)

1           24       Use of water connection in the home                         WASH       Community, non-        OHT/ Excel
                                                                                            health                 (*)

1           25       Improved excreta disposal (latrine/toilet)                  WASH       Community, non-        OHT/ Excel
                                                                                            health                 (*)

1           26       Hand washing with soap                                      WASH       Community, non-        OHT/ Excel
                                                                                            health                 (*)

1           27       Hygienic disposal of children's stools                      WASH       Community, non-        OHT/ Excel
                                                                                            health                 (*)

1           28        Promotion of the use of clean fuels and technologies       ENV        Community, non-        Excel
                     for cooking (**)                                                       health

Platform 2: Periodic outreach services


2           29       Measles vaccine                                             EPI         Outreach, first       OHT
                                                                                            level facility

2           30       Polio vaccine                                               EPI         Outreach, first       OHT
                                                                                            level facility

2           31       HPV vaccine                                                 EPI         Outreach, first       OHT
                                                                                            level facility

2           32       Rotavirus vaccine                                           EPI        First level facility   OHT

2           33       Pentavalent vaccine                                         EPI        First level facility   OHT

2           34       DPT vaccination                                             EPI        First level facility   OHT

2           35       Hib vaccine                                                 EPI        First level facility   OHT

2           36       Hep B vaccine to prevent liver cancer                       EPI        First level facility   OHT

2           37       BCG vaccine                                                 EPI        First level facility   OHT

2           38       Pneumococcal vaccine                                        EPI        First level facility   OHT

2           39       Yellow Fever vaccine (**)                                   EPI         Outreach, first       Excel
                                                                                            level facility




                                                                    13
2   40   Meningitis vaccine (**)                                  EPI          Outreach, first        Excel
                                                                              level facility

2   41   Japanese Encephlopathy Vaccine (**)                      EPI          Outreach, first        Excel
                                                                              level facility

2   42   Neglected Tropical Diseases: Preventive chemotherapy     NTD          Community,             Excel
         (PC) including post-PC surveillance (**)                             outreach

2   43   Neglected Tropical Diseases: Vector management (**)      NTD          Community,             Excel
                                                                              outreach

2   44   Neglected Tropical Diseases: Disease management          NTD          Community,             Excel
         including active case finding (**)                                   outreach, first level
                                                                              facility

2   45   Vector control for malaria                               Malaria     Community               Excel

2   46   Chemoprevention in vulnerable populations (**)           Malaria     Outreach                Excel

2   47   Clean practices and immediate essential newborn care     RMNCH       Community               OHT
         (home)

2   48   Family planning                                          RMNCH        Community,             OHT
                                                                              outreach, first level
                                                                              facility

2   49   Outreach to injecting drug users                         HIV/AIDS     Community,             OHT
                                                                              outreach

2   50   Needle exchange for injecting drug users                 HIV/AIDS     Community,             OHT
                                                                              outreach

2   51   Interventions focused on female sex workers              HIV/AIDS     Community,             OHT
                                                                              outreach

2   52   Interventions focused on men who have sex with men       HIV/AIDS     Community,             OHT
                                                                              outreach

2   53   Condoms for HIV/AIDS                                     HIV/AIDS    Community               OHT

2   54   Iodine supplementation for pregnant women and for        Nutrition    Community, first       Excel
         children (**)                                                        level facility

2   55   Daily iron and folic acid supplementation (pregnant      Nutrition    Community, first       OHT
         women)                                                               level facility

2   56   Daily Iron folic acid, postpartum, anaemic women (**)    Nutrition    Community, first       OHT
                                                                              level facility

2   57   Breastfeeding counselling and support                    Nutrition    Community,             OHT
                                                                              outreach, first level
                                                                              facility

2   58   Complementary feeding counselling and support            Nutrition    Community, first       OHT
                                                                              level facility

2   59   Nurturing care counselling for early child development   RMNCH        Community, first       Excel
                                                                              level facility

2   60   Support for maternal depression                          RMNCH        Community, first       Excel
                                                                              level facility

2   61   Home fortification of food with multiple micronutrient   Nutrition    Community, first       OHT
         powders (children 6-23 months)                                       level facility



                                                       14
2            62        Vitamin A supplementation in infants and children 6-59      Nutrition     Community, first       OHT
                       months                                                                   level facility

2            63        Intermittent iron supplementation in children               Nutrition    Community               OHT

2            64        Daily iron supplementation for children 6 to 23 months      Nutrition    Community               OHT
                       (where anaemia is >= 40%)

2            65        Management of moderate acute malnutrition (children)        Nutrition     Community, first       OHT
                                                                                                level facility

2            66        Feeding counselling and support for infants and young       Nutrition    Outreach                OHT
                       children in emergency situations (**)

2            67        Offer to help quit tobacco use: Brief intervention          NCD          First level facility    OHT/Excel
                                                                                                                        (*)

2            68        Screening and brief intervention for hazardous and          NCD          First level facility    OHT/Excel
                       harmful alcohol use                                                                              (*)

2            69        Physical inactivity: Brief advice as part of routine care   NCD          First level facility    OHT/Excel
                                                                                                                        (*)

2            70        Basic palliative care for breast, cervical and colorectal   NCD/cancer   Community,              Excel
                       cancer (**)                                                              outreach, hospital
                                                                                                outpatient

Platform 3: First level clinical services

3            71        Safe abortion                                               RMNCH        First level facility,   OHT
                                                                                                hospital outpatient,
                                                                                                hospital inpatient

3            72        Post-abortion case management                               RMNCH        First level facility,   OHT
                                                                                                hospital outpatient,
                                                                                                hospital inpatient

3            73        Ectopic case management (medical)                           RMNCH        Hospital inpatient      OHT

3            74        Tetanus toxoid immunization (pregnant women)                RMNCH        First level facility    OHT

3            75        Syphilis detection and treatment (pregnant women)           RMNCH        First level facility    OHT

3            76        Basic antenatal care ( 4 visits)                            RMNCH        First level facility,   OHT
                                                                                                hospital outpatient

3            77        Hypertensive disorder case management                       RMNCH        First level facility,   OHT
                                                                                                hospital outpatient,
                                                                                                hospital inpatient

3            78        Management of pre-eclampsia (Magnesium sulphate)            RMNCH        First level facility,   OHT
                                                                                                hospital outpatient,
                                                                                                hospital inpatient

3            79        Labor and delivery management - normal delivery             RMNCH        First level facility,   OHT
                                                                                                hospital outpatient,
                                                                                                hospital inpatient

3            80        Active management of the 3rd stage of labour                RMNCH        First level facility,   OHT
                                                                                                hospital outpatient,
                                                                                                hospital inpatient

3            81        Management of eclampsia (Magnesium sulphate)                RMNCH        First level facility,   OHT
                                                                                                hospital outpatient,
                                                                                                hospital inpatient



                                                                       15
3   82    Neonatal resuscitation (institutional)                 RMNCH     First level facility,   OHT
                                                                           hospital outpatient,
                                                                           hospital inpatient

3   83    Treatment of local infections (Newborn)                RMNCH     First level facility,   OHT
                                                                           hospital outpatient

3   84    Kangaroo mother care                                   RMNCH     First level facility,   OHT
                                                                           hospital outpatient

3   85    Feeding counselling and support for low-birth-weight   RMNCH     Community, First        OHT
          infants (**)                                                     level facility,
                                                                           hospital outpatient,

3   86    Antibiotics for preterm premature rupture of           RMNCH     First level facility,   OHT
          membranes (pPRoM)                                                hospital outpatient

3   87    Maternal Sepsis case management                        RMNCH     First level facility,   OHT
                                                                           hospital outpatient

3   88    Newborn sepsis - Injectable antibiotics                RMNCH     First level facility    OHT

3   89    Clean postnatal practices                              RMNCH     First level facility    OHT

3   90    Mastitis                                               RMNCH     First level facility,   OHT
                                                                           hospital outpatient

3   91    Chlorhexidine for cord care                            RMNCH     First level facility,   OHT
                                                                           hospital outpatient

3   92    Treatment of syphilis                                  RMNCH     First level facility,   Excel
                                                                           hospital outpatient

3   93    Treatment of gonorrhoea (**)                           RMNCH     First level facility,   Excel
                                                                           hospital outpatient

3   94    Treatment of chlamydia (**)                            RMNCH     First level facility,   Excel
                                                                           hospital outpatient

3   95    Treatment of trichomoniasis (**)                       RMNCH     First level facility,   Excel
                                                                           hospital outpatient

3   96    Treatment of lower abdominal pain and Pelvic           RMNCH     First level facility,   Excel
          Inflammatory Disease (PID) - lower abdominal pain                hospital outpatient
          (**)

3   97    Treatment of urinary tract infection (UTI) (**)        RMNCH     First level facility,   Excel
                                                                           hospital outpatient

3   98    Vitamin A supplementation for treatment of             RMNCH     First level facility    OHT
          xerophthalmia in women of reproductive age (**)

3   99    Vitamin A supplementation for treatment of             RMNCH     First level facility    OHT
          xerophthalmia in children (**)

3   100   Pneumonia treatment (children)                         RMNCH     First level facility    OHT

3   101   Antibiotics for treatment of dysentery in children     RMNCH     First level facility,   OHT
                                                                           hospital outpatient

3   102   Vitamin A for measles treatment (children)             RMNCH     First level facility    OHT

3   103   Intermittent preventive treatment of malaria in        Malaria   First level facility    OHT
          pregnancy (iptp)




                                                         16
3   104   Malaria diagnosis and treatment (children under five)     Malaria     First level facility    OHT

3   105   Malaria diagnosis and treatment (population aged 5        Malaria     First level facility    OHT
          years and above, including pregnant women)

3   106   TB: first line                                            TB          First level facility    Excel

3   107   TB: second line                                           TB          First level facility    Excel

3   108   Collaborative TB/HIV activities, and management of        TB          First level facility    Excel
          co-morbidities

3   109   TB: diagnostic                                            TB          First level facility    Excel

3   110   Drug substitution for injecting drug users                HIV/AIDS    First level facility    OHT

3   111   Voluntary counselling and testing                         HIV/AIDS    First level facility    OHT

3   112   Male circumcision                                         HIV/AIDS    First level facility    OHT

3   113   Prevention of mother-to-child transmission (PMTCT)        HIV/AIDS    First level facility,   OHT
                                                                                hospital outpatient

3   114   Post-exposure prophylaxis                                 HIV/AIDS    First level facility,   OHT
                                                                                hospital outpatient

3   115   ART (Second-Line Treatment) for adults                    HIV/AIDS    First level facility,   OHT
                                                                                hospital outpatient

3   116   Paediatric ART                                            HIV/AIDS    First level facility,   OHT
                                                                                hospital outpatient

3   117   Cotrimoxazole for children                                HIV/AIDS    First level facility,   OHT
                                                                                hospital outpatient

3   118   HIV/AIDS service package for transgender populations      HIV/AIDS    Outreach                Excel
          (**)

3   119   HIV/AIDS service package for prisoners (**)               HIV/AIDS    Outreach                Excel

3   120   Pre-exposure prophylaxis (PrEP) (**)                      HIV/AIDS    First level facility    Excel

3   121   Intermittent iron-folic acid supplementation              Nutrition   Community,              OHT
          (menstruating women where anaemia is public health                    outreach, first level
          problem)                                                              facility, hospital
                                                                                outpatient

3   122   Intermittent iron and folic acid supplementation (non-    Nutrition   First level facility,   OHT
          anaemic pregnant women) (**)                                          hospital outpatient

3   123   Vitamin A supplementation in pregnant women               Nutrition   First level facility,   OHT
                                                                                hospital outpatient

3   124   Calcium supplementation for prevention and treatment      Nutrition   First level facility,   OHT
          of pre-eclampsia and eclampsia                                        hospital outpatient

3   125   Nutritional care and support (HIV+ pregnant and           Nutrition   First level facility    OHT
          lactating women) (**)

3   126   Nutritional care and support for pregnant and lactating   Nutrition   First level facility    OHT
          women in emergencies

3   127   Intermittent FAF, postpartum, non-anemic pregnant         Nutrition   First level facility    OHT
          women (**)




                                                       17
3   128   Screening for risk of CVD/diabetes                        NCD          First level facility    OHT

3   129   Follow-up care for those at low risk of CVD/diabetes      NCD          First level facility    OHT
          (absolute risk: 10-20%)

3   130   Treatment for those with very high cholesterol but low    NCD          First level facility    OHT
          absolute risk of CVD/diabetes (< 20%) OHT

3   131   Treatment for those with high blood pressure but low      NCD          First level facility    OHT
          absolute risk of CVD/diabetes (< 20%)

3   132   Treatment for those with absolute risk of CVD/diabetes    NCD          First level facility    OHT
          20-30%

3   133   Treatment for those with high absolute risk of            NCD          First level facility,   OHT
          CVD/diabetes (>30%)                                                    hospital outpatient

3   134   Treatment of cases with rheumatic heart disease (with     NCD          First level facility,   OHT
          benzathine penicillin)                                                 hospital outpatient

3   135   Standard glycemic control                                 NCD          First level facility,   OHT
                                                                                 hospital outpatient

3   136   Intensive glycemic control                                NCD          First level facility,   OHT
                                                                                 hospital outpatient

3   137   Neuropathy screening and preventive foot care             NCD          First level facility,   OHT
                                                                                 hospital outpatient

3   138   Screening and Treat pre-cancerous lesions (Cervical       NCD/cancer   First level facility,   Excel
          cancer: VIA, HPV+VIA)                                                  hospital outpatient

3   139   Colorectal Cancer screening                               NCD/cancer   Community, First        Excel
                                                                                 level facility,
                                                                                 hospital outpatient

3   140   Post-cancer surveillance (breast, cervical, colorectal)   NCD/cancer   First level facility,   Excel
                                                                                 hospital outpatient

3   141   Extended palliative care for breast cancer for breast,    NCD/cancer   Community,              Excel
          cervical and colorectal cancer                                         outreach, hospital
                                                                                 outpatient

3   142   Asthma: Inhaled short acting beta agonist for             NCD          First level facility,   OHT
          intermittent asthma                                                    hospital outpatient

3   143   Asthma: Low dose inhaled beclometasone + short-           NCD          First level facility,   OHT
          acting beta 2-agonists (SABA)                                          hospital outpatient

3   144   Asthma: High dose inhaled beclometasone + short-          NCD          First level facility,   OHT
          acting beta 2-agonists (SABA)                                          hospital outpatient

3   145   Chronic obstructive pulmonary disease (COPD):             NCD          First level facility,   OHT
          Smoking cessation                                                      hospital outpatient

3   146   Chronic obstructive pulmonary disease (COPD):             NCD          First level facility,   OHT
          Inhaled salbutamol                                                     hospital outpatient

3   147   Chronic obstructive pulmonary disease (COPD): Low-        NCD          First level facility,   OHT
          dose oral theophylline                                                 hospital outpatient

3   148   Chronic obstructive pulmonary disease (COPD):             NCD          First level facility,   OHT
          Ipratropium inhaler                                                    hospital outpatient

3   149   Basic psychosocial treatment for anxiety disorders        MNS          First level facility,   OHT



                                                          18
                      (mild cases)                                                            hospital outpatient

3           150       Basic psychosocial treatment and anti-depressant            MNS         First level facility,   OHT
                      medication for anxiety disorders (moderate-severe                       hospital outpatient
                      cases)

3           151       Basic psychosocial treatment for mild depression            MNS         First level facility,   OHT
                                                                                              hospital outpatient

3           152       Basic psychosocial treatment and anti-depressant            MNS         First level facility,   OHT
                      medication of first episode moderate-severe cases                       hospital outpatient

3           153       Intensive psychosocial treatment and anti-depressant        MNS         First level facility,   OHT
                      medication of first episode moderate-severe cases                       hospital outpatient

3           154       Basic psychosocial support and anti-psychotic               MNS         First level facility,   OHT
                      medication                                                              hospital outpatient

3           155       Intensive psychosocial support and anti-psychotic           MNS         First level facility,   OHT
                      medication                                                              hospital outpatient

3           156       Basic psychosocial treatment, advice, and follow-up for     MNS         First level facility,   OHT
                      bipolar disorder, plus mood-stabilizing medication                      hospital outpatient

3           157       Intensive psychosocial intervention for bipolar disorder,   MNS         First level facility,   OHT
                      plus mood-stabilizing medication                                        hospital outpatient

3           158       Basic psychosocial support, advice, and follow-up, plus     MNS         First level facility,   OHT
                      anti-epileptic medication                                               hospital outpatient

Platform 4: Specialized care

4           159       Labor and delivery management - emergency obstetric         RMNCH       Hospital inpatient      OHT
                      care

4           160       Pre-referral management of labor complications              RMNCH       First level facility,   OHT
                                                                                              hospital outpatient

4           161       Management of obstructed labor                              RMNCH       Hospital inpatient      OHT

4           162       Antenatal corticosteroids for preterm labor                 RMNCH       Hospital inpatient      OHT

4           163       Induction of labor (beyond 41 weeks)                        RMNCH       Hospital inpatient      OHT

4           164       Newborn sepsis - Full supportive care                       RMNCH       Hospital inpatient      OHT

4           165       Treatment of postpartum hemorrhage                          RMNCH       First level facility,   OHT
                                                                                              hospital inpatient

4           166       Treatment of severe illness in children (diarrhea,          RMNCH       Hospital inpatient      OHT
                      pneumonia, malaria)

4           167       Management of severe malnutrition (children)                Nutrition   Community, First        OHT
                                                                                              level facility,
                                                                                              hospital inpatient

4           168       Retinopathy screening and photocoagulation                  NCD         First level facility,   OHT
                                                                                              hospital outpatient

4           169       Treatment of new cases of acute myocardial infarction       NCD         Hospital outpatient     OHT
                      (AMI) with aspirin

4           170       Treatment of cases with established ischaemic heart         NCD         Hospital outpatient     OHT
                      disease (IHD) and post MI




                                                                    19
4            171       Treatment for those with established cerebrovascular        NCD                Hospital outpatient     OHT
                       disease and post stroke

4            172       Mammography                                                 NCD/cancer         Hospital outpatient     Excel

4            173       Cervical cancer treatment: stage 1 to stage 4               NCD/cancer         Hospital inpatient      Excel

4            174       Colorectal cancer treatment: stage 1 to stage 4             NCD/cancer         Hospital inpatient      Excel

4            175       Breast cancer treatment: stage 1 to stage 4                 NCD/cancer         Hospital inpatient      Excel

4            176       Asthma: Theophylline + High dose inhaled                    NCD                First level facility,   OHT
                       beclometasone + SABA                                                           hospital outpatient

4            177       Asthma: Oral Prednisolone + Theophylline + High dose        NCD                First level facility,   OHT
                       inhaled beclometasone + SABA                                                   hospital outpatient

4            178       COPD: Exacerbation treatment with antibiotics               NCD                First level facility,   OHT
                                                                                                      hospital outpatient

4            179       COPD: Exacerbation treatment with oral prednisolone         NCD                First level facility,   OHT
                                                                                                      hospital outpatient

4            180       COPD: Exacerbation treatment with oxygen                    NCD                First level facility,   OHT
                                                                                                      hospital outpatient

4            181       Intensive psychosocial treatment and anti-depressant        MNS                First level facility,   OHT
                       medication for anxiety disorders (moderate-severe                              hospital outpatient
                       cases)

4            182       Intensive psychosocial treatment and anti-depressant        MNS                First level facility,   OHT
                       medication of recurrent moderate-severe cases on an                            hospital outpatient
                       episodic basis

4            183       Intensive psychosocial treatment and anti-depressant        MNS                First level facility,   OHT
                       medication of recurrent moderate-severe cases on a                             hospital outpatient
                       maintenance basis

4            184       Surgical and trauma care (***)                              Surgery            Hospital outpatient,    Excel
                                                                                                      hospital inpatient

Additional programmatic interventions incl. activities addressing socioeconomic determinants

             185       Cash transfers for girls in hyper-endemic countries with    HIV/AIDS           Non-health              Excel
                       low rates of secondary school enrolment (**)

             186       Cash transfer to poor women to deliver in facilities (**)   RMNCH              National level          Excel

             187       Programme support costs include training, monitoring,       ENV, EPI,          National level          Excel
                       supervision, programme administration costs. (**)           HIV/AIDS,
                                                                                   NCD, Malaria,
                                                                                   MNS, NTD,
                                                                                   Nutrition,
                                                                                   RMNCH,
                                                                                   Surgery, TB

Notes to table: ENV= Environmental health; EPI = Expanded Program on Immunization, MNS = Mental Health and Substance Use; NCD =
Non Communicable Disease; NTD= Neglected Tropical Diseases, OHT = OneHealth Tool, RMNCH = Reproductive, Maternal, Child and
Newborn Health.
(*) Health impact was projected within the OHT projections, while costs were modelled in Excel. (**) No health impact modelling directly
associated with this intervention. (***) Costs for surgical and trauma care is not modelled on a bottom-up patient perspective but rather
from a health systems perspective ensuring that the necessary resources are made available. Health impact is not estimated.




                                                                       20
2.3. Scenarios towards UHC
Given the uncertainty around the current capacity of health systems to absorb additional resources, we have
modelled two scenarios with differing levels of ambition. Firstly, we worked with experts in each technical area
to interpret global targets and their implications, in order to inform an ambitious scenario towards reaching
global targets by 2030. The ambitious scenario considers strengthening health systems towards global
benchmarks, and an accompanying expansion of the full package of services towards 95% coverage for most
country categories, albeit at different speeds. It implies strengthening the foundations and institutions within
health systems to enable these to support models of care that provide responsive, quality health services. It
entails addressing six essential gaps (Box S1), by modelling investments towards attainment of benchmarks
within each respective health system building block, where examples include attaining targets of facility density
within the infrastructure component, and attaining high governance scores within the component for governance
and regulation.

However, while global best practice targets can be, and have been, set for where countries should strive to be in
2030, our model recognizes that not all countries may fully achieve these targets.

Therefore, to illustrate the advancement that can be made under a more resource constrained scenario, we
designed a progress scenario which models progress towards global targets whilst taking into account limits on
absorptive capacity and health systems in distress. The purpose of the progress scenario is to illustrate a scenario
where progress towards UHC will not occur evenly on all fronts, and where not all of the SDG targets may be
met by 2030, but where significant progress can still be made, in particular through scaling up the provision of
health interventions through the lower level service delivery platforms (policy, population-wide, and periodic
schedulable and outreach delivery).



Box S1. Addressing six essential gaps for health systems 19
     •    Financing: Invest in financial engineering to build a unified and transparent financial management
          system and procurement procedures, ensuring secure and transparent financial flows and enhancing
          accountability
     •    Health workforce: Invest in pre-service education for the PHC workforce, especially education
          pathways of six months to three years, with the parallel development of deployment and retention
          strategies in rural and remote areas
     •    Pharmaceuticals/medical products: Invest in supply chains, diagnostic facilities, stocks
     •    Health information: Invest in unified underlying health information systems, including surveillance
     •    Governance: invest in local health governance systems through district health management (including
          supervision, monitoring, performance management, health facility management committees, etc.) and
          community engagement
     •    Service delivery: invest in basic infrastructure and equipment




19
   WHO (2016), FIT – Foundations, Institutions, Transformation. Information Brochure, Department of Health Systems Governance and
Financing.



                                                                21
In the “Ambitious” scenario, gaps are significantly closed by 2030, such that systems attain the aspirational
targets identified and require their corresponding level of resource use. In the “Progress” scenario, substantial
advancement is obtained, but there will still be a gap remaining for many countries by 2030 in most areas. The
progress scenario may be considered by some to be a more realistic scenario which considers the limited
capacity in many settings to absorb new funding and efficiently translate this into equitable service delivery.

Both scenarios illustrate the need for countries to proceed in a step-wise fashion. In both scale-up scenarios,
fragile states will require a certain period of stability before frontloaded capital investments in (re)constructing
health systems can take place. Moreover, as discussed above, service provision can be scaled up faster in certain
delivery platforms than in others.

In both scenarios, health systems are modelled to be scaled up towards set benchmarks by 2030. The modelling
for the three most resource intensive health system components (health workforce, infrastructure, and supply
chain) is interlinked and closely related to the scope of services provided. Other health system investments
(health information systems, Emergency Risk Management, governance and health financing) are more
independent of the service package and relate to strengthening institutions. Table S6 provides an overview of
the components that are modelled within the analysis and differences across the two scenarios.

Table S6. Investments to transform health systems

      Strategic investment area                                          Ambitious Scenario             Progress Scenario
 Health workforce: This component builds on the                     Costs relate to the additional      The progress
 recommendation by the Global Strategy on Human Resources           health workers employed,            scenario assumes a
 for Health that a ratio of 4.45 health workers per 1,000           assuming that the difference        slower pace in
 population represents a density fit for purpose for a              between the current country         additional production
 transformed workforce to reach the SDGs. Drawing upon the          baseline densities and the 2030     methods.
 model developed for the Global Strategy, targets for health        targets will be entirely closed
 worker density per population are set by country, taking into      by 2030. It also includes           The anticipated
 account rural-urban population distributions.                      continued in-service training,      expansion within the
                                                                    including all hazard training       current pre-service
 Different service delivery platforms have different health         (but whose costs are part of        education system
 workforce requirements, which are taken into account by our        Emergency Risk Management           remains the same as
 model. For instance, the necessary number of health workers        below).                             in the ambitious
 who deliver outreach services based in rural centers varies in                                         scenario, but efforts
 line with the projected % of rural population in each country.     The modelled increase in            to close the gap are
 The scale-up of the workforce is matched to the facility           workforce relates both to the       more modest (two-
 infrastructure scale-up.                                           modelled expansion within the       thirds of the gap
                                                                    current pre-service education       closed). However,
 Taking into account current pre-service training capacity, we      system, as well as additional       the production of
 model the number of additional health workers expected to be       production methods that need to     “other cadres”, who
 added to the workforce each year (assuming that doctors            be put in place in order to close   are key in providing
 require five years of training, new doctors needed to close an     the gap.                            services in service
 SDG target related gap only enter the labor market in 2021;                                            delivery platforms 1
 similarly nurses and other cadres of workers are added in 2019                                         and 2 are scaled up
 and 2017, respectively).                                                                               to close 90% of the
                                                                                                        gap by 2030,
 Our analysis also estimates the health workforce required to                                           reflecting the service
 deliver health services as part of the 5 service delivery                                              provision targets.
 platforms outlined in Figure 1, using a bottom-up approach
 built into the OneHealth Tool , whereby each intervention is
 associated with a specific health worker provider time.

 Infrastructure and equipment: Targets take into account the        Costs relate to health centers      The progress
 health services to be provided as part of each service delivery    built to reach targets of 1 per     scenario follows the
 platform, as well as projected future population growth and        12000 population (urban)/ 6000      same benchmark as
 rural-urban migration. Costs include the construction of new       population (rural); District        the ambitious
 facilities, their equipment, and the recurrent costs that these    hospitals (1 per 100,000 urban      scenario, but it is
 will accrue, including maintenance. Vehicle costs are also         and 1 per 50,000 rural              considered that
 considered, including the referral chain i.e., ambulances and      population) and Provincial          reaching the targets
 drivers. Additionally, we model improvements in making water       hospitals (1 per 100,000            would take more
 and power lines available in those facilities not yet connected,   population). Required               than 15 years, and


                                                             22
in order to expand quality of care and access to basic services.     equipment and recurrent costs         that the benchmarks
Finally, we include investments for both new facilities and a        are estimated per facility type,      will not be fully
proportion of existing facilities to meet safe hospital standards,   including vehicle costs, but do       attained by 2030.
taking actions to promote the resilience of new and existing         not include single use medical        The gap between the
hospitals and other health facilities.(1)                            devices, such as syringes or          current and the
                                                                     sutures, nor implantables or          needed number of
                                                                     assistive devices. Due to the         facilities is closed by
                                                                     complex situations and limited        two-thirds by 2030,
                                                                     years to build facilities, post-      except for health
                                                                     conflict and vulnerable               centers, where 90%
                                                                     countries (C and F1) are              of the gap is closed
                                                                     modelled as only closing 80 and       by 2030.
                                                                     90% of the gap, respectively, in
                                                                     required health facilities. The
                                                                     connection of facilities to utility
                                                                     systems include only the
                                                                     additional recurrent cost of
                                                                     utility bills, and not the cost of
                                                                     connecting water, sanitation
                                                                     lines to facilities.
Supply chain: Costs are modelled based on the estimated              Fixed and recurrent costs both        The progress
additional volume and value of the consumables transported           relate to the volume of               scenario entails a
through the system, which is determined by the commodities           consumables transported               lower number of
required to provide the entirety of services modelled as being       through the system, by country        commodities and
delivered in our model.                                              and year, and thus are directly       thus lower resource
                                                                     linked to the service targets         needs.
Supply chain estimates come from the numbers of                      within the model.
commodities, including medicines and medical devices, related
to delivering all health interventions across the various service
delivery platforms. This does not include shipping or insurance
costs, but does include in country recurrent costs, construction
of new warehouses, new trucks, and buffer stock. The recurrent
cost of the supply chain is considered to be a cost fraction of
the value of commodities being sent through the
system, determined by a variety of factors, including
population density, and logistics scores from the Wold Bank
Logistics Performance Index.


For temperature-sensitive vaccines we specifically estimate          For cold chain estimates, the         The progressive
costs related to cold chain, which are determined by estimating      infrastructure needs for each         scenario will follow
the cold space required to store the volumes of vaccines             country required to deliver the       the ambitious
required to be able to provide immunization services in the          vaccine coverages of the              scenario and will just
subsequent year, as the storage capacity needs to be in place        ambitious scenario are                calculate a smaller
before mobilization and supplies of vaccines can be increased.       identfied, estimating a gap of        capacity gap, based
                                                                     cold chain capacity. This gap is      on the progressive
                                                                     filled over time with new             scenario`s lower
                                                                     facilities scaled up at the same      vaccination coverage
                                                                     speed as similar-complexity           levels. However, it
                                                                     health facilities. Similarly, the     should be noted that
                                                                     facilities and their cold storage     the countries that
                                                                     equipment currently in                require additional
                                                                     existence in the 67 countries are     resources to replace
                                                                     deemed to be in need of               20% of their cold
                                                                     upgrading between now to              storage equipment
                                                                     2030, to incorporate new              will still require the
                                                                     technology and replace outdated       same resources in
                                                                     equipment. Almost all of this         this scenario, as the
                                                                     replacement is expected to fall       need to upgrade the
                                                                     under planned, continued              existing scenario,
                                                                     government expenditure, but it        and its urgency and
                                                                     is estimated that a small group       need, remains the
                                                                     of low income countries will          same.
                                                                     need to raise additional
                                                                     resources to finance the
                                                                     replacement of 20% of their
                                                                     existing equipment.

                                                              23
 Governance: The necessary activities of a health system              The gap between a countries          This scenario is
 required to improve its governance structure, measured as            current CPIA score and an ideal      similar to the
 improving a country`s “Country Policy and Institutional              score of 6 is identified, whereby    ambitious scenario,
 Assessment” (CPIA) Score20. These include strategic planning,        the activities related to            but sees countries
 consensus building, regulation and accreditation.                    governance are carried out so        only improve their
                                                                      that a country`s score is            governance to reach
                                                                      increased every other year,          a CPIA score of 5.
                                                                      reaching the highest level of 6
                                                                      in every country by 2030.
 Health information systems: we estimate the resources                Following the Roadmap for            The components that
 required to build up resilient information systems across the        Health Measurement for the           are linked to health
 sector. This includes a financial information system, a health       Post-2015 Agenda, the                facilities scale up
 workforce information system, a public health institute, an          ambitious scenario outlines a        line with their
 information system division in the Ministry of Health, a             series of activities and             construction or
 facility-based information system and the periodic occurrence        capabilities a fully functioning     repair in this
 of surveys. This includes specialized health workers,                health information system            scenario. Where
 equipment, and recurrent costs and governance-related                should have, Several                 components follow a
 activities. Scale up of the facility-based information system is     components are linked to the         separate scale up
 linked to the scale up of facilities.                                building or refurbishment of         pattern of
                                                                      facilities, while others follow a    progression, the
                                                                      categorization of information        speed of
                                                                      systems that transition through      improvement or
                                                                      various levels to an uppermost       scale up is reduced.
                                                                      level of health information
                                                                      system development.
 Health financing: this component estimates the resources             After an assessment of               As the resources
 required to improve health financing towards achieving               countries that have either           required are based on
 Universal Health Coverage, through strengthening the                 recently started or will shortly     estimated countries
 purchasing functions of social health insurance institutions and     start reforming their health         changes in GGHE,
 Ministries of Health who have public service provision21.            financing systems, it was            the progressive
                                                                      estimated that countries should      scenario applies the
                                                                      be spending 1 to 2% more of          same methodology
                                                                      their General Government             to a less ambitious
                                                                      Health Expenditure (GGHE) to         projection of GGHE.
                                                                      strengthen the administrative
                                                                      portions of SHIs and MoHs in
                                                                      order to achieve more effective
                                                                      reform of their health financing
                                                                      functions.
 Emergency Risk Management and post emergency and
 conflict relief:                                                     The ambitious scenario sets out      The progress
 We estimate the resources required for preparing for, and            a series of activities and targets   scenario deviates
 responding to, health emergencies and conflict. This includes        all countries should meet to best    very slightly from
 three main components:                                               minimize the effects of              the ambitious
                                                                      disasters, best handle and           scenario, with the
      •    Emergency Preparedness, risk mitigation, and               coordinate response to               only difference being
           emergency response includes general disaster               emergencies, and to establish        where resource
           preparedness and emergency management, as well as          compliance with the                  needs are linked to
           main activities related to establishing the                International Health                 other targets, such as
           International Health Regulations (2005). Costs             Regulations (2005). It also          new facilities.
           include activities required to improve a country`s         considers the rebuilding or
           capacity to minimize the effect of a disaster, and its     repairing 100% of conflict
           response to the same, including the creation of            affected health facilities and its   Risk mitigation and
           Emergency Management teams within Ministries of            related costs, as well as the        emergency response:
           health, with functional emergency operation centers        costs primarily related to health    the same level of
           for coordinated response, laboratory capacity, and         worker time in disaster-relief       investments in most
           national action plans for emergency preparedness.          situations.                          components, except
           The scale-up of laboratories is linked to the scale-up                                          a fewer number of
           of health facilities. Furthermore, all countries in our                                         labs where fewer
           sample require strengthening of their IHR core                                                  health facilities are
           capacities, moving towards “sustainable capacity” in                                            being built.

20
  World Bank CPIA building human resources rating, a part of the Social Inclusion and Equity CPIA cluster.
21
  While revenue-raising is an important part of health financing, it has negligible incremental costs for this exercise, as its
costs should be borne outside the health sector.



                                                               24
         each.
    •    Post conflict reconstruction. For the reconstruction or
         repair of health facilities after conflicts, the number                                        Post conflict
         of facilities that were either destroyed or severely                                           reconstruction: same
         damaged that are counted within the baseline of                                                costs as in the
         health facilities, and have not been fixed or rebuilt,                                         ambitious scenario.
         are modelled as being fixed and rebuilt over three
         years after a period of stability in post-conflict                                             Post-emergency
         settings. This applied to the Conflict and Vulnerable                                          relief: same methods
         category countries, where data was available.                                                  as in the ambitious
    •    Post-emergency relief entails the additional                                                   scenario, but with
         workforce costs in countries currently recovering                                              lower costs as the
         from emergencies or conflicts, as well as countries                                            volume of health
         currently in conflict where our model assumes a                                                workers is lower in
         peace agreement will hold. It is mainly comprised of                                           the progress
         a hazard pay given to health workers, the number of                                            scenario.
         which are estimated in proportion to the share of the
         population affected by conflict, or by a maximum
         quarter of workers who will be delivering services in
         such circumstances. This applies to all Conflict and
         Vulnerable category countries.
    •    It is assumed that national contingency funds and the
         World Bank Group’s Pandemic emergency financing
         facility would constitute the first line of resources
         available for responding to an emergency. Their costs
         fall outside the health sector, and are not included
         here

Health service delivery:
                                                                    Scale-up curves vary across the     The set of services is
    We consider four service delivery platforms, representing       four country typologies. Scale-     the same in both
    different modes for providing patients with information,        up towards global targets follow    scenarios. The
    counselling, essential preventive commodities, screening,       the same curves as published        differences between
    diagnosis, treatment, and follow-up.                            estimates where available.          the two lies mainly
                                                                                                        in the rate at which
    Platform 1: Policy and population-wide interventions            In general, interventions under     systems are expected
    focus on policies and information communication that can        platforms 1 and 2 are rapidly       to transform and
    be delivered to the population en masse at relatively low       scaled up across country            expand.
    cost, to support changes in behaviours among risk groups        groups, although the more
    in the population, whether for preventive purposes (i.e.,       advanced groups (HS2, HS3) do       For platforms 1 and
    reduce smoking, promote physical exercise, sleep under a        have steeper curves and attain      2, extensive progress
    mosquito net) or to ensure an appropriate response to a         benchmarks earlier than 2030,       towards universal
    health problem (i.e., ensure that a child with diarrhoea        whereas conflict and foundation     coverage is modelled
    takes oral rehydration salts and increased intake of fluids).   countries only attain               for all country
    Such interventions can be rapidly expanded at low cost          benchmarks for their groups in      settings. The model
    even as countries are in the process of building up the         the final year 2030. Moreover,      maintains fairly
    foundations of their health systems.                            conflict countries require a        rapid scale-up curves
                                                                    certain lag time before the         for interventions
    Platform 2: Periodic schedulable and outreach services:         country situation becomes           delivered as part of
    includes services which are provided routinely and              stabilised and service expansion    these two platforms,
    periodically. The key characteristic of these interventions     can start.                          and we estimate that
    is that they are standardized and have low information                                              they will attain close
    asymmetry yet require an individual transaction with a                                              to the same levels of
    qualified health worker. They require contact with health                                           coverage by 2030 as
    workers, but through brief and schedulable interventions                                            in the ambitious
    (e.g., immunization, routine antenatal care, iodine                                                 scenario. Therefore,
    supplementation). Because of their relatively high level of                                         we close the
    standardization, a number of interventions can be               Coverage pathways for               coverage gap
    delivered through health workers with short term training,      Platform 3 services varies more     between current
    meaning that they can therefore be rapidly scaled up even       substantially between the           standards and the
    in resource constrained systems.                                country groups, as coverage for     ambitious scenario
                                                                    these services is modelled to       “target” coverage by
    Platform 3: First level clinical services: includes mainly      increase slowly in initial years    90%.
    individualized health-care interventions that are specific to   in Conflict and Low income
    the patient’s needs, delivered through primary level health     countries, but to follow a linear
    facilities. Examples include treatment of sexually              scale-up in HS2 and HS3             For platforms 3 and
    transmitted infections, treatment of TB and treating and        countries.                          4 however the gap

                                                             25
      managing non-communicable diseases such as diabetes.                                              between current
      Typically these services require interaction with a health     Coverage pathways for platform     standards and the
      worker that has a certain level of skills and access to        4 are the most diverse across      ambitious scenario
      diagnostic tools, therefore requiring the foundations of       country groups, and this is also   coverage target is
      health systems to be built up before services can be           where end targets for 2030         closed by two-thirds
      expanded.                                                      diverge the most, since Conflict   (67%) within the
                                                                     and Vulnerable countries are       progress scenario.
      Platform 4: Specialized care: would typically be               expected to not be able to         This is not to say that
      delivered by highly skilled health personnel, and rely on      expand these services until        countries should
      specialized diagnostic and referral systems. Examples of       significant investments have       adopt lower targets,
      interventions include diagnosis and treatment of cancer,       been made to strengthen the        but that within our
      management of obstructed labor, and management of              foundations of their health        model, more time
      severe malnutrition. Services in this category are typically   systems, and therefore these       would be required
      heterogenous, highly individualized and transaction            countries do not attain the 95%    for countries to attain
      intensive. They have a high asymmetry of information           targets for a set of services.     the targets.
      between users and providers and require high quality
      service provision.                                             Supplementary material
                                                                     provides more detail on
                                                                     intervention targets.



 Equity considerations:
 Our model considers equity at multiple levels.                                                         The progress
                                                                                                        scenario includes
 First, many health system interventions will support efforts to     Example: governance includes       similar investments
 leave nobody behind through investments in governance and           costs for extending population     in Governance.
 health information systems.                                         participation.
                                                                                                        The progress
 Secondly, our infrastructure and health workforce models are        Example: health workforce          scenario maintains a
 structured to represent reorientation of care models towards        densities are higher in rural      focus on rural health
 close-to-client primary health care, with explicit consideration    areas.                             workers and a higher
 of rural vs urban needs.                                                                               proportion of the gap
                                                                                                        is closed for health
 Third, on the health services side, we model a rapid expansion                                         centers than for
 of “pro-poor” interventions and strategies such as bed nets and     Example: bed nets are rapidly      hospitals.
 neglected tropical disease interventions through population         scaled up in all countries.
 wide and the periodic schedulable and outreach strategies, and                                          Progress scenario
 scale these up to high coverage targets.                                                               maintains high
                                                                                                        investments in pro-
 Fourth, we incorporate costs for specific health activities for     Example: outreach services for     poor strategies.
 marginalised populations such as alcoholics, drug users,            injecting drug users.
 transgender populations and prisoners; and making health
 services adolescent -friendly.                                      Example: conditional cash          Progress scenario
                                                                     transfers for expecting mothers    maintains high
 Fifth, we model interventions to address social determinants of     to support facility-based          investments in
 health, including cash transfers, and improving environmental       delivery; general cash transfers   strategies targeting
 conditions (water and sanitation, clean cooking fuels).             to promote families to seek        vulnerable groups.
                                                                     preventive health care and for
                                                                     children to continue schooling.
                                                                                                        Conditional cash
                                                                                                        transfers for skilled
                                                                                                        delivery are included
                                                                                                        in the Progress
                                                                                                        scenario.




2.4 Scale-up curves

Table S4 highlighted typical service characteristics which identify which specific constraints related to the
organization of service provision are present. In low and middle income countries where health systems face
problems of inadequate human resources, services with high transaction intensity will need to be expanded at a
more limited pace. Our model designs a pathway for each country through which health systems are


                                                               26
strengthened and service coverage is expanded. Given that our model uses year-specific projections for the
number of people reached with each intervention, we designed stylized scale-up curves to model the potential
expansion of service coverage within each platform and across different settings, exemplified through the four
country groups.

In theory, the scale up of service provision can follow a wide range of curves. Figure S3 presents a variety of
curves used within our model.

Figure S3. Types of health service scale-up curves used in the modelling


     Coverage (%)




                                                                                   Year


Each intervention is linked to one of three families of scale-up curves “s”. The first two platforms (policy and
population wide interventions and periodic outreach services) are both associated with scale-up curve type “s1”
which indicates that relatively rapid scale-up is possible.22

The expansion of facility based care is associated with scale-up curve type “s2” which is less frontloaded than
“s1” because service expansion relies on already having a strengthened health system with functioning and
accessible facilities that are adequately staffed and able to provide quality outpatient care. The scale-up curves
of health services for this platform therefore closely follow those of health system infrastructure and health
workforce.

Finally, specialized care interventions require relatively more specialized resources, both in terms of skilled
health workers and specialized equipment and facilities, which take longer to acquire and will rely more heavily
on investments in the health system, and as such are scaled-up within our model according to curve type “s3”.

Figure S4 shows typical stylized scale-up curves for selected interventions from the four service delivery
platforms in a stylized Health System 1 country. Each intervention is assigned to a delivery platform which is
associated with a specific shape of the scale-up curve, differentiated for the four country groups. For certain
interventions, we do not apply the stylized scale-up curves but instead use globally projected targets and scale




22
  There are two exceptions within this group however: one is policy and regulation interventions, for which another set of curves are
applied to take into account institutional build-up –see below; and the second exception are the interventions primarily funded outside of the
health sector (water, sanitation and hygiene, and clean cooking stoves which both require significant hardware investments and are therefore
also scaled up using curve type s2).



                                                                     27
up curves(such as for Rotavirus vaccine, in the example below, where we use GAVI projected targets and scale
up curves).23 Baseline coverage levels are country-specific.

Figure S4. Examples of platform- and intervention-specific scale-up curves (stylized example for a HS1
country)


       Coverage (%)




                                                                             Year



Note: Figure S4 presents data for one specific country, for the following interventions:
Platform 1: Policy and population-wide (curve s1): Hand washing with soap
Platform 2 Periodic schedulable and outreach services (curve s1): Rotavirus vaccine
Platform 3 First level clinical (curve s2): Malaria treatment in adults
Platform 4 Specialized care (curve s3): Screening for risk of cardiovascular disease/diabetes.
The Figure presents a stylized example only, for the Ambitious scenario. Every country and intervention has a unique starting point and end
point to which curves are applied.



Moreover, the scale-up curves vary across the five country typologies. As Figure S5 shows, scale-up curves are
modelled to take on different shapes depending on the country context. It should be noted, however, that the
least difference between groups within our model is for services delivered in platforms 1 and 2, where even in
the most fragile contexts we can still expect rapid progress towards universal health coverage for these types of
interventions and services, should resources be made available.




23
     GAVI, the Vaccine Alliance. 2015 Strategic demand forecast.



                                                                    28
Figure S5. Examples of scale-up curves from start to end point for services type “s2” (stylized examples
by country group)




                                                                                    Year




Table S7 outlines the stylized curves for service types s1, s2, and s3, and presents the generic SDG 2030 targets
applied within our model for the ambitious scenario.

Table S7: Stylized curves and assumptions for target coverage achieved by 2030, Ambitious scenario, by
country group

Country group          Curve type (s1, s2, s3) and Generic 2030 target for the Ambitious scenario (in parentheses)

                      s1 (platforms 1 and 2)                         s2 (platforms 1 and 3)        s3 (platform 4)

Conflict              Minor frontload from 2017 or 2019              Slow initial                  Exponential

                      (in general reach 90% coverage by 2030)        (in general countries reach   (in general countries reach
                                                                     80% coverage by 2030)         80% coverage by 2030)

Vulnerable systems    Minor frontload from 2016                      Slow initial                  Exponential


                                                                     (in general countries reach   (in general countries reach
                      (in general reach 90% coverage by 2030)
                                                                     90% coverage by 2030)         80% coverage by 2030)


Health System 1       Minor frontload from 2016                      Major frontload from 2019     Slow initial

                      (in general reach 95% coverage by 2030)        (in general countries reach   (in general countries reach
                                                                     95% coverage by 2030)         95% coverage by 2030)


Health System 2      Heavy frontload                                 Linear                        S-shaped adjusted
                                                                     (in general countries reach
                     (in general reach 95% coverage by 2028)         95% coverage by 2028)         (in general countries reach
                                                                                                   95% coverage by 2030)

Health System 3      Heavy frontload                                 Minor frontload               Linear

                     (in general reach 95% coverage by 2025)         (in general countries reach   (in general countries reach
                                                                     95% coverage by 2025)         95% coverage by 2030)




                                                                29
With respect to policy interventions targeted at reducing non communicable disease (such as policies to restrict
the use of tobacco), the model works in a step-wise fashion. The policy is either considered to be in place or not.
Countries are first classified according to their current policy status, which is assumed to correspond to a certain
level of investment. They are then modelled to shift towards full policy implementation in a given year
depending on their country group.


The coverage curves for the Ambitious scenario were designed with the ambitious SDG agenda in mind,
including targets for universal coverage of essential health interventions and their associated health impact.


Within the Progress scenario, we assume somewhat lower targets across systems as well as services:


        •     First of all, we take as our starting point that even in the more limited progress scenario, extensive
              progress can be made towards universal coverage in all settings when it comes to population-wide and
              outreach services. We therefore keep the rapid scale-up curves for interventions delivered as part of
              platforms 1 and 2, and we estimate that they will attain close to the same levels of coverage by 2030 as
              in the ambitious scenario. Therefore, we close the coverage gap between current standards and the
              ambitious scenario “target” coverage by 90% for interventions in packages 1 and 2. 24
        •     With respect to platforms 3 and 4, the gap between current standards and the ambitious scenario
              coverage target is closed by two-thirds (67%) within the progress scenario. This is not to say that
              countries should adopt lower targets, but that within our model, universal targets are not reached by
              2030, and countries will require more time for these to be reached.
        •     The 2030 targets set for coverage within the Progress scenario are thus intervention-specific and
              country-specific, depending on the current country baseline and the distance to the global benchmark.
        •     Health systems investments in the Progress scenario mirror (and drive) the assumptions in scale up of
              service delivery assumptions. For the health workforce component, a scale-up of “Other” health
              workers to close the gap by 90% is modelled, to increase the number of community health workers and
              health workers who provide outreach services as part of platforms 1 and 2, whereas the gap for the
              required numbers of nurses, midwifes and doctors is only closed by two-thirds, given that this is the
              qualified workforce needed to provide services in platforms 3 and 4, where the Progress scenario
              models a convergence towards gaps being closed by two-thirds. Similarly, the infrastructure model
              assumes that the difference between the currently existing infrastructure and the density benchmark is
              only reduced by two-thirds for hospitals, while 90% of the gap is closed for health centers..

The purpose of presenting two scenarios is to present different resource implications, to stimulate debate at the
global and country level regarding what strategies can be implemented and what targets should be set for 2030,
and to highlight the usefulness of scenario generation to compare and discuss alternative investment profiles. All
countries can achieve universality progressively, and the conditions for doing so will vary across settings.


As shown in Table S7, our model aligns the targets used within the ambitious scenario with previously
published global strategies and benchmarks. For areas where there were no published strategies with specific
targets, we applied the 2030 targets outlined in table S7, by country group, which are referred to as the “generic
SDG targets” applied within our model.




24
     Policy interventions attain the same full implementation level in the Progress scenario but in a later year than in the Ambitious scenario.



                                                                         30
Table S8: Type of coverage target applied within model to be achieved by 2030, by scenario and country
group

Programme area             Ambitious scenario2030                    Progress scenario 2030 targets
                           targets


RMNCH, Nutrition,          Generic SDG targets (80, 90, 95%          Platforms 1 and 2: close 90% of the gap to reach the targets *
WASH                       by service type and country group)
                                                                     Platforms 3 and 4: close two-thirds of the gap to reach the targets *


Child immunization         GAVI forecasts                            Closes two-thirds of the gap to GAVI forecast targets*25


HIV/AIDS                   UNAIDS Fast-track                         Closes two-thirds of the gap to reach the Fast-track targets*26


Malaria                    WHO Global Malaria Strategy               Closes two-thirds of the gap to reach targets outlined in the
                                                                     strategy.


TB                         Applies targets from the Stop TB          Applies targets from the Stop TB Partnership Global Plan to End
                           Partnership Global Plan to End TB         TB 2016-2020 for all countries.
                           2016-2020 for 64 countries.

                           Applies targets from Menzies
                           (2016) for 3 countries.27


NCD policy                 Implement policies in all countries,      Implement most policies in all countries, with a later schedule for
interventions              with schedule depending on the            implementation, depending on the group classification
                           group classification
                                                                     Certain policies are only implemented in HS2 and HS3 countries,
                                                                     such as the brief interventions for tobacco, alcohol and physical
                                                                     inactivity. Policies related to diet and salt intake are not scaled up
                                                                     in conflict and vulnerable countries.


NCD screening and          Reduce unmet need by half.                Closes two-thirds of the gap to reach the targets in the ambitious
treatment                                                            scenario.


Neglected tropical         WHO Investment case on                    Closes two-thirds of the gap to reach the targets outlined in the
disease                    neglected tropical diseases.28            NTD investment case* (except for Preventive chemotherapy (PC)
                                                                     including post-PC surveillance, where the same targets are attained
                                                                     as in the ambitious scenario)


*Gap refers to the 2030 target minus the country-specific baseline data point (2015)




25
   Portnoy et al, 2015.
26
   Stover et al, 2016.
27
   Menzies et al. (2016).
28
   WHO (2015), Investing to overcome the global impact of neglected tropical diseases.



                                                                     31
            Section 3: Start and end points within models


This section presents a summary of the sources of data for start points and 2030 targets set within our model.

Table S9: Sources for assumptions on baseline and 2030 targets within the model
Area of analysis       Source for assumptions and                     2030 targets: ambitious                  2030 targets: Progress
                       baseline data                                  scenario                                 scenario


Infrastructure


Number of facilities   Global Health Observatory data (1) and         Benchmarks for the minimum               Closes two-thirds of the gap
                       country-specific country planning and          number of facilities required to         closed in the ambitious
                       health system documents, complemented          deliver care to the population uses      scenario, except for health
                       by information from country                    the same targets as applied by WHO       centers, where 90% of the gap
                       representatives attending the country          for HLTF 2009 (2), further adjusted      is closed.
                       review meeting.                                to allow for people-centred primary
                                                                      health care. The benchmarks are
                                                                      attained only for HS1, HS2 and HS3
                                                                      countries, while these are nearly
                                                                      estimated as being reached for the C
                                                                      and V groups.

                                                                      The benchmarks are one urban
                                                                      health center per 12,000 people, one
                                                                      rural health center per 6,000 people,
                                                                      one urban district hospital per
                                                                      100,000 people, one rural district
                                                                      hospital per 50,000 people, and one
                                                                      provincial hospital per 1 million
                                                                      people.


Safe Hospitals         Costs for facilities to meet safe hospital     All new facilities are equipped to       Same methodology as in the
                       standards are based on the WHO                 meet safe fospital standards. The        ambitious scenario, but with
                       Comprehensive Safe Hospital Framework          retrofitting of existing facilities is   fewer new facilities being
                       and country specific studies. Estimates on     determined by the number of              built.
                       the baseline numbers for the share of each     facilities that have urgency to
                       type of facilities that require retrofitting   withstand hazards (thus maintaining
                       to fit safe hospital standard was provided     functionality in emergencies and
                       by expert opinion, across country types.       disasters), as measured by Natural
                                                                      Disaster Propensity2429. Both of
                                                                      these are estimated to be a fixed %
                                                                      of new building costs.


Health workforce


Number of workers      Global Health Observatory data (1),            Target human resource densities,         Closes two-thirds of the gap
                       complemented by information from               and mix of cadres, estimated as part     in the ambitious scenario,
                       country representatives attending a            of work for the Global Strategy for      except for “other” workers,
                       country review meeting.                        Human Resources for Health,              where 90% of the gap is
                                                                      Adjusted “other” workers based on        closed.
                                                                      rural population distribution, to
                                                                      reflect different country local needs.

                                                                      While final ratios vary across



29
     Component of the Index for Risk Management – INFORM, http://www.inform-index.org/. Accessed March 2016.



                                                                      32
                                                                    countries and settings within our
                                                                    model, as a general rule, the targets
                                                                    set are for 4.45 doctors, nurses and
                                                                    midwives and 2.15 “other” 30 health
                                                                    workers per 1000 people, with an
                                                                    additional 2 “other” workers per
                                                                    1000 rural population.

                                                                    The benchmarks are attained only
                                                                    for HS1, HS2 and HS3 countries,
                                                                    while these are estimated as being
                                                                    nearly reached for the C and V
                                                                    groups.


Supply chain


Fixed and Recurrent    Costs are based on the additional volume     The total volume and value of             Follows the same approach as
costs                  and value in the supply chain as a result    commodities related to the                the ambitious scenario, and
                       of additional interventions or higher        interventions modelled as being           here estimates are based on
                       intervention coverage of the model.          delivered determines costs of             the relatively smaller volume
                       There is no assessment of current volume     running the supply chain, and the         of commodities related to the
                       of commodities passing through the           required infrastructure and               lower levels of coverage of
                       system.                                      equipment (warehouses, trucks) to         interventions.
                                                                    handle these commodities.
                                                                    Estimates are derived using an
                                                                    updated version of the JSI / USAID
                                                                    deliver model.31


Cold chain             Existing cold chain equipment and            Target cold chain capacity                The projected expansion of
                       volume capacity shared by WHO/IVB            determined by projections of the          cold chain capacity is lower,
                       (Gavi grant proposals, cMYP and EVM          increased volume of immunization          since it aligns with the lower
                       results).                                    commodities, which in turn is             coverage targets for vaccines
                                                                    determined by coverage levels of          in this scenario.
                       Baseline vaccine coverage and                immunization interventions.
                       projections derived from GAVI
                       projections.


Health information systems




30
   “Others” refers to the other cadres of health workers in the WHO Global Health Workforce Statistics database, which include, dentists,
pharmacists, laboratory health workers, community and traditional health workers, and health management and support health workers.
31
   For more information, see: http://deliver.jsi.com/dlvr_content/resources/allpubs/policypapers/EstiCostGlobSuppMDG.pdf



                                                                    33
Health Facility        Specialized staff, such as demographers         We estimate the resource needs to        The progress scenario follows
Information            and statisticians, is modelled to cover a       strengthen the health information        the same methodology as the
Systems                certain population catchment, which             system at the facility level. This is    ambitious one, but reflects the
                       corresponds to that of district hospitals. It   dominated by specialized human           lower number of district
                       is then assumed that the current amount         resources not considered in the          hospitals modelled as being
                       of this staff is parallel to the existing       health workforce component above,        built or refurbished.
                       number of district level hospitals in           that contribute specifically to health
                       country. Where we assume a need to              information and surveillance
                       refurbish hospitals to meet Safe Hospital       system. Targets here follow the
                       standards, we also assume that these are        targets and scale up of new district
                       understaffed and do not have health             hospitals, and the refurbishment of
                       information system staff.                       them, where district hospitals that
                                                                       are assumed to need refurbishment
                                                                       also are assumed to not already have
                                                                       this specialized staff.


Financial              Costs are estimated for strengthening the       Specialized staff, activities and        The progress scenario follows
Information System     system of tracking financial resources in       meetings are costed which are            the same methodology, except
(FIS)                  the health system. Expert opinion from          modelled as improving the level of       for assuming slower
                       the HIS department has identified current       FIS development. Countries move          improvements and
                       level of FIS development for our 67             up levels until reaching the top level   movements up along the
                       countries.                                      of a mature financial information        different levels of FIS
                                                                       system.                                  maturity.


Surveys                Costs are estimated for the needs for 3         A full schedule of surveys is carried    A subset of countries with
                       periodic surveys beyond censuses a              out in all countries, estimating         considered difficulty to carry
                       country is already expected to carry out.       resources for household visits and       out surveys are estimated as
                       The number of each type of survey               interview costs.                         only being able to carry out
                       needed is proportional to population.                                                    half as many surveys until
                                                                                                                2020, where they are
                                                                                                                estimated as being able to
                                                                                                                carry out the full schedule of
                                                                                                                them.


Health Workforce       Costs are estimated at creating and             Specialized human resources              Countries are split as in the
Information System     managing a health workforce information         comprise the majority of costs           survey component, where a
                       system.                                         estimated, where a portion of the        subset of countries is
                                                                       target is estimated as one data clerk    estimated as only being able
                                                                       per 500,000 population, while a          to scale up to 2/3 of targets by
                                                                       team of technical professionals of       2020, and then move towards
                                                                       different sizes is estimated as          final targets in the last 10
                                                                       needed across 5 different population     years.
                                                                       brackets. Scale up follows a curve
                                                                       where all countries reach 80% of
                                                                       targets by 2024.


National Statistical   Costs are estimated at creating and             Capital investments, such as             The progress scenario varies
Office, Public         staffing the governance structure to            equipment, and recurrent costs,          from the ambitious one in the
Health Institute and   manage the health information system of         including specific specialized           same manner as in the health
Governance             a country, including a national statistical     human resources, are scaled up           workforce information system
                       office, a national public health institute,     following the same assumptions as        component.
                       and a health information department in          in the above component.
                       the ministry of health.


Governance


                       Costs are assessed for activities to            We estimate costs for activities that    Within the progress scenario,
                       improve a country`s level of governance.        will bring each country towards a        countries currently scoring
                       To measure a country`s level of                 CPIA score of 6.                         less than 5 are brought
                       governance, we employ the Country


                                                                       34
                       Policy and Institutional Assessment            The type and intensity of activities    towards a CPIA score of 5.
                       (CPIA; World Bank, 2007). Countries are        is determined by expert opinion and
                       scored in the range 1-6, where 6 equates       draws upon the same methodology
                       to optimal performance.                        used as in HLTF 2009(2). Each
                                                                      country is modelled as improving its
                                                                      CPIA score after following different
                                                                      regulatory and planning activities
                                                                      for three years, until reaching a
                                                                      CPIA score of 6.


Health Financing Policy


Strengthening the       Contact and surveys with WHO regional         Costs are estimates as a percentage     Similar approach as above, in
purchasing function,   and country offices identified countries       of GGHE, the source for which are       that within the progress
through new health     that have just embarked on health finance      projections carried out based on        scenario, only that the more
finance reform         reform or are likely to do so in the near      global growth and fiscal projection     moderate financing scenario,
                       future. Countries already undergoing           scenarios (see below for more           and resulting smaller GGHE
                       reform were not evaluated in terms of          details.)                               values, was used.
                       current progress, and assumed to be able
                       to continue reform as a continuation of        The source for the percentage share
                       current resources devoted to these             of general government health
                       processes.                                     expenditures drew upon the work
                                                                      done for Social Health Insurance
                                                                      administrative costs as part of the
                                                                      2009 High Level Task force on
                                                                      Innovative Finance, as well as
                                                                      country level National Health
                                                                      Accounts data, identifying
                                                                      administrative costs at between 1
                                                                      and 2% of GGHE.


Emergency Risk Management


Post-Conflict          The numbers of facilities destroyed or         All severely damaged or destroyed       Same as in the ambitious
Reconstruction         severely damaged in conflict were              facilities are repaired or rebuilt.     scenario.
                       identified from a variety of WHO and
                       external sources, including official
                       multiyear reconstruction plans. Several
                       plans had associated cost projections
                       which provide data on resources needed
                       to rebuild or repair severely damaged
                       facilities. We applied the ratio of these
                       costs to the cost of constructing new
                       facilities in the same country, to allow for
                       the estimation of repair and
                       reconstruction for countries for which no
                       cost data was obtained.


Emergency Relief       Costs for emergency relief in pre-existing     A country-specific proportion of        Same as in the ambitious
                       humanitarian or conflict settings was          modelled number of health workers       scenario, but reflecting the
                       estimated as being captured by a               receive additional hazard pay for the   lower density of health
                       calculation of hazard pay for health           estimated duration of emergency         workers modelled in this
                       workers not currently in place.                relief, estimated for post-conflict     scenario.
                       Information on populations affected by         contexts as being equivalent to the
                       conflict were identified from WHO              share of a country`s population
                       Humanitarian Response plans. No                affected by conflict, and as being
                       estimates or considerations of current         25% of the modelled workforce for
                       amounts of emergency relief were               non-conflict humanitarian scenarios,
                       considered for countries not currently in      such as countries recovering from




                                                                      35
                       conflict.                                    Ebola.


Laboratory Services    Based on published literature for primary    Based on the minimum number of          Closes two-thirds of the gap
                       data of existing laboratories at district,   laboratories required per capita as     identified in the ambitious
                       provincial and national level32 and          identified by the Georgetown            scenario.
                       extrapolated baseline values.                University Laboratory Capacity
                                                                    Costing Estimates Tool based on
                                                                    WHO and CDC Technical
                                                                    Guidelines for Integrated Disease
                                                                    Surveillance and Response in the
                                                                    African Region, 2010.

                                                                    The benchmarks are attained for all
                                                                    countries by 2030 within our model,
                                                                    and are considered to be one district
                                                                    level lab per 150,000 people, one
                                                                    provincial level lab per 1,500,000
                                                                    people and one national reference
                                                                    lab per country.


Emergency              Country self-assessments of compliance       Activities carried out to raise the     The methodology is the same
Preparedness and       with core capacities of the International    scores of the indicators within the     as in the ambitious scenario,
Response, and          Health Regulations were used to identify     core capacities of the International    as this component is deemed
International Health   the current starting point of countries      Health Regulations. For emergency       to be of strategic importance
Regulations (2005)     towards meeting full achievement of          preparedness and response               to all countries, and full
                       these core capacities. For non-IHR           components considered separately        attainment of targets
                       components, used national planning           from the IHRs, the WHO Health           necessary to ensure minimal
                       documents and WHO Health                     Emergencies program provided            effects of cross-border
                       Emergencies program working                  guidance on each of these (ie. The      pandemics and epidemics.
                       documents and expert opinion.                size and equipment profile of a
                                                                    national poision control center).


Demography, epidemiology and current coverage of health services


Service coverage       OHT includes pre-populated country           See tables S6 and S7 above.             See tables S6 and S7 above.
modelled within        profiles that include demographic and
OHT                    epidemiological data specific to the
                       country.33

                       Default coverage data within OHT
                       originating from DHS and MICs surveys,
                       and/or expert opinion where surveys not
                       available34.


Service coverage       Baseline coverage adopted from existing      See tables S6 and S7 above.             See tables S6 and S7 above.
modelled in Excel      documents (see table S7 above.)


Notes to table:
(1) Global Health Observatory http://www.who.int/gho/en/
(2) Constraints to Scaling Up the Health Millennium Development Goals: Costing and Financial Gap Analysis. Background Document
for the Taskforce on Innovative International Financing for Health Systems. Working Group 1: Constraints to Scaling Up and Costs
http://who.int/choice/publications/d_ScalingUp_MDGs_WHO_finalreport.pdf


32
   Elbireer et al., The Good, the bad and the unknown : quality of clinical laboratories in Kampala, Uganda. PLos One, 2013 May 30 ; 8(5),
Schroeder, Lee F. and Amukele, Timothy. Medical Laboratories in Sub-Saharan Africa That Meet International Quality Standards.
American Society for Clinical Pathology, 2014; 141 : 791-795, Scott et. al, Establishing a simple and sustainable quality assurance
program and clinical chemistry services in Eritrea. Clin Chem, 2007, Nov ; 53(11) : 1945-53, and communication from the Namibian
Institute of Pathology, and the Ministry of Health of Bhutan.
33
   For details, see www.avenirhealth.org/software-onehealth.php
34
   For immunizations, the Gavi Strategic Demand Forecast 2015 was used.



                                                                    36
             Section 4: Cost and impact projection methods


This section describes the methods and tools used to estimate the resources required and the potential impact of
expanding health intervention coverage towards the SDG targets and UHC.

4.1 Country-specific projections

The analysis considers the unique context of every country when modelling investment needs. The country-
specific cost and impact outputs take into account the demographic and epidemiological context of individual
countries, including projected urbanization,35 as well as the current health system structure, and country-specific
prices for inputs. Modelling is set up to model standards of performance, grounded in empirical data, where
possible (Table S8 above).


4.2 Defining health sector costs vs costs in other sectors (“below and above the line”)

In this paper we focus the discussion on the resource needs required in the health sector. However, within our
analysis we have also examined costs that would fall outside health sector expenditure but that have some
impact upon health. We employ terminology traditionally used within health accounts, that of “above the line”
for those costs that we classify as health sector spending, and “below the line” for costs that would not be
funded through health expenditure. Below the line costs were estimated for clean cook stoves, cash transfers for
poor populations, pre-service education of health workers, and the hardware investments required for water,
sanitation and hygiene (WASH).

4.3 Ingredients-based bottom-up costing

The general approach is an ingredients-based costing (Quantities x Prices). Within each area, we specify the
inputs required to carry out activities in order to attain the benchmarks. Inputs are defined relative to total
population, population density, or to other appropriate denominators such as number of districts or the projected
number of health facilities per country and year. Prices are country-specific, where possible (see below).

The non-use of unit costs implies that economies of scale (in terms of decreasing and/or increasing unit costs) is
not taken into account. Instead, we consider that in certain settings, such as more sparely populated rural settings,
there may be a need for more fixed resources for smaller populations than in urban settings, and as such, the
implicit cost per capita is higher in most rural settings than in urban ones. A typical example is the health
workforce and infrastructure modelling, where our model assumes a need for higher density of infrastructure
and health workers in rural areas than in urban areas.

For service delivery costs, each intervention is associated with specific inputs and prices. Cost projections are
needs-based, taking into account country-specific epidemiology and coverage trajectories. This differs
significantly from an approach which would project an increase in average per capita utilization visits and
associated costs. A needs-based approach allows us to identify which interventions drive the costs, and to model
the impact of preventive interventions on the need for curative care.

4.4 Tools

Our estimates draw upon pre-existing models and estimations, including global strategies and plans in each
respective area. Several health system models draw upon the methods used for the HLTF (2009), applying
updates of the same tools (e.g., Governance, Infrastructure) or using Excel spreadsheets designed for other
recent assessments (i.e., WHO’s Global strategy on human resources for health: workforce 2030). There is an
explicit effort, to the extent possible, to be consistent with other estimates on resource needs for the 2016-2030
period where those costs have already been made public, and to make use of the same estimates and projection


35
     United Nations Population Division, World Urbanization Prospects the 2014 revision. https://esa.un.org/unpd/wup/



                                                                     37
models, but making sure to take out costs for shared resources such as health worker time, to avoid double
counting.

Our analysis thus makes use of established tools and methods, many of which have been peer reviewed and
published. Most of the health service scale-up and related impact is modelled within the OneHealth Tool (OHT)
version 5.47, a software product whose development is overseen by the UN Inter Agency Working Group on
costing (IAWG-COSTING), and carried out by Avenir Health36. OHT includes pre-populated country profiles
that include demographic and epidemiological data specific to the country. The tool is also pre-populated with
cost assumptions around consumables, and the health workforce inputs required, per service provided. Table S5
above indicates which interventions are modelled within OHT.


OHT is developed within Spectrum which is a suite of models that aim to provide policymakers with analytical
tools to support priority setting and decision making processes. As such, OHT incorporates a variety of impact
estimation models – including the Lives Saved (LiST) tool, the FamPlan model, and a number of models for
Non-Communicable Diseases, – in order to project the costs and health impacts of scaling up specific
interventions and activities in a given country.

Health impact is estimated through the OHT impact models that are directly linked to year- and country-specific
intervention targets. Box S2 provides additional detail on the OHT models used for the analysis, as well as
additional sources of data for projected impact. The key added value from projecting service coverage within
the OHT is the linkage of separate disease impact projection models through a central demographic model,
which ensures that deaths averted are not “double counted” but also allows us to benefit from the interaction of
the interventions on different indicators in the tool (an example being a change in fertility rates from family
planning affecting the number of children in need of a measles vaccination).



Box S2. Demographic and epidemiological models included within the OneHealth Tool and used for the
analysis

For the majority of the health impact projections, we used the impact projection models built into the OneHealth
Tool.

The DemProj model includes a demographic profile for every country, based on data produced by the
Population Division of the United Nations. DemProj projects the population over time by age and sex, based on
assumptions about fertility, mortality, and migration. The population projections within DemProj are used by
the other modules to support calculations on the population in need for each intervention, the associated cost and
health impact.

The Lives saved Tool (LiST) estimates the effect on maternal and child mortality and morbidity of scaling up a
range of child and maternal health interventions, including malaria interventions. The model has been developed
under the guidance of the Child Health Epidemiology Reference Group (CHERG). The LiST model uses the
user’s inputs on projected changes in the coverage of health interventions to adjust mortality rates and cause of
death structure over time. The outputs produced include changes in population level of risk factors (such as
wasting or stunting rates) and cause-specific mortality (including neonatal, children aged 1–59 months, maternal
mortality, and stillbirths). The association between an input (change in intervention coverage) with one or more
outputs is driven by intervention-specific effectiveness for reduction of the probability of that outcome
(mortality of risk factor).

The AIDS Impact Model (AIM) uses historic data combined with user-inputted coverage targets to project the
consequences of the AIDS epidemic including: the number of people infected with HIV, AIDS deaths, and the

36
     http://who.int/choice/onehealthtool/en/



                                                       38
number of people needing treatment. The projections are based on UNAIDS estimates and projections of adult
prevalence, which is combined with information on the age and sex distribution of prevalence and progression
to death in order to estimate the number of new adult infections by age and sex. Estimates of new infant
infections are based on HIV prevalence among pregnant women and the rate of mother-to-child transmission,
which is dependent on infant feeding practices and the coverage of prophylaxis with antiretrovirals (ARVs).
New infections progress over time to a symptomatic stage where antiretroviral treatment (ART) is required.
Those who receive first-line and or second-line ART will have extended survival. People at any stage are
subject to non-AIDS mortality at the same rates as those who are not infected.

FamPlan uses targets set for family planning to model projected changes in fertility. The user can enter future
contraceptive prevalence goals along with assumptions about the proximate determinants of fertility and the
characteristics of the family planning program (method mix, source mix, discontinuation rates). The model
estimates the number of users and acceptors of different methods by source, the number of pregnancies which
are likely to terminate in spontaneous and induced abortions, and the overall fertility outcomes.

The NCD impact model estimates prevalence, and incidence of NCDs including cardiovascular disease, lung
health, diabetes, and mental, neurological, and substance abuse disorders. The population in the base year is
allocated to health states based on initial prevalence. Users can specify scale-up of preventive interventions
which reduce incidence, and curative interventions, which reduce case fatality rates. A module within the NCD
impact module calculates the prevalence of risk factors for NCDs such as tobacco use, which additionally
influences the prevalence and incidence of the diseases. Policy interventions can be implemented at differing
levels of intensity which affect the prevalence of risk factors for NCDs, and ultimately the prevalence of NCDs.

With respect to health interventions modelled outside the OneHealth Tool, we drew upon previous estimates
where available to compute the additional health impact that would be attained.

Tuberculosis:

TB-related health impact estimates draw upon the Stop TB Partnership Global Plan to End TB 2016-2020.

Cancer:

Impact projections for Cervical, Colorectal and Breast cancer were estimated in Excel using country-specific
incidence projections to 2030 (Globocan, 2016). Impact (deaths averted) for ambitious, progress and baseline
scenarios was calculated using the following equation:

                  ℎ         =   ∗  ℎ             ∗   

Where  is the cancer-stage (I-IV),   is the annual total incidence,  ℎ    is the
proportion of deaths that would be averted at 100% coverage and    is the projected annual coverage
rate.

For cervical cancer screening, impact (deaths averted) for ambitious, progress and baseline scenarios was
calculated using the following equation:

     ℎ         =  30 − 49 ∗   ! ∗  ℎ                 ∗   

Where  is the cancer-stage (I-IV),  30 − 49 is the total female population between ages 30 & 49 (UN
World Population Prospects) ,   ! is the proportion of positive cases identified by screening,
 ℎ    is the proportion of deaths that would be averted at 100% coverage and    is
the projected annual coverage rate.

Exclusion

We were unable to include impact for some health interventions due to a lack of available models. Table


                                                       39
S5indicates the list of interventions for which we have not modelled gains in mortality and morbidity.

References for Box S2.

Demproj:

USAID. (2008) DemProj - A Computer Program for Making Population Projections.

LiST:

Garnett GP, Cousens S, Hallett TB, Steketee, R, Walker N. Mathematical models in the evaluation of health
programmes. Lancet 2011; 378: 515–25.

AIM:

Stover J. (2007) AIM: a computer program for making HIV/AIDS projections and examining the social and
economic impact of AIDS. Glastonbury, CT: Futures Institute.

Stover J, Johnson P, Zaba B, et al (2008). The Spectrum projection package: improvements in estimating
mortality, ART needs, PMTCT impact and uncertainty bounds. Sex Trans Infect; 84: i24-i30.

FamPlan:

USAID. FamPlan - A Computer Program for Projecting Family Planning Requirements.
http://www.healthpolicyinitiative.com/Publications/Documents/1256_1_FampmanE.pdf

The above listed references are available together with other reference materials at: www.avenirhealth.org.




4.5 Cost projection models

Commodity costs were generated by the OneHealth Tool. For interventions not included in the OHT Excel
spreadsheets were used. Costs are country- and year-specific. We incorporated a 10% mark-up for wastage.
Costs for activities to support programme administration and scale-up were estimated for each programme
(Maternal and child health, SRHR, immunization, Nutrition, malaria, HIV/AIDS, NCDs, cancers,
Mental Health and Substance Use, neglected tropical diseases, and environmental health) using the WHO-
CHOICE standardised programme costs (www.who.int/choice) and using a tracer intervention approach for
each programme. This entailed taking the current coverage of service provision (for the tracer intervention, by
country), estimating the gap to reach universal coverage, and multiplying country-specific programme costs by
the coverage gap. These programme administration costs include costs for training health workers, monitoring
and evaluation of programme performance, supervision, information campaigns and general programme
management. Some areas already had projected programme administration costs (e.g., TB) in which case we
used the pre-existing estimates. Programme cost estimates for adolescent health -i.e., improving the quality and
accessibility of health services to provide priority health interventions for adolescents- were estimated drawing
upon the approach by Deogan et al. (2012). Costs include general programme coordination at national and
district level, development and distribution of national standards for Adolescent Friendly Health Services
(AFHS), in-service training on AFHS, information and communication activities, and upgrade of infrastructure
and equipment to adolescent friendly standards.


We included costs specifically to provide financial incentives to women seeking to deliver at formal health
facilities. We used the same methodological approach as was used for the WHO HLTF (2009) analysis. Thus
the cost for incentives was calculated based on their provision to the total eligible population, and not just the


                                                         40
incremental proportion of the population that is currently not delivering in facilities. In addition to the cost of
cash transfers, costs are also incurred for administering the program, identifying poor women and paying
providers for their services. Given the findings from various countries we assumed that 30% additional costs,
calculated as a proportion of the cash transfers, would be the absolute minimum for administration costs.


4.6 Prices

Within our model we apply prices sourced from publicly available references and databases. Where possible,
prices are differentiated by country. As a general rule, price assumptions for drugs and commodities refer to
generic drugs and the lowest (median) price selected in the international market. 37 The WHO-CHOICE
database provides country specific prices for both traded and non-traded goods. 38 Where additional prices were
needed but were not contained in the list of previously mentioned sources, we also made use of additional data
sources for prices such as construction costs, 39 vaccine prices,40 etc.

Prices are reported in 2014 USD. Prices from the WHO-CHOICE database, which were available in 2010 USD,
were inflated to 2014 using country specific inflators. When costs were drawn from other pre-existing
publications (such as NTDs), we adjusted the costs to 2014 USD.

As a general rule, price assumptions within our model do not vary with volume nor over time. Thus, for
example, there is no inbuilt consideration of volume discounts for drug purchases. Similarly, we have not
modelled an increase or decrease in future prices 41 (e.g., salaries might be expected to increase with GDP
growth, and prices of certain drugs or medicines may be expected to decrease). The reason for not modelling
changes in prices over time is uncertainty. For many current medications it is likely that biosimilars will be
forthcoming in the future patent landscape; however, predictions remain uncertain.

4.7 Modelling increases in life expectancy

Summary measures of health such as life expectancy, healthy life expectancy and healthy life years gained
provide a general assessment of country progress towards strong primary care and universal health coverage.
The OneHealth Tool (OHT) projections, including Spectrum impact modules (AIM, GOALS, LIST, DemProj,
FamPlan, NCD), produce estimates on changes to population and deaths by age, taking into account coverage of
interventions to prevent or treat various diseases. Estimates on life expectancy were calculated in Excel, drawing
upon outputs from Spectrum/OHT, complemented by additional data when required. The Spectrum model
tracks the population by single age as people are born, grow older, and die, and produces outputs on modelled
deaths by age. We used these outputs to adjust/construct standard life tables42 to estimate life expectancy at
birth, and drawing upon GBD2010 disability weights by region,43 to calculate the healthy life years gained due
to scale up of interventions

We calculated life expectancy for three scenarios: the first is life expectancy at birth in 2015, the base year of
our analysis. The second is life expectancy at birth in 2030 based on projecting current intervention
implementation forward without additional investment. The third is life expectancy at birth in 2030 projecting
the health impacts of increased investments. Comparing the life expectancy at birth under scenario with
additional investments to the projected life expectancy at birth in 2030 with a constant coverage scenario, allows


37
   MSH International Drug Price Indicator Guide http://erc.msh.org/mainpage.cfm?file=1.0.htm&module=DMP&language=English
38
   http://www.who.int/choice/cost-effectiveness/inputs/en/
39
   Data entracted from SPON`s construction costs handbooks, Compass International 2016 Construction Costs Yearbook, and IADB
Infrastructure project reports.
40
   Portnoy et al (2015), costs of vaccine programs across 94 low-and middle-income countries.
41
   Traztuzumab for treating breast cancer is an exception, where a forecasted drop in its price is taken into account.
42
   Life tables: http://www.who.int/healthinfo/statistics/LT_method.pdf?ua=1&ua=1 WHO methods and data sources for life tables 1990-
2015 (Global Health Estimates Technical Paper WHO/HIS/IER/GHE/2016.8)
43
   For Disability weights, see Salomon et al. (2012).



                                                                  41
us to estimate the LE gained through the scale-up of the interventions, whilst implicitly taking into account the
background projected increase in LE built-into the UN pop projections.

The 2030 projected life expectancy at birth within the scale-up scenarios includes the impact of scaling up care
HIV/AIDS, maternal and child health, and a set of non-communicable diseases (cardiovascular disease, diabetes,
asthma, COPD), epilepsy, and mental, neurological, and substance abuse disorders, as modelled through the
OHT. Additional data available for Cancers, TB and NTDs were available from models with the same
underlying methodology which we were able to incorporate into the calculations using an Excel-based
calculation approach.44 We additionally explicitly show the impact of avoiding still births on life expectancy
increases. Intrapartum and Antepartum stillbirths are counted differently to avoided deaths following a live birth.
A body of literature suggests that sentience begins at 28 weeks gestation, thus we would consider the fetus as a
being from this point in time and would therefore include these data in health gain calculations.45 Although
sentience exists, there appears to be consensus that each stillbirth avoided should not be valued the same as
neonatal death following live birth.46 Thus each intrapartum still birth avoided is weighted at 75% and each
antepartum stillbirth avoided is weighted at 25% of a neonatal death.

Table S10a: Modelled increase in life expectancy, selected countries, Ambitious scenario

                                                      Life
                                                  Expectancy
                                                   increase              Share of LEB increase due to conditions within model
                                       N          LEs      LEB        HIV     TB       Malaria   RMNCH       Stillbirth    NCD     MNS
By country group
Total for country subset               18         3.24       4.91      4%     11%          1%       29%            6%      46%          2%
Conflict (N=2)                         2          1.74       3.12      0%      1%          0%       57%           10%      31%          1%
Vulnerable (N=2)                       2          5.24       8.37      3%      9%         16%       57%            8%       6%          1%
HSS 1 (N=2)                            2          3.89       6.73      2%     16%          1%       50%           14%      16%          1%
HSS 2 (N=6)                            6          3.27       5.50      5%     17%          2%       43%           10%      23%          1%
HSS 3 (N=6)                             6         1.17       3.83      3%         5%       0%         7%            1%     81%          3%
By income group
LIC (N=3)                              3          4.74       8.02      3%     10%          9%       57%           12%       9%          1%
LMIC (N=10)                            10         3.13       5.31      5%     16%          1%       42%            9%      25%          1%
UMIC (N=5)                              5         1.13       3.83      3%         4%       0%         7%            1%     83%          3%


* MNS = Mental Health and Substance Use; NCD = Non Communicable Disease; RMNCH= reproductive, maternal, newborn and child
health.


We ran projections for 18 countries representing 60% of the global burden of disease (2010) and 79% of the
population of the 67-country set.47 This analysis is intended as indicative only. We report two life expectancy
results: the first compares life expectancy at birth in 2030 in the scale up scenario with life expectancy at birth in
2030 in the flatline scenario. This is a conservative estimate of life expectancy gain, and referred to in table s10
below as LES. Alternatively; we compare life expectancy at birth in 2030 in the scale up scenario with life
expectancy at birth in 2015, referred to as LEB in table s10. We report in our main results the full life expectancy
gain between 2015 and 2030. Health system investments are required even in the absence of scale up of

44
     De Vlas et al (2016); Stop Tb partnership (2015), Menzies NA et al (2016).

45
   Quereshi, Z U (2015) ; Phillips and Millum, (2015 ).
46
   Jamison DT, Shahid-Salles SA, Jamison J, et al.(2006)
47
   Bangladesh, Brazil, China, Democratic Republic of the Congo, Egypt, Ethiopia, India, Indonesia, Iran, Iraq, Mali, Mexico, Myanmar,
Nigeria, Pakistan, Philippines, Viet Nam, and Yemen.



                                                                      42
interventions to support the current status of intervention implementation in a growing population, thus we
believe that LEB is valid as without these health system investments it may not be possible to continue to
implement interventions at current scale. The proposed additional health investments would generate more than
double the life expectancy gain than is expected in the flat line scenario. Tables S10a and s10b present life
expectancy results for the 18 countries.

Table S10b: Modelled increase in life expectancy, selected countries, Progress scenario
                                        Life
                                   Expectancy
                                     increase        Share of LEB increase due to conditions within model
                          N        LEs       LEB  HIV TB       Malaria RMNCH Stillbirth          NCD MNS
By country group
Total for country subset          18      2.46   4.34    4%     12%          1%     25%           5%    51%         2%
Conflict (N=2)                    2       1.36   2.74    0%      1%          0%     54%           8%    34%         2%
Vulnerable (N=2)                  2       3.90   7.03    4%     11%         16%     55%           6%     7%         0%
HSS 1 (N=2)                       2       2.97   5.81    2%     18%          1%     49%         12%     17%         1%
HSS 2 (N=6)                       6       2.53   4.77    5%     17%          2%     40%          8%     28%         1%
HSS 3 (N=6)                       6       0.90   3.56    3%      5%          0%      2%          1%     86%         3%
By income group
LIC (N=3)                         3       3.58   6.86    3%     11%          9%     56%         11%     10%         0%
LMIC (N=10)                       10      2.42   4.60    4%     17%          1%     39%          8%     30%         1%
UMIC (N=5)                        5       0.88   3.58    2%       5%         0%      1%           1%    89%         3%




The quality as well as quantity of health impact is important. In addition to the life expectancy, the number of
healthy life years gained due to this set of interventions was estimated in the OHT projection models where
possible, and from additional sources for TB and NTD. Increases in healthy life years lived within the Spectrum
impact models are calculated based on comparisons between continuation of the status quo, and implementation
of interventions to prevent or treat diseases, resulting in more people alive and healthy, and reduced disability of
the population. Across the 67 countries, 81 million healthy life years would be gained in 2030, with a total gain
of 535 million healthy life years over the course of the SDG period (Table S11). A calculation such as this is
crucial for diseases for which treatment focusses on quality of life rather than cure. For example, mental,
neurological and substance abuse disorders contribute only 3% of projected life expectancy gain, but 15% of the
projected healthy life years gained.



Table S11: Modelled increase in healthy life years by cause, 67 countries

           HIV         TB              Malaria   RMNCH        Stillbirths   NCD     MNS        NTD         Total

  2015           -           -             -        -              -           -       -           -           -
  2016         0.2          0.0           0.0      0.2            0.0         0.3     0.5         0.4         1.6
  2017         0.6          0.2           0.0      0.7            0.1         0.4     1.0         0.8         3.8
  2018         1.1          0.4           0.1      1.5            0.2         1.2     1.6         1.3         7.2
  2019         1.7          0.8           0.1      2.5            0.3         1.8     2.2         1.8         11.1
  2020         2.4          1.6           0.2      3.7            0.5         2.5     2.8         2.2         15.9
  2021         3.2          2.5           0.3      5.1            0.7         3.3     3.4         2.7         21.3
  2022         4.0          3.3           0.4      6.7            1.0         4.2     4.1         3.2         27.0
  2023         4.7          3.7           0.5      8.6            1.4         5.3     4.8         3.7         32.8



                                                          43
  2024        5.5         3.9         0.6        10.7        1.9         6.4         5.5         4.3        38.9
  2025        6.3         4.0         0.8        12.9        2.5         7.7         6.3         4.8        45.2
  2026        7.0         4.5         1.0        15.4        3.1         9.1         7.0         4.9        51.9
  2027        7.8         4.7         1.2        18.0        3.7         10.6        7.8         4.9        58.7
  2028        8.6         4.9         1.5        20.7        4.4         12.2        8.6         5.0        65.9
  2029        9.4         5.0         1.7        23.6        5.2         13.9        9.4         5.1        73.3
  2030        10.2        5.2         2.0        26.6        6.0         15.7        10.2        5.2        81.1
  Total       72.7        44.8       10.3       156.8        31.0        94.5        75.3       50.4       535.8


It is important to note three reasons why our life expectancy impact numbers are underestimates:

Firstly, the increase in LES only captures those interventions modelled in the OHT Spectrum platform.

Secondly, the estimated increase in LEB reflects the built-in assumptions around extended longevity as in UN
pop projections. However within our projection model we scale-up health systems significantly faster than what
would happen if business as usual continued. Thus the UN pop projections do not adequately capture the extent
to which UHC is strengthened within our model.

Third, the estimated healthy life year gain is projected only through to 2030, whereas the health impact of the
many preventive actions implemented within the care package will only become apparent beyond the 15 year
time period modelled.

At the same time, some of the underlying increase in general longevity as projected in the UN pop datasets
would capture the conditions that we are explicitly modelling, without the increased intervention coverage but
capturing the forthcoming impacts of preventive interventions. Thus, there is a chance of overestimation for the
estimates for LES. We therefore consider the scenario “S” estimates to be a conservative, minimum
measurement of life expectancy gain, with the comparisons to baseline, “B”, an optimistic estimation of future
gains.




4.8 Overall limitations of the SDG cost and impact projections

The main limitation of the modelling approach used is its scope, insofar that it limits what activities and health
services to include within our resource estimates. We relied on existing models and treatment protocols to model
resource needs, including only interventions whose effectiveness has been demonstrated, for which there is a
general consensus on their appropriateness for inclusion in discussions around UHC, and for which there is
reasonable information about current country specific coverage levels. However, there are interventions which
would be important to include within a country-level comprehensive package of services and for which no
available model could be identified, such as for interventions to reduce suicide, or treatment of cancers not
included in our analysis (e.g, oral cancer, child leukaemia) or hepatitis. Other areas important for public health
which we were not able to address include assistive technologies and oral health, given little information about
current coverage levels.

In addition to health sector interventions, we have modelled out resource needs for additional areas such as
WASH and indoor air pollution. However, these investments represent only a selected part of the overall
multisectoral needs for health. In terms of reducing mortality and overall improvement in health, important gaps
within our analysis include addressing road traffic injuries, as well as other injuries, including those linked to
violence and crime, limiting exposure to chemicals, and reducing violence against women.




                                                        44
Due to overall uncertainty in relation to predicting what the future will bring, we do not model changes in
technology over time, with a few exceptions where we adapt the projections to allow for more efficient
behaviours. This includes, for example, the modelling of family planning services, where we model a shift
towards modern methods within the scale-up of general contraceptive prevalence. Another example is within the
cold chain modelling, where the model assumes a shift towards renewable energy sources (i.e., solar panels)
over time.

We do not have data on actual cost structures in countries. Therefore, our approach of adding the incremental
per capita need on top of the total current spending is an approach with many limitations. For example, health
worker salaries constitute a large share of the resource need. However, while our model projects the salaries that
health workers will receive by country, the actual wage bill structure in country may be different.

In terms of addressing the knowledge gap on current cost structures, the technical review meeting provided an
opportunity for validating prices of cost drivers. The validation process focused on the components within the
model that drive the cost and impact projections, and inputs were provided from participating countries.48




48
  Representatives from Bangladesh, Brazil, China, Egypt, Ethiopia, India, Indonesia, Iran, Mexico, Myanmar, Nigeria, Pakistan, the
Philippines, and Viet Nam were present.



                                                                   45
           Section 5. Methods for projecting available financing


5.1 Outputs

The projections developed cover the period from 2016 to 2030 and concern three health expenditure aggregates:

      1.   Total health expenditure (THE)
      2.   Domestic total health expenditure (D-THE)
      3.   General government health expenditure (GGHE)

Simple and transparent estimation methods (detailed below) were adopted to allow a common approach for all
countries using only universally available inputs such as gross domestic product forecasts and population
projections. Range estimates have been favoured over point estimates because of the inherent uncertainty
associated with long-term projections.




5.2 Inputs

The key variables and their sources used in the projections are shown in Table S10.

Table S12: Sources for assumptions on baseline and 2030 targets within our model

Acronym         Variable                                        Source
THE             Total health expenditure                        WHO Global Health Expenditure Database
D-THE           Domestic total health expenditure               WHO Global Health Expenditure Database
GGHE            General government health expenditure           WHO Global Health Expenditure Database
POP             Population                                      UN World Population Prospects, 2015 Revision, Total
                                                                Population Medium Fertility Variant
GGE             General government expenditure                  IMF, World Economic Outlook, April 2016
GDP             Gross domestic product                          IMF, World Economic Outlook, April 2016

The health expenditure input data is based on financing agents consistent with the System of Health Accounts
(SHA 1.0). It includes both current and capital health expenditure. General government health expenditure
includes social health insurance as well as foreign development assistance for health channelled through
government as budget support. It is assumed that these external resources are progressively replaced by
domestic resources.

Due to the absence of a dependable source with a complete set of GDP projections to 2030 for all WHO
member countries, the non-parametric method of bootstrapping was used to obtain GDP growth estimates
beyond the IMF’s projections to 2021 (further details provided below).



5.3 Units of Measurement

All amounts are measured in constant 2014 USD. Exchange rates are held constant for future periods based on
2014 annualised exchange rates. All aggregated results are calculated as unweighted simple averages unless
otherwise indicated.


5.4 Scenarios


We consider 4 scenarios: flatline, business as usual, moderate progress, and optimistic.



                                                        46
5.5. Estimation Methods

5.5.1     Total health expenditure (THE)
Total health expenditure, which includes, among other components, government health expenditure, social
health insurance, voluntary private health insurance, out-of-pocket spending, and aid, provides the overall
envelope of available resources for health in a country. Although THE and its per capita amount by themselves
reveal little about the quality, efficiency and equity of a country’s health care system, THE gives insights into
potential levels of attainable health care given the available resources. This is particularly important when it
comes to meeting essential health needs in severely resource constrained environments such as in conflict
settings and low income countries.



Total health expenditure for each country for years 2015-2030 in each scenario is given by




                 "#
     "#$,& = ' (    *       × . / δ1 2 3 × )$,&
                  ) $, &+,




where



    •    c is the country and t is the year
    •    GDP is gross domestic product based on IMF-WHO projections
         THE is total health expenditure
         ./ δ1 2 is a health-economy expansion function
    •
    •


This top-down approach gives THE as a function of just GDP and the share of health expenditure to the total
economy. Our flatline scenario golds the last 5-year average of THE as a share of GDP constant whereas our
business as usual scenario alters health’s share of the economy based on country specific historical trends given
by linear regression of observed health expenditure data from 1995 to 2014. The moderate progress and
optimistic scenarios are based on normative increases in THE as a share of GDP that would constitute a
favourable expansion of available resources for health, specifically, a 1% point increase over 2015 to 2030
under the moderate scenario and a 2% point increase under the optimistic scenario (e.g. THE as % of GDP
increases from an initial 3% to 4% under the moderate scenario and to 5% under the optimistic scenario. See
also the section on scenarios below).

Although the relationship between health expenditure and economic development has been studied extensively,
it should be noted that the above health-economy expansion function is not an elasticity. Without a more
complex model and the availability of projections of other predictive variables, the use of an elasticity is not
possible.




                                                        47
Figure S6. THE%GDP and GDP per capita in 2014, selected countries with population > 500,000




As illustrated in the scatterplot diagram above, the direct relationship between GDP and THE as a share of GDP
per capita is weak in part due to the presence of aid in low and middle income countries. The assumed
expansion of THE as a share of GDP under the progress and ambitious scenarios is nevertheless consistent with
theoretical literature and observed historical trends.



5.5.2     Domestic total health expenditure (D-THE)


Domestic total health expenditure, which is equal to total health expenditure minus external resources as a
source, provides an additional perspective on available resources and is a measure constructed exclusively for
this exercise.

Domestic total health expenditure for each country for years 2015-2030 in each scenario is given by




                           -"#
        -"#$,& = ' (           *     × ./ γ1 2 3 × )$,&
                            ) $, &+,




where



    •     c is the country and t is the year
    •     GDP is gross domestic product based on IMF-WHO projections
    •     D-THE is domestic total health expenditure


                                                       48
    •     ./ γ1 2 is a health-economy expansion function


As for total health expenditure discussed above, the projections of D-THE are based on a top-down approach
and consier the same equivalent scenarios as THE (see the scenario-specific analysis below for the specific
parameters).



5.5.3     General government health expenditure (GGHE)

General government health expenditure, which also encompasses social health insurance, represents a large part
of total health expenditure in most countries and plays a central role in advancing universal health coverage by
reducing financial barriers and impoverishment through prepayment and pooling. Projections of GGHE can
therefore provide some general insights into both the quantity and quality of future health spending. This
emphasis on government spending is also consistent with the Addis Ababa Action Agenda, which focuses on
the need to raise domestic resources.

Government health expenditure for each country for years 2015-2030 in each scenario is given by




                       ))#                          ))"#
        ))"#$,& = 6'7(     *     × . /8$ 2 93 × '7(      *      × . /:$ 2 93; × )$,&
                      ) $, &+,                     ))# $, &+,




where

    •     c is the country and t is the year
    •     GDP is gross domestic product based on IMF-WHO projections
    •     GGE is general government expenditure
          GGHE is general government health expenditure
          ./8$ 2 is a fiscal space expansion function
    •

          ./:$ 2 is a health prioritisation function
    •
    •

This bottom-up approach gives GGHE as a function of fiscal space, health prioritisation and GDP. Our different
scenarios are established by altering the fiscal space and health prioritisation variables (α and β) based on
different assumptions and analyses of historical data. Essentially, alpha determines by how much and how
quickly government revenues increase and beta determines how health’s share of the total government
expenditure evolves. The product of the fiscal space and health prioritisation gives the often discussed share of
GGHE to GDP.

                                      ))"#     ))#   ))"#
                                           = ↑     ×      ↑
                                       )     )    ))#
In our flatline, business as usual, and moderate progress scenarios, historical values and trends were projected
forward to establish variations of what might be considered the range of probable future available resources.
Under the flatline scenario, the 5-year average values for fiscal space and health prioritisation were held
constant. Under the business as usual scenario, a linear regression of observed health expenditure data from
1995 to 2014 was used to establish the country specific parameters of alpha and beta. If the historical trend


                                                        49
growth in fiscal space and health prioritisation was not statistically significant, the last values for fiscal space
and health prioritisation were held constant. The moderate progress scenario assumes certain improvements on
top of the continuation of historical trends and our optimistic scenario sets normative levels that would mark a
dramatic and significant positive change given different starting points (see table below and summary in section
on scenarios). The possibility that the business as usual and moderate scenarios, which are based on country-
specific historical trends, could exceed the normative levels of the optimistic scenario was not ruled out.

Table S13: Assumptions on increases in allocation towards health

                                  GGE%GDP                                          GGHE%GGE
                                  Fiscal Space                                   Health Prioritisation
                    If <20% in 2014 increase to 25% by 2030              If <7% in 2014 increase to 10% by 2030

                   If >20% and <40% in 2014, increase of 5%            If >7% and <13% increase by 3% (max 15%)
                             (max 40%) by 2030                                         by 2030
                    If >=40% in 2014, flatline (i.e. hold ratio            If >=13% in 2014 make 15% in 2030
                                  constant)                               (N.B. If >15% in 2014 reduce to 15%)




5.5.4    Gross domestic product (GDP)

Gross domestic product is a readily available key measure of economic development and a central variable to
the above outlined methods of projecting health expenditure. Despite the existence of GDP forecasts for many
countries, a dependable source with a complete set of GDP growth projections for all WHO Member Countries
out to 2030 does not exist, as far as the authors are aware. Given the considerable uncertainty and potentially
arbitrary nature of GDP predictions beyond 5-10 years, a simple, defensible method to get range estimates for
all countries was required.

On the advice of counterparts from the World Bank (WB) and the International Monetary Fund (IMF), we
prepared GDP growth projections for 2022 to 2030 with lower and upper ranges using the IMF’s projections to
2021 and each country’s historical data. The non-parametric method of bootstrapping to obtain expected average
growth rates with a lower and upper bound for years 2022-2030 was adopted for its simplicity and its non-
reliance on statistical assumptions about the normality of the data. The bootstrap was based on 1000 sample
replications (draw and replacement) and greater weighting was placed on more recent years (2011-2021)
assuming that current growth patterns would have more influence on future growth out to 2030. Specifically, the

                                                     ,                                      A
         = =  &> / ∑AB
                      C,      1995: =, =                = 0.00285; 1996 ∶ =A =                 = 0.0057
                                                    GH,                                    GH,

where

    •    W are weights of sampling probabilities (e.g. weight for 1995 observations = 0.2%, for 1996 =0.6%
         and so on)

A sensitivity analysis comparing the GDP growth rates obtained using the Bootstrap method with available
GDP growth rates from other sources showed our projected range estimates to be consistent and robust.

Where GDP forecasts to 2030 were available from an official institution of a country, such as the national
central bank, these were used in place of IMF-WHO projections.


5.6 Scenario-specific analysis



                                                                  50
5.6.1       Scenario Parameters
Four scenarios were modelled to establish the available resources that might be potentially available with
different priorities and financing policies.

     1.     Flatline – expenditure ratios remain unchanged from their current values
     2.     Business as usual – historical trends in expenditure growth continue
     3.     Moderate progress – positive trends accelerate and increase notably
     4.     Optimistic – dramatic and significant positive changes in health expenditure

As outlined below, these scenarios were modelled slightly differently for the bottom-up approach used to
estimate GGHE compared with the top-down approach used for the THE and D-THE envelopes.


Table S14: Assumptions

 Assumptions for GGHE estimates


                                GGE%GDP                        GGHE%GGE
                                                                                                 GDP                 Population
                                Fiscal Space                 Health Prioritisation

 Flatline                          Average value 2010-2014 held constant

                         Change based on country specific historical trends given by
 Business as Usual
                                  ordinary least squares over 1995-2014
                           Change based on double the magnitude of country specific
 Moderate progress
                          historical trends given by ordinary least squares over 1995-
 (variant a) – “double
                            2014. If trends are in negative direction, average trend of
 effort”
                                               income group adopted.
 Moderate progress         Change based on best performer within income group over        IMF-WHO Projections
 (variant b) – “best        1995-2014. Value for largest fiscal space expansion and                                UN Population
                                                                                                   :
 performer”                health reprioritisation may come from different countries.                           Projections : Medium
                                                                                          Low, Medium & High
                                                                If <7% in 2014,                                   Fertility Variant
                         •     If <20% in 2014,
                                                          •                                 Growth Variants
                                                                increase to 10% by
                               increase to 25% by
                                                                2030
                               2030
                                                          •     If >7% and <13% in
                         •     If >20% and <40% in
                                                                2014, increase by 3%
 Optimistic                    2014, increase of 5%
                                                                (max 15%) by 2030
                               (max 40%) by 2030
                                                          •     If >=13% in 2014 make
                         •     If >=40% in 2014,
                                                                15% in 2030 (N.B. If
                               flatline (i.e. hold ratio
                                                                >15% in 2014 reduce to
                               constant)
                                                                15%)

 Assumptions for THE and D-THE


                                                 THE%GDP
                                                                                                 GDP                Population
                                               Health-Economy

 Flatline                          Average value 2010-2014 held constant

                         Change based on country specific historical trends given by      IMF-WHO Projections
 Business as Usual                                                                                                 UN Population
                                  ordinary least squares over 1995-2014                            :
                                                                                                                Projections : Medium
 Moderate progress              Increase of 1%point by 2030 (e.g. 3% to 4%)               Low, Medium & High
                                                                                                                  Fertility Variant
                                                                                            Growth Variants
 Optimistic                    Increase of 2%points by 2030 (e.g. 3% to 5%)




5.6.2       Scale-up curves for financial projections
Rather than adopting a linear growth pathway, the change in the annual increase of available resources (i.e.
scale-up) is determined by each country’s categorisation. For conflict and the first subset of foundation countries,
the expansion of available resources is delayed and then accelerates in later years consistent with the fact these
countries need to rebuild and/or strengthen their health system foundations. HS3countries, in contrast, should in

                                                                  51
principle be able to raise and absorb additional resources rapidly with their more developed health systems and
better governance structures. Hence, the scale-up curve for transformation countries assumes early rapid growth
before tapering (i.e. front loading). HS1 and HS2 countries, which possess by definition health systems in
between those of HS3and conflict countries, exhibit a smoother and steadier growth pathway over the entire
period. The scale-up curves used for these country groups are the same as those used in the SDG costing
exercise (see figure S3).



5.6.3    Limitations of methods used to project available financing
One of the first limitations of the projections relates to the available historical data that combines current and
capital health expenditure, which should ideally be analysed and projected separately. Another important
limitation of the baseline data is the inclusion of on-budget foreign aid under general government health
expenditure. This is due to the aggregated reporting of expenditure by financing agents. This means that for
some countries, a certain amount of GGHE is actually from external sources, which complicates the discussion
of domestic revenue-raising. It is assumed as part of this study that this amount is progressively replaced by
truly domestic public revenue.

Next, the top-down projections of THE give only the size of the health resources envelope and not its
breakdown. The bottom-up estimations of GGHE are to also be treated separately and not combined with the
projections of THE. This study does not attempt to estimate other disaggregated values of THE, such as
voluntary health insurance or out-of-pocket spending, which will be influenced by future health financing
reforms among other factors. Hence, it is also not the purpose of this analysis to consider future changes to the
shares of health expenditure in relation to THE or GGHE.

Another challenge is the uncertainty of future aid flows and allocations, which are highly variable and politically
determined. This study does not provide estimates of future aid even though external resources are likely to
continue to be an important source of revenue for many countries, especially those in the conflict and foundation
groups.

Finally, it should be noted that the analysis relies on GDP growth projections that are inherently difficult to
predict far into the future. To mitigate this, the study uses the best available forecasts from the IMF then builds
upon these by constructing a range of average expected growth rates although this implicitly assumes the
continuation of historical long-term growth.




                                                         52
             Section 6. Review processes


This work was guided by regular consultation and review processes. A WHO and UNAIDS expert group met
monthly to provide inputs on the methodological framework, scope of analysis, and modelling approach. The
group included representatives from individual disease areas and health system building blocks.

In July 2016, WHO organised an expert review and country feedback meeting to discuss the methodology and
preliminary results of the analysis. Participants included international experts and academics, and
representatives from 14 low and middle income countries jointly accounting for more than 75% of population
covered in the analysis.49

Discussions considered the methodological approach, presentation of results and key messages. Moreover
country participants reviewed country specific input assumptions and provided feedback on these. The main
focus for country specific review was assumptions for cost drivers, as well as assumptions for health impact
projections.

Participant feedback was discussed and informed a revision in methods, where there was agreement that
methods could be improved, for example on costing of emergency risk management, as well as the projections
of available health financing. Moreover, country feedback informed the revision of data within the models with
regards to country specific data points such as the number of existing health workers, the salary cost of a nurse,
current health service coverage for specific interventions, and epidemiology such as maternal mortality or the
current prevalence of tobacco use. Feedback was only incorporated when references were provided.


             Section 7. Efficiency considerations


In recognition of the fact that current health systems in low and middle income countries may not currently be
operating at high levels of efficiency and there may be scope for improvement, we design scenarios around
efficiency, the purpose of which is to demonstrate how expectations on efficiency would affect the estimated
potential funding gap in countries. While expectations for zero wastage may be unrealistic, we consider
scenarios that consider improving current system efficiencies which would effectively free up resources and
lower overall projected costs. The converse argument is that weak capacity in low-income countries increases
the costs of making improvements, and that current inefficiencies could be assumed to also be prevalent in
future systems – at least for the short term - such that projected marginal costs should be higher.

7.1 Methods

Based on the World Health Report (WHR) 2010, we adopt specific assumptions around the share of health
expenditure that is inefficiently used. We consider two alternative scenarios both for the Progress and
Ambitious resource needs estimates, where one assumes a lower level of efficiency and another assumes a
higher level of efficiency.

7.1.1 Less efficient scenario

The WHR2010 reported that 20-40% of resources in health systems are currently wasted. As such, actual
implementation in country may incur additional resources beyond what we have modelled as our standard
scenario, which assumes that incremental costs 2016-2030 reflect efficient practice. The five categories of
inefficiencies considered in WHR2010 were human resources, medicines, hospitals, leakages, and the
intervention mix. We consider our model to specify the requirements for human resources, hospitals, and


49
     Bangladesh, Brazil, China, Egypt, Ethiopia, India, Indonesia, Iran, Mexico, Myanmar, Nigeria, Pakistan, Philippines, and Viet Nam.



                                                                      53
intervention mix rather rigorously. We therefore only model additional resource requirements due to potential
inefficiency for two categories: medicines and leakages. As such, we apply a cost increase on the incremental
cost projections for 2016-2030. For medicines we apply a constant ratio across all years (15%) for all countries,
based on WHR2010, on the projected additional commodity cost. With respect to leakages, the inefficiency loss
is estimated to range from 5-10% for LMICs, and we therefore apply a 10% mark-up on the estimated overall
additional costs. The same assumptions are applied to the cost projections for the two scale-up scenarios
(ambitious and progress).



7.1.2 More efficient scenario

For the more efficient scenario we consider that the incremental investments will reflect efficient practise (as in
the standard scenario), and we make the adjustments for increasing efficiency on the current total health
expenditure (THE) by country. We model a shift towards more effective practices over time such that the
potential efficiency savings outlined in WHR2010 are reached by 2030, that is that 3% to 5% of THE could be
saved due to increased efficiency in the procurement of medicines in low income countries, and 2% to 5% in
middle income countries. We use the lower range for the Progress scenario, and the higher range for the
Ambitious scenario. These are applied on the THE projections for 2016-2030 for each respective scenario
(ambitious and progress). Efficiency is assumed to gradually increase over time in a linear fashion such that the
high estimate is attained by 2030. Similarly for leakages, we assume a gradually more efficient system such that
by 2030 an estimated 10% of THE is released for more productive purposes in the ambitious scenario, and
similarly 5% of THE can be released in the progress scenario. This scenario is equivalent to expanding fiscal
space by reorienting health systems towards more effective practises.



7.2 Results

Here we present results for low-income countries, which are the countries that have the greatest needs to reach
SDG benchmarks. The purpose is to highlight how expectations on efficiency would affect the estimated
potential funding gap for these countries

Table S15. Projected funding gap in 2030 (difference between projected costs and projected available
funding), with inefficiency-efficiency range, Total for low income countries (N=28), by scenario, billion
US$2014

                                                   Ambitious scale-up, optimistic   Progress scale-up, moderate
                                                   financing projections            financing projections
 Regular scenario (costs)                                        16                                7
 Less efficient scenario (costs)                                 20                               10
 Increasing efficiency scenario (funds released)                  7                                4




                                                                54
Figure S7. Projected funding gap in 2030 (difference between projected costs and projected available
funding) in billions US$2014, Total for low income countries (N=28)




                     Figure S7A. Progress scale-up and moderate financing projections




                    Figure S7B. Ambitious scale-up and optimistic financing projections




                                                    55
          Section 8. Additional results, tables and figures


This section presents additional tables on projected investments and related costs.

8.1 Total additional cost by investment area



Table S16. Total Additional Costs, by investment area, 2016-2030, Billion USD 2014, Ambitious Scenario (67 countries total)

Investment area           2016   2017      2018       2019      2020       2021       2022    2023    2024    2025    2026    2027    2028    2029    2030
Infrastructure and        26.1   37.7      53.7       61.4      88.4       78.9       94.8    91.7    101.1   105.9   118.6   118.6   120.6   127.2   93.2
Equipment
Health workforce          6.3    20.2      29.6       50.6      69.5       84.4       96.5    107.0   116.0   122.8   129.0   134.9   140.4   145.5   149.9
Health information        0.0    0.2       0.1        0.4       0.5        0.7        0.6     0.6     0.5     0.5     0.5     0.7     0.6     0.6     0.6
systems
Supply chain              4.9    5.0       5.6        6.8       7.9        8.6        9.5     10.4    11.3    12.1    13.1    14.0    14.9    15.7    16.5
Health financing policy   0.0    0.1       0.2        0.3       0.5        0.6        0.8     0.9     1.1     1.3     1.5     1.8     2.0     2.3     2.6
Governance                1.4    1.6       1.6        1.6       1.7        1.8        1.7     1.8     1.8     1.8     1.9     1.8     1.8     1.8     1.8
Emergency Risk
Management                0.0    0.0       0.3        0.3       0.4        0.8        1.1     1.3     1.4     1.5     1.6     1.7     1.9     2.0     2.1
Commodities and           6.1    13.2      18.5       23.3      28.2       32.6       36.7    41.5    46.2    50.2    54.6    59.2    63.2    67.3    71.8
supplies
Emergency Relief          1.1    1.1       1.1        1.1       0.9        0.8        0.8     0.5     0.5     0.0     0.0     0.0     0.0     0.0     0.0
Reconstruction costs in   0.0    0.2       0.3        1.1       2.0        1.6        0.0     0.0     0.0     0.0     0.0     0.0     0.0     0.0     0.0
conflict and fragile
settings
Additional health         10.8   14.9      17.3       19.5      22.4       21.7       22.2    23.9    25.5    31.3    27.4    28.7    29.2    26.8    28.7
programme costs*
Total                     56.8   94.2      128.5      166.6     222.4      232.6      264.6   279.6   305.4   327.4   348.3   361.4   374.6   389.1   367.1




                                                                                       56
Table S17. Total Additional Costs, by investment area, 2016-2030, Billion USD 2014, Progress Scenario (67 countries total)

Investment area             2016        2017         2018         2019         2020        2021         2022         2023         2024         2025        2026         2027         2028         2029        2030
Infrastructure and
Equipment                   22.5        36.0         46.6         53.1         70.6        65.5         76.0         76.4         83.1         87.5        107.9        99.7         104.1        107.8       80.0
Health workforce            5.7         13.7         20.8         31.6         41.1        49.5         56.4         62.4         67.7         72.1        76.5         80.5         84.3         88.1        91.5
Health information
systems                     0.0         0.2          0.1          0.3          0.4         0.5          0.4          0.4          0.4          0.4         0.4          0.7          0.5          0.5         0.5
Supply chain                2.0         3.7          4.7          5.7          6.8         7.7          8.5          9.5          10.4         11.3        12.2         13.1         14.0         14.7        12.9
Health financing policy     0.0         0.0          0.1          0.1          0.2         0.2          0.3          0.4          0.5          0.5         0.6          0.8          0.9          1.0         1.1
Governance                  1.4         1.4          1.4          1.4          1.5         1.6          1.5          1.5          1.5          1.5         1.6          1.5          1.5          1.5         1.5
Emergency Risk
Management                  0.0         0.0          0.2          0.3          0.4         0.5          0.6          0.8          0.7          0.8         0.8          0.9          0.9          1.0         1.0
Commodities and
supplies                    4.6         10.5         14.7         18.4         22.2        25.9         29.3         33.0         37.0         40.3        43.8         47.6         51.0         53.9        57.3
Emergency Relief            1.1         1.1          1.2          1.2          1.0         0.9          0.8          0.6          0.6          0.0         0.0          0.0          0.0          0.0         0.0
Reconstruction costs in
conflict and fragile
settings                    0.0         0.2          0.3          1.1          2.0         1.6          0.0          0.0          0.0          0.0         0.0          0.0          0.0          0.0         0.0
Additional health
programme costs*            9.0         12.1         12.9         14.6         17.1        16.4         16.8         18.5         19.7         25.4        21.3         22.5         22.9         21.7        22.4
Total                       46.4        79.0         103.0        127.9        163.1       170.3        190.6        203.5        221.6        239.9       265.1        267.2        280.1        290.2       268.3


*”Additional health programme costs” include those that are programme-specific but do not refer to specific medicines, drugs or lab tests. This includes costs for programme-specific administration staff, supervision
and monitoring relative to the services for which the programme provides leadership and oversight (e.g., the national malaria programme provides implementation guidance, monitors and supervises service delivery
for malaria). It also includes mass media campaigns and demand generation.




                                                                                                          57
8.2 Strengthening key components of the health system



Table S18. Strengthening key components of the health system, 2016-2030

                                                                                Number of     Number of      Number
                                                                                                                                                    Number of                                  Number of other
                                                 Number of       Number of         Rural        Urban           of                                                    Number of nurses and
                             Total Health                                                                                                       physicians added to                             cadres of health
                                                Rural Health    Urban Health      District      District    Provincial   Total Health Workers                         midwives added to the
 Country Groups             Facilities Built,                                                                                                       the health                                workers added to the
                                                Centers Built   Centers Built    Hospitals     Hospitals     Hospitals    Added 2016-2030                               health workforce
                              2016-2030                                                                                                          workforce 2016-                               health workforce
                                                 2016-2030       2016-2030      Built 2016-   Built 2016-     Built                                                        2016-2030
                                                                                                                                                       2030                                       2016-2030
                                                                                   2030          2030       2016-2030

 Ambitious scale-up
 scenario
 All                            415,034           261,449         116,266         29,826         6803          690           23,567,016             3,014,527              10,480,173             10,072,316
 Conflict-affected states        9,870             6,108           2,730           749           278            4              519,505                85,140                198,513                 235,852
 Vulnerable systems             34,814             24,775          5,997          3,072          829           141            1,699,054              258,214                703,005                 737,834
 HS1                            53,959             40,522           6946          5,635          779           77             3,198,961              494,926               1,287,331               1,416,704
 HS2                            206,840           133,931          54,230         15,478         2738          463           13,770,370             1,837,014              5,692,770               6,240,585
 HS3                            109,551            56,113          46,364          4892          2178           4             4,379,127              339,233               2,598,554               1,441,340
 Progress scale-up
 scenario
 All                            377,948           244,020         108,117         20,540        4,776          494           14,244,625             1,882,100              6,086,331               6,276,194
 Conflict-affected states        8,883             5,646           2,534           522           177            3              360,410                52,364                142,817                 165,228
 Vulnerable systems             31,490             22,991          5.671          2,095          635           97              944,007               130,828                346,496                 466,683
 HS1                            49,544             38,394          6,656          3,890          546           58             1,545,949              215,361                565,285                 765,303
 HS2                            189,605           125,810          50,823         10,737        1,903          331            8,391,114             1,133,926              3,346,060               3,911,128
 HS3                            98,426             51,179          42,432         3,295         1,515           5             3,003,145              349,620               1,685,674                967,851




                                                                                                58
8.3 Scenarios of Additional Resource Needs vs. Additional projected Financing



Figure S8. Scenarios of Additional Resource Needs vs. Additional projected Financing, US$ 2014 billion, HS1 Countries (N=26)




                                                                                59
Figure S9. Scenarios of Additional Resource Needs vs. Additional projected Financing, US$ 2014 billion, HS2 Countries (N=16)




                                                                         60
Figure S10. Estimated incremental resource needs and projected additional available financing in year 2030, average per capita and per country group
(US$ 2014)




                                                                          61
8.4 Estimated Required Health Expenditure, as percentage (%) of projected GDP*



Table S19. Estimated Required Health Expenditure, as percentage (%) of projected GDP*
                                                                                                                                                        Total Projected Required Health
                                                       Total Health Expenditure                   Total Projected Additional Resource Needs
          Country groups            N                                                                                                                           Expenditure **
                                                         (as % of GDP) 2014                                  (as % of GDP), 2030
                                                                                                                                                              (as % of GDP) 2030
 Progress scenario                           Average               Min            Max              Average             Min            Max            Average          Min           Max
 All 67                         67            5.6%                 2.2%           10.8%             3.5%              0.1%            18.7%            6.5%          1.9%           22.6%
 Conflict-affected states (C)   4             4.1%                 2.7%           4.8%              9.3%              1.3%            18.7%           12.6%          4.5%           22.6%
 Vulnerable systems (V)         11            6.8%                 3.6%           10.8%             6.1%              3.4%            10.6%            9.5%          5.5%           12.9%
 HS1                            15            5.5%                 2.2%           8.0%              6.0%              1.5%            12.3%            8.6%          2.1%           16.5%
 HS2                            16            4.8%                 2.7%           7.5%              1.5%              0.5%            3.5%             3.6%          1.9%           5.9%
 HS3                            21            6.2%                 4.0%           9.1%              0.4%              0.1%            0.9%             4.3%          2.1%           8.9%

 Ambitious scenario                          Average               Min            Max              Average             Min            Max            Average          Min           Max
 All 67                             67        5.6%                 2.2%           10.8%             4.6%              0.2%            17.9%            7.5%          2.1%           20.5%
 Conflict-affected states (C)       4         4.1%                 2.7%           4.8%              9.7%              1.6%            17.9%           13.0%          4.7%           20.5%
 Vulnerable systems (V)
                                    11        6.8%                 3.6%           10.8%             8.5%              5.0%            14.2%           11.9%          7.2%           16.2%
 HS1
                                    15        5.5%                 2.2%           8.0%              8.0%              1.9%            16.3%           10.7%          2.5%           20.5%
 HS2                                16        4.8%                 2.7%           7.5%              2.0%              0.5%            4.4%             4.0%          2.1%           6.7%
 HS3                                21        6.2%                 4.0%           9.1%              0.5%              0.2%            1.1%             4.4%          2.1%           9.2%
*Average/Min/Max refers to the average (mean), lowest and highest country values within each group. ** Total required expenditure estimated as Total Health Expenditure (THE) in 2014 plus
the additional projected resource needs. There are two scenarios for projected THE: one moderate and one optimistic scenario. In this table the Progress scenario costs are compared with the
Moderate scenario projected THE, and the Ambitious scenario costs are compared against the Optimistic scenario projected THE; which is why the total projected required THE as a share of
GDP is higher for some countries in the Progress scenario than in the Ambitious scenario.




                                                                                             62
8.5 Resource needs by service delivery platform
We map the resource requirements to the four service delivery platforms. Intervention-specific costs such as
commodities and supplies are directly associated with a specific platform. The allocation of health workforce
costs by platform is based on bottom-up estimations of required full –time equivalent health workers per
intervention, year and country from the OneHealth Tool simulations. Costs related to infrastructure are only
included under platforms 3 (health centres) and 4 (hospitals). Overarching functions such as those related to
governance, financial administration and emergency preparedness are presented separately.

More than half of additional resources will be required to support service delivery through first level clinical
services. This is where the majority of investments in health workforce and infrastructure will be required to
ensure that primary level quality care is accessible. This is also the platform to which most health interventions
have been mapped, both preventive and curative (see section 2.2). Investments in specialized care entail setting
up and running district hospitals to provide referral care.

Figure S11. Additional resource needs by service delivery platform (Ambitious scale-up scenario, 67
countries, 2030)




Table S20 illustrates the investment profile across country typologies. Primary level care will require the bulk
of additional resources across all settings. The greatest relative investment in referral care is required in
vulnerable and low income (HS1) countries where infrastructure investments have been overlooked.

Table S20. Additional resource needs by service delivery platform, Ambitious scale-up scenario, total by
country typology, year 2030

                    Costs included
                                                                              Conflict   Vulnerable   HS1   HS2   HS3
Share of costs, per service delivery platform
Platform 1:          Commodities specific to interventions included under
Policy and           platform 1, a share of health worker time allocated to
population-wide      platform 1, Policy interventions aimed at changing
interventions        behaviour                                                     6%           6%    8%     8%   13%
Platform 2:          Commodities specific to interventions included under
Periodic             platform 2, a share of health worker time allocated to
schedulable and      platform 2, outreach activities to vulnerable groups
outreach services                                                                  9%           7%    5%     6%    3%
                     Commodities specific to interventions included under
Platform 3: First    platform 3, a share of health worker time allocated to
level clinical       platform 3; Constructing, equipping and running health
services             centres                                                      52%         51%     51%   57%   60%


                                                                  63
                     Commodities specific to interventions included under
                     platform 4, a share of health worker time allocated to
Platform 4:          platform 4; Constructing, equipping and running hospitals;
Specialized care     conditional cash transfers for facility deliveries.                 22%            27%        28%         20%      13%
                     All costs related to Supply chain, Governance, Health
Overarching          financing, Emergency preparedness, and Health Information
functions            Systems; Overall programme management.                              11%             9%        9%          10%      11%
Affordability of supporting overarching functions* with platforms 1 and 2
Number of countries for which estimated additional costs in 2030 (ambitious
scenario) exceed projected additional available total health expenditure in 2030
(optimistic scenario)                                                                        1             0        0             0           0
Number of countries for which estimated additional costs in 2030 (ambitious
scenario) exceed projected additional available general government health
expenditure in 2030 (optimistic scenario)                                                    2             2        2             0           0

*This includes the overarching costs of governance, health financing policy, health information systems, a share of supply chain costs, and
overall programme management, in addition to specific activities and commodities associated with platforms 1 and 2.



In settings where clinical services are still underdeveloped and health workforce density is low, there is still
potential for rapidly moving towards full coverage with those interventions that can be delivered through the
first two platforms: policy, population-wide, and periodic schedulable and outreach delivery. A comparison of
the projected costs with the estimates additional available total health expenditure in 2030 reveals that only one
(conflict affected) country may not be able to afford universal provision of interventions provided through these
platforms. A comparison of costs with projected additional government health expenditure indicates that 61 out
of 67 countries should be able to fully fund these interventions through government generated revenue streams.



8.6 Breakdown of disease specific resource needs
Figure S12 and Table S21 present a breakdown of projected costs that are specific to each disease control/
prevention programme area. These estimates include commodities specific to each intervention as well as
additional programmatic interventions such as in-service training, outreach and monitoring activities.

Figure S12. Additional investments in specific disease control/prevention programmes, US$ 2014 billion,
Ambitious scenario (all 67 countries)


            US$ billion




                                                                                                 Year


                                                                     64
Table S21. Additional disease prevention and control /programme-specific costs 2016-2030, by
programme area, US$ 2014 billion (all 67 countries), Ambitious scale-up scenario


Programme area
                                              US$ bn 2016-2030        Percentage share
TB                                                     57                   6%

HIV/AIDS                                               102                  11%

Malaria                                                51                   5%

Sexual and reproductive health                         50                   5%

Maternal, adolescent and child health                  42                   4%

Child immunization                                     35                   4%

Nutrition                                              46                   5%
Non communicable disease (including
                                                       421                  44%
cancer)
Mental health and substance use                        31                   3%

Neglected tropical diseases                            40                   4%

Environmental health                                   89                   9%

Total                                                  963                 100%
Numbers may not sum to the total because of rounding




                                                                 65
References for supplementary material


Deogan C, Ferguson J, Stenberg K. Resource needs for adolescent friendly health services: estimates for 74 low-
and middle-income countries. PLoS One. 2012;7(12):e51420. doi: 10.1371/journal.pone.0051420. Epub 2012
Dec 27.

De Vlas SJ, Stolk WA, le Rutte EA, et al. Concerted efforts to control or eliminate neglected tropical diseases:
how much healthwill be gained? PLoS Negl Trop Dis 2016; 10: e000438.

Fund for Peace, Fragile states index 2016. http://fsi.fundforpeace.org/

Jamison DT, Shahid-Salles SA, Jamison J, et al. Incorporating Deaths Near the Time of Birth into Estimates of
the Global Burden of Disease. In: Lopez AD, Mathers CD, Ezzati M, et al., editors. Global Burden of Disease
and Risk Factors. Washington (DC): The International Bank for Reconstruction and Development / The World
Bank; 2006. Chapter 6. Available from: https://www.ncbi.nlm.nih.gov/books/NBK11805/ Co-published by
Oxford University Press, New York.

Menzies NA, Gomez GB, Bozzani F, et al. (2016). Cost-effectiveness and resource implications of aggressive
action on TB in China, India and South Africa: a combined analysis of nine models. Lancet Global Health. 4
(11), e816-e826. doi.org/10.1016/S2214-109X(16)30265-0.

Phillips, J., & Millum, J. (2015). Valuing Stillbirths. Bioethics, 29(6), 413–423.
http://doi.org/10.1111/bioe.12120

Portnoy A, Ozawa S, Grewal S, et al., 2015. Costs of vaccine programs across 94 low- and middle-
incomecountries. Vaccine. 2015 May 7;33 Suppl 1:A99-108. doi: 10.1016/j.vaccine.2014.12.037.

Qureshi, Z. U., Millum, J., Blencowe, et al. (2015). Stillbirth should be given greater priority on the global
health agenda. The BMJ, 351, h4620. http://doi.org/10.1136/bmj.h4620

Salomon JA, Vos T, Hogan DR, et al. Common values in assessing health outcomes from disease and injury:
disability weights measurement study for the Global burden of Disease Study 2010. 2012 The Lancet 380(9859):
2129-2143.

Stop TB partnership (2015), Global Plan to End TB: the paradigm shift 2016-2020.
http://www.stoptb.org/global/plan/plan2/

Stover J, Bollinger L, Izazola JA, et al. (2016) What Is Required to End the AIDS Epidemic as a Public Health
Threat by 2030? The Cost and Impact of the Fast-Track Approach. PLoS ONE 11(5): e0154893.
doi:10.1371/journal.pone.0154893

WHO (2010). The World Health Report. Health systems financing: the path to universal coverage.

WHO (2015), Global Technical Strategy for Malaria 2016–2030.

WHO (2015), Tracking universal health coverage: first global monitoring report.

WHO (2015), Investing to overcome the global impact of neglected tropical diseases.

WHO Global Health Observatory. http://www.who.int/gho/en/ Accessed 24 May 2016.

WHO (2016), FIT – Foundations, Institutions, Transformation. Information Brochure, Department of Health
Systems Governance and Financing.

World Bank (2004). World Development Report 2004. Making Services Work for Poor People.


                                                         66
United Nations, Department of Economic and Social Affairs, Population Division, World Population Prospects:
The 2012 Revision, New York, 2013




                                                    67
