MNH PROGRAMME SUMMATIVE EVALUATION

MNH PROGRAMME

SUMMATIVE EVALUATION

Volume II

Annex VIII: Value For Money

Assessment

January 2020

Laarstraat 43

B-2840 Reet Belgium

MNH Kenya – Summative Evaluation – Vol 2. Annex VIII

hera / Final summative report / January 2020 i


hera / Final summative report / January 2020 ii

Table of Contents

1 Introduction ................................................................................................................................................... 1

2 Bungoma County Health Expenditure 2014-2018 ......................................................................................... 2

3 DFID MNH budget (overall and for Bungoma County) ................................................................................ 10

3.1 MANI, MANI Bungoma, and MANI CICF ................................................................................................... 10

3.2 CICF and CICF Bungoma ........................................................................................................................... 12

3.3 LSTM – MiH (overall expenditure) ........................................................................................................... 16

3.4 LSTM-MiH in Bungoma ............................................................................................................................ 19

4 Cost-effectiveness analysis .......................................................................................................................... 21

4.1 Additional DFID MNH expenditure vs. domestic MNH expenditure in Bungoma ................................... 21

4.2 Baseline .................................................................................................................................................. 21

4.3 Cost-effectiveness ratios .......................................................................................................................... 22

5 Additional notes on efficiency and sustainability ........................................................................................ 28

5.1 MiH / LSTM .............................................................................................................................................. 28

5.2 MANI HSS Bungoma ................................................................................................................................. 28

5.3 CICF .................................................................................................................................................. 29

List of Figures

Figure 1. Distribution of total health expenditure (THE) by Diseases / conditions, FY 2012/13 and FY

2015/16 ............................................................................................................................................................ 9

Figure 2. MiH Expenditure per Quarter (GBP) ................................................................................................ 18

Figure 3. MiH Expenditure per Quarter (in percentage) ................................................................................ 18

Figure 4. Direct Unit Cost per (direct) trainee (in-service training) (GBP)...................................................... 19

List of Tables

Table 1. Bungoma County budget 2014-2018 (KES) (Government, DPs “on budget”, and Linda Mama) ....... 4

Table 2. Share of health budget of county budget from the “Government’s equitable share of national

revenue” ........................................................................................................................................................... 5

Table 3. Share of health budget of total county budget (Percentage)............................................................. 5

Table 4. Overall Bungoma County health budget 2014-2018 (GBP) (Government, DPs “on budget”, and

Linda Mama) ..................................................................................................................................................... 7

Table 5. Bungoma County Financial Statements (KES) ..................................................................................... 8

Table 6. Bungoma County Financial Statements (KES) by Programme ............................................................ 8

Table 7. Bungoma County: health expenditure and MNH expenditure 2014-2018 (GBP) ............................ 10


hera / Final summative report / January 2020 iii

Table 8. MANI project budget and expenditure 2015-2018 (GBP) ................................................................ 10

Table 9. MANI expenditure per Consortium Partner ..................................................................................... 11

Table 10. MANI-HSS Bungoma expenditure per year (GBP and %) ................................................................ 12

Table 11. MANI-CICF expenditure 2015-2018 (GBP and %) .......................................................................... 12

Table 12. CICF Grants: Category, Duration and budget (GBP) ....................................................................... 12

Table 13. CICF: Expenditure per type of grants .............................................................................................. 14

Table 14. CICF projects in Bungoma (budgets) ............................................................................................... 14

Table 15. CICF Bungoma: Actual Expenditure per year (GBP) ........................................................................ 14

Table 16. MiH expenditure in Bungoma County ............................................................................................ 20

Table 17. Total DFID MNH Bungoma (GBP) ................................................................................................... 20

Table 18. Bungoma County Health Expenditure 2014 - 2018 ........................................................................ 21

Table 19. National and Bungoma MMR in 2014 (baseline) ............................................................................ 22

Table 20. MNH Burden of Disease 2014-2018 (IHME accessed in August 2019) ........................................... 23

Table 21. CEA Bungoma County ..................................................................................................................... 24

Table 22. CEA Bungoma County (GBP, 2015-2018) ........................................................................................ 25

Table 23. CEA Bungoma County (GBP, 2014-2016, Formative Evaluation) ................................................... 26

Table 24. CEA 3MDGF (US$, 2015, Myanmar) ............................................................................................... 27


hera / Final summative report / January 2020 iv

Abbreviations

3MDGF Three Millenium Development Goals Fund (Myanmar)

AIA Appropriation In Aid

ANC Antenatal Care

BEmONC Basic Emergency Obstetric & Neonatal Care

BoD Burden of Disease

CDO Context – Delivery - Outcome

CEA Cost-Effectiveness Analysis

CEmONC Comprehensive Emergency Obstetric & Neonatal Care

CER Cost-Effectiveness Ratio

CHMT County Health Management Team

CHV Community Health Volunteer

CICF County Innovation Challenge Fund

CS Caesarean Section

CSO Civil Society Organisation

CU Community Unit

DALY Disability Adjusted Life Year

Danida Danish Development Aid Agency

DHIS2 On-line health management information system used in Kenya

DP Development Partner

DSF Demand Side Financing

EmONC Emergency Obstetric & Neonatal Care

ESM-UBT Every Second Matters – Uterine Balloon Tamponade

EU European Union

FBO Faith Based Organisation

FGD Focus Group Discussion

GBP British Pound

GDP Gross Domestic Product

FMHC Free Maternal Health Care

HF Health Facility

HMIS Health Management Information System

HRH Human Resources for Health

HSS Health System Strengthening

ICER Incremental Cost-Effectiveness Ratios

ICT Information and Communication Technology

IDA International Development Association (World Bank)

IHME Institute for Health Metrics and Evaluation

INGO International Non-Governmental Organisation

KDHS Kenya Demographic and Health Survey

KEMSA Kenya Medical Supply Authority

KES Kenyan Shilling


hera / Final summative report / January 2020 v

KII Key Informant Interview

KMTC Kenya Medical Training College

LSTM Liverpool School of Tropical Medicine

M&E Monitoring and Evaluation

MANI Maternal and Newborn Initiative

MDG Millennium Development Goal

M&E Monitoring and Evaluation

MGH Massachusetts General Hospital

MICS Multiple Indicator Cluster Survey

MiH Making it Happen

MKU Mount Kenya University

MGH Massachusetts General Hospital

MM Maternal Mortality

MMR Maternal Mortality Ratio

MNCH Maternal, Newborn and Child Health

MNH Maternal and Newborn Health

MoH Ministry of Health

MPDSR Maternal and Perinatal Death Surveillance and Response

NHA National Health Accounts

NHIF National Hospital Insurance Fund

NMR Neonatal Mortality Rate

OR Operational Research

PBF Performance Based Financing

PMB Programme Management Budget

PNC Postnatal Care

QI Quality Improvement

RAG Red-Amber-Green (rating scale)

RMFLF Road Maintenance Fuel Levy Fund

RMNCH Reproductive, Maternal, Neonatal and Child Health

RMNCAH Reproductive, Maternal, Newborn, Child and Adolescent Health

SBA Skilled Birth Attendance

SBCC Social Behavioural Change Communication

SBR Still Birth Rate

SC Sub-county

SCI Save the Children

SCHMT Sub-County Health Management Team

TA Technical Assistance

TBA Traditional Birth Attendant

THE Total Health Expenditure

THS-UC Transforming Health System for Universal Care

ToC Theory of Change


hera / Final summative report / January 2020 vi

ToT Trainer of Trainers

TWG Technical working group

UBT Uterine Balloon Tamponade

U5MR Under Five Mortality Rate

UHC Universal Health Coverage

UN United Nations

US$ United States Dollar

VfM Value for Money

WB World Bank

WDI World Development Indicators (database)

WHO World Health Organization


hera / Final summative report / January 2020 1

1 INTRODUCTION

As mentioned in the inception report (October 2018), this VfM analysis will not repeat the approaches and

metrics used by MANI (HSS-Bungoma) and LSTM: both implementing agencies have their own chain of

results and related indicators as well as their own VfM approach used for their periodic reports to DFID. The

present VFM analysis focuses on efficiency and effectiveness ‘largo sensu’ to address evaluation questions

pertaining to efficiency, cost-effectiveness (for Bungoma county -all interventions consolidated i.e. MANI,

MiH and CICF). It also addresses some aspects of sustainability.

The VFM analysis requires detailed analyses of expenditure and outputs/outcomes in Bungoma county,

including:

• Bungoma county overall financial resources for health and more specifically for MNH, per year (2014-

2018). That information mainly comes from (a) secondary sources (e.g. the Kenya Gazette (Senate

Bills), the County Allocation of Revenue Bills, annual National and County Budget analysis from MoH,

NHIF/Linda Mama financial reports, and CHMT and SCHMT annual reports) and (b) from Bungoma

county financial statements 2015/16 till 2017/18 provided by the county Government.

• DFID MNH additional (incremental) resources to the existing domestic and other external resources,

per implementer (MANI Bungoma, MANI CICF, CICF projects, MiH/LSTM), and per year (from 2014 to

2018).

• Bungoma County and sub-counties MNH outcomes. We use the same baseline values of Maternal

Mortality Ratio and Neonatal Mortality Rate as used for the VfM section of the Formative Evaluation

Report but adapted to 2014. Coverage rates (2014-2018) of deliveries in health facilities in Bungoma

County and sub-counties come from DHIS2.

The information listed above allows for a cost-effectiveness analysis for Bungoma County as a whole, based

on the specific burden of disease (related to MNH), MNH coverage trends, and incremental MNH

expenditure (MANI, LSTM and CICF projects in Bungoma). Cost-effectiveness is evaluated according to WHO

standards whereby an intervention is considered highly cost effective if the estimated cost per DALY averted

is less than the annual per-capita GDP and considered cost-effective if the cost is less than three times the

annual per-capita GDP.

A sensitivity analysis has been applied both to the efficacy rate1 and to the attributability to DFID funding,

with a RAG rating system based on WHO thresholds and on the Kenya GDP per capita 2018 (expressed in

current US$ and converted into GBP).

Beyond the cost-effectiveness analysis that pertains to the three projects/implementers (MANI Bungoma,

MANI-CICF and MiH/LSTM) in a single county (Bungoma), the VfM analysis also addresses some efficiency2

issues separately for each of the 3 components (MiH, MANI HSS, and CICF).

1 An efficacy rate of 100% would mean that there is no maternal/neonatal death/DALY anymore among additional deliveries in health facility since the DFID support, which is very unlikely. We use efficacy rates varying from 25% to 100%. 2 Overall administrative/management/overhead cost on total cost



2 BUNGOMA COUNTY HEALTH EXPENDITURE 2014-2018

The analysis process is as follows:

• Step 1: Computation of Bungoma County annual budgets and (when available) expenditures

• Step 2: Computation of the share of the overall county budget that is allocated to the health sector

• Step 3: Conversion of financial years into calendar years (to make possible comparisons with DFID

MNH annual budgets/expenditures based on calendar years)

• Step 4: Conversion of KES amounts in GBP (using annual average exchange rates)

• Step 5: Computation of the share of the county health budget allocated to MNH

Data sources for county budget:

• 2013/14: Kenya Gazette Supplement n° 67 (Senate Bills n°1), 7th May 2013

• 2014/15: Kenya Gazette Supplement n° 52 (Senate Bills n°13), 16th April 2014

• 2015/16: Kenya Gazette Supplement n° 41 (Senate Bills n°5), 17th April 2015

• 2016/17: Kenya Gazette Supplement n° 123 (Acts n°22), Acts 2016, 27Th July 2016

• 2017/18: Kenya Gazette Supplement n° 109 (Acts n°23), Acts 2017, 13th July 2017

• 2018/19: Kenya Gazette Supplement n°18 (National Assembly Bills n°7), 6th March 2018

• 2018/19: Kenya Gazette Supplement n° 97 (Acts n°8), Acts 2018, 4th July 2018

• 2018/19: Kenya Gazette Supplement n° 128 , Senate Bills n°29, 12th October 2018

Data sources for health budget as a share of total budget (counties):

• MoH, National and County Budget Analysis FY 2014/15



• Bungoma County health financial statements 2015/16, 2016/17, and 2017/18 (budget and

expenditure) provided by Bungoma Government

Exchange rates3

2013 2014 2015 2016 2017 2018 2019

GBP / KES 135.79 144.17 150.61 132.05 131.15 131.34 128.15

Explanation on County Budget as published by the Kenya Gazette

County Budgets (e.g. 32% of total national revenues in 2016/17) comprise of:

3 https://www.poundsterlinglive.com accessed early July 2019 (annual mid exchange rate)

https://www.poundsterlinglive.com/



1. Governments’ equitable share of revenue raised nationally based on the revenue sharing formula

(column D of the First Schedule) – Transferred to the respective County Revenue Fund

2. Conditional allocations from national government revenue:

a. Health: Conditional grants for level 5 hospitals (column B of the Second Schedule)

b. Health: Special purpose grants supporting emergency medical services – only Lamu and

Tana River (column C of the Second Schedule) or “Supplement for construction of county

headquarters” (from 2017/18)

c. Health: Conditional allocations for free maternal health care (column D of the Second

Schedule) (transferred to County Governments on a reimbursement basis). This has been

cancelled after the introduction of Linda Mama in 2017/2018. Now the column D includes

the “Conditional allocation for development of Youth Polytechnics”.

d. Health: Conditional allocations to compensate county health facilities for foregone user fees

revenue (column E of the Second Schedule)

e. Health: Conditional allocations for leasing of medical equipment (column F of the Second

Schedule) (managed by the National Government)

f. Conditional allocations from the Road Maintenance Fuel Levy Fund for the repair and

maintenance of county roads (column G of the Second Schedule)

3. Conditional allocations financed by proceeds of loans or grants from development partners to each

county government (column E of the Third Schedule)

a. Health: Conditional allocation financed by a loan from the World Bank to supplement

financing for county health facilities (column B of the Third Schedule). It is now WB-THS-UC

since 2018

b. Health: Conditional allocation financed by a grant from the Government of Denmark to

supplement financing for county health facilities (column C of the Third Schedule, but

column G since 2018)

c. IDA (WB) credit (National Agriculture and Rural Inclusive Growth Project (NARIGP) (Column

C Third Schedule)

d. IDA (WB) Kenya Climate Smart Agriculture Project (KCSAP) (column D of the Third Schedule)

e. IDA (WB) credit: Kenya Devolution Support Project (KDSP) level 1 grant (column E of the

third Schedule – since 2018).

f. IDA (WB) credit : Kenya Urban Support Project (KUSP) (column F of the third Schedule –

since 2018)

g. EU Grant (Instruments for devolution advice and support) (IDEAS) (Column H of Third

Schedule)

h. Conditional allocations to ensure continuation of services financed by loans and grants from

other development partners (column D of the Third Schedule – till 2017) (managed by the

National Government)



Table 1. Bungoma County budget 2014-2018 (KES) (Government, DPs “on budget”4, and Linda Mama)

Source Category 2013/14 KES 2014/15 KES 2015/16 KES 2016/17 KES 2017/18 KES 2018/19 KES

National Government and DPs "on

bugdet"

Equitable Share of National Revenue

6,180,666,881 7,373,210,075 7,675,726,717 8,282,207,063 8,758,000,000 8,949,000,000

Conditional Allocations DANIDA Grant

14,750,000 7,375,000 22,161,451 29,362,500

Conditional Allocations Other Loans and Grants

711,417,035 182,000,000 62,188,798

Allocations for FMHC

170,630,600 189,574,801

Kenya devolution support project (level 1 grant FY2016/17 allocation)

54,474,504 58,329,310

WB THS UC 110,000,000 100,000,000

WB/IDA Loan for national agricultural and rural inclusive growth project

50,000,000 140,435,163

IDA/WB Kenya urban support project

300,977,100

IDA/WB Kenya urban support project Urban institutional grants

41,200,000

Sweden agricultural sector development support programme

22,404,685

EU Water tower protection and climate change mitigation and adaptation programme

80,000,000

Lease of medical equipment

95,744,681 95,744,681 200,000,000

Allocation to support abolishment of user fees in Health Centres and Dispensaries

38,017,014 33,282,912 32,837,307 32,837,307

4 It does not include projects and programmes directly funded by other DPs (e.g. does not include DFID MNH)



Source Category 2013/14 KES 2014/15 KES 2015/16 KES 2016/17 KES 2017/18 KES 2018/19 KES

Conditional allocation from the Road Maintenance Fuel Levy Fund (RMFLF)

107,348,422 327,025,699 235,619,376

Conditional allocation to county emergency fund

93,617,021

Conditional allocation for development of youth polytechnics

76276625 65500000

Linda Mama MNH 18,800,000 ?

Total 6,180,666,881 8,084,627,110 8,377,834,455 8,512,439,776 9,607,509,065 10,255,665,441

Linda Mama funding, although not part of the county health budget stricto sensu (reimbursements are

transferred directly to the health facility (HF) account), has been considered, with figures coming from (a)

the analysis of Linda Mama system done by Options in May 2018 (see Annex 1), and (b) from financial

information coming from the Linda Mama and NHIF websites.

Allocation to health for each budget line (see table 2 below):

• For Governments’ equitable share of revenue (first row of county budget) (from MoH budget

analysis: see “ Data source for health budget as a share of total budget (counties)” above).

Table 2. Share of health budget of county budget from the “Government’s equitable share of n ational revenue”

2014 2015 2016 2017 2018

Bungoma 28% 25% 20% 25% 25%

• For all budget lines earmarked to health: 100%

• For Linda Mama line: 100%

• For all other budget lines -2 f) and 3 c)-, etc.: 0%

Table 3. Share of health budget of total county budget (Percentage)

Source Category

% to

health

13/14

% to

health

14/15

% to

health

15/16

% to

health

16/17

% to

health

17/18

% to

health

18/19

National

Government

and DPs "on

budget"

Equitable Share of

National Revenue 28 25 20.15 25.13 25.25 25

Conditional Allocations

DANIDA Grant 100 100 100 100 100 100

Conditional Allocations

Other Loans and Grants 0 0 0 0 0 0

Allocations for FMHC 100 100 100 100 100 100



Kenya devolution support

project (level 1 grant

FY2016/17 allocation)

0 0 0 0 0 0

WB THS UC 100 100 100 100 100 100

WB/IDA Loan for national

agricultural and rural

inclusive growth project

0 0 0 0 0 0

IDA/WB Kenya urban

support project 0 0 0 0 0 0

IDA/WB Kenya urban

support project Urban

institutional grants

0 0 0 0 0 0

Sweden agricultural sector

development support

programme

0 0 0 0 0 0

EU Water tower

protection and climate

change mitigation and

adaptation programme

0 0 0 0 0 0

Lease of medical

equipment 100 100 100 100 100 100

Allocation to support

abolishment of user fees

in HC and D

100 100 100 100 100 100

Conditional allocation

from RMFLF 0 0 0 0 0 0

Conditional allocation to

county emergency fund 0 0 0 0 0 0

Conditional allocation for

development of youth

polytechnics

0 0 0 0 0 0

Linda Mama MNH 100 100 100 100 100 100

By (a) applying the conversion of financial year into calendar year (2013=half of 2013/14; 2014=half of 2013

+ half of 20145; etc. ; (b) applying the average annual exchange rates GBP/KES; and (c) applying the

percentage of budget allocated to health, we get the following table:

5 The analysis of the Bungoma county financial statements 2015/16 to 2017/18 shows that the recurrent budget execution rate is rather high (>90%), which can be explained by the fact that staff remuneration (« Compensation to Employees ») represents from 71% to 84% of the total recurrent health budget. It allows then to consider that the budget is equally spent during the 12 months of the FY.



Table 4. Overall Bungoma County health budget 2014-2018 (GBP) (Government, DPs “on budget” 6, and Linda Mama)

Source Category GBP 2013 GBP 2014 GBP 2015 GBP 2016 GBP 2017 GBP 2018

National

Governmen

t and DPs

"on budget"

Equitable

Share of

National

Revenue

6,372,29

1 12,394,705 11,254,105 13,737,136 16,365,664 16,935,606

Conditional

Allocations

DANIDA

Grant

0 0 48,968 83,775 112,606 196,147

Allocations

for FMHC

(Free

Maternal

Health

Care)

0 0 566,465 1,363,898 722,740 0

WB THS UC 0 0 0 0 419,367 799,452

Lease of

medical

equipment

0 0 317,856 362,532 365,020 1,125,874

Allocation

to support

abolishmen

t of user

fees in HC

and D

0 0 126,210 269,973 252,079 250,018

Linda Mama MNH 0 0 0 0 69,225 69,225

Total 6,372,29

1

12,394,70

5

12,313,60

4

15,817,31

4

18,306,70

0

19,376,32

2

Notes: 2013= half year budget; Linda Mama amount is for the second half of 2017 and the first 4 months of 2018

The last step consists in determining the MNH share of the county total health budget. The Bungoma county

financial statements don’t provide this kind of information (see table 5 below):

6 It does not include projects and programmes directly funded by other DPs (e.g. does not include DFID MNH)



Table 5. Bungoma County Financial Statemen ts (KES)

Budget

2015/16 AE 2015/16

Exec. Rate

Budget 2016/17 AE 2016/17 Exec. Rate

Budget 2017/18 AE 2017/18 Exec. Rate

Gross 1,876,470,979 1,867,809,014 100% 2,229,099,687 2,139,891,458 96% 2,702,721,862 2,602,419,751 96%

Appropriation in Aid (AIA) 250,000,000 181,997,380 73% 210,020,000 224,195,696 107% 271,035,600 271,613,781 100%

Net 1,626,470,979 1,685,811,634 104% 2,019,079,687 1,915,695,762 95% 2,431,686,262 2,330,805,970 96%

Compensation to employees 1,346,434,164 1,231,259,372 91% 1,439,638,804 1,547,250,285 107% 1,719,782,992 1,958,945,411 114%

Maintenance 15,600,000 0 0% 7,500,000 2,841,579 38% 1,305,400 1,010,100 77%

Operations 264,436,815 226,985,624 86% 571,940,883 388,885,760 68% 527,738,117 370,850,459 70%

The budget/expenditure per programme is neither more informative on MNH expenditure (see table below, from the same source: note that totals of both tables

do not match, which comes from errors in the paper documents provided by the Bungoma County Government).

Table 6. Bungoma County Financial Statements (KES) by Programme

Budget

2015/16 % in col.

AE 2015/16 % in col.

Budget 2016/17

% in col.

AE 2016/17 % in col.

Budget 2017/18

% in col.

AE 2017/18 % in col.

Programme 1: general administration and planning and support services

1,730,900,000 83% 1,668,809,014 99% 2,036,213,921 84% 1,874,189,641 88% 2,519,493,142 93% 2,233,560,781 97%

Programme 2: Preventive promotive and rehabilitative

5,100,000 0% 3,100,000 0% 140,908,712 6% 35,397,500 2% 11,408,800 0% 0 0%

Programme 3: Curative health services

325,600,000 16% 10,200,000 1% 216,345,027 9% 204,292,832 10% 153,913,120 6% 63,792,172 3%

Programme 4: Health Research

11,400,000 1% 0 0% 3,992,968 0% 3,201,884 0% 8,000,000 0% 2,153,860 0%

Programme 5: Maternal and child health

1,100,000 0% 1,100,000 0% 22,569,600 1% 22,569,600 1% 9,907,600 0% 0 0%

Total 2,074,100,000 100% 1,683,209,014 100% 2,420,030,228 100% 2,139,651,458 100% 2,702,722,662 100% 2,299,506,813 100%



The Kenya National Health Accounts (NHA) 2015/167 presents total health expenditure by group of “disease

conditions”8 (Figure 1 below):

Figure 1. Distribution of total health expenditure (THE) by Diseases / conditions, FY 2012/13 and FY 2015/16

According to the NHA 2015/16, the total Reproductive Health (RH) expenditure represents 12.1% of Total

Health Expenditure (THE). This ratio seems consistent with other studies already mentioned in the VfM

analysis of the Formative Evaluation report: the ratio is 14% in Ethiopia9 (and 11% for Child Health: THEch on

THE), and 16.4% in Afghanistan10. Country case studies done by Carlyn Mann et al show the following ratios:

• Reproductive and Maternal Health expenditure on THE: from 8% in Malawi to 20% in Tanzania

• Child Health Expenditure on THE: from 8% in Tanzania to 22% in Malawi.

• A Public Health Expenditure Review done in Tanzania11 shows that Reproductive and Child Health

represent 31% of Local Government Health Expenditure.

For cost-effectiveness analysis (CEA) purpose, we decided to keep the NHA 2015/16 rate of 12.1% (RH/MNH)

for each year of the study. On the one hand, it could be an overestimation: (1) MNH expenditure per capita

may be lower than the national average in the counties and sub-counties supported by the DFID MNH

programme 12; and (2) RH is broader than MNH. On the other hand, the DFID MNH budget exceeds the strict

limits of maternal and neonatal health: e.g. infrastructure, green technology, HSS/capacity building at

national and county level (e.g. training in leadership and governance, health financing, HMIS, HRH

management, etc.) go far beyond MNH. If we try to estimate the share of MNH expenditure on RH

expenditure, we should then as well estimate the share of the DFID MNH budget that was exclusively spent

7 Ministry of Health, 2017. Kenya National Health Accounts 2015/16. Nairobi : Ministry of Health. 8 HIV/AIDS, TB, Malaria, Reproductive Health (RH), NCDs, nutritional deficiencies, vaccine-preventable diseases, diarrhoeal diseases and respiratory infections. 9 Ethiopia Federal Ministry of Health. April 2014. Ethiopia’s Fifth National Health Accounts 2010/2011. Addis Ababa, Ethiopia. 10 Islamic Republic of Afghanistan, Ministry of Public Health. Afghanistan National Health Accounts with Sub-Accounts for Reproductive Health 2011-2012. October 2013. 11 Directorate of Policy and Planning, Ministry of Health and Social Welfare. July 2012. Health Sector Public Expenditure Review, 2010/11. Dar es Salaam, Tanzania and Health Systems 20/20 project, Abt Associates Inc. 12 It happens quite often that domestic budget does not increase when substantial external aid is present



for MNH: this would lead to somewhat artificial calculations (and quasi impossible at this stage). Therefore,

the attribution of incremental impact to the DFID programme will take into consideration the DFID MNH

budget compared to county MNH expenditure calculated as 12.1% of total county health expenditure

(excluding DFID contribution) table below).

Table 7. Bungoma County: health expenditure and MNH expenditure 2014 -2018 (GBP)

Expenditure 2014 2015 2016 2017 2018

Total Health (GBP) 12,394,705 12,313,604 15,817,314 18,306,700 19,376,322

MNH (GBP) 1,499,759 1,489,946 1,913,895 2,215,111 2,344,535

Bungoma population 1,609,749 1,627,374 1,719,013 1,774,180 1,637,392

Total health per capita (GBP) 7.70 7.57 9.20 10.32 11.83

MNH per capita (GBP) 0.93 0.92 1.11 1.25 1.43

3 DFID MNH BUDGET (OVERALL AND FOR BUNGOMA COUNTY)

3.1 MANI, MANI BUNGOMA, AND MANI CICF

OPTIONS provided us with the 2 following tables (see below): budget and actual expenditure per phase and

per category of cost, and cost per partner (as part of the total inputs and expenses of table 7).

Table 8. MANI project budget and expenditure 2015-2018 (GBP)

GBP Inception Implement-

ation Extension Total

Current

Amendment

Revised

Total

Programme

Spend up to

Dec 2018

Inputs

Long Term

Input Days 110,369 3,044,524 1,135,980 4,290,873 980,077 5,270,950 6,319,019

Short Term

Input Day 268,624 1,746,053 110,480 2,125,157 103,635 2,228,792 1,016,300

Management

Input Day 28,051 486,680 54,702 569,433 70,800 640,233 770,250

Total Cost of

Fees 407,044 5,277,257 1,301,162 6,985,463 1,154,512 8,139,975 8,105,569

Expenses

Total Travel

Costs 31,538 290,839 58,976 381,353 54,416 435,769 241,960

Total Living

Costs 24,542 278,190 23,183 325,915 20,020 345,935 119,944

Total

Equipment

Costs

66,400 346,003 261,277 673,680 6,462 680,142 429,465

Total Activity

Costs 139,155 1,653,469 1,792,624 810,123 2,602,747 2,713,661

Of which PBF 704,405

Total Other

Costs 42,030 592,522 634,552 118,583 753,135 872,835

Total Cost of

Expenses 303,665 3,161,023 343,436 3,808,124 1,009,604 4,817,728 4,377,865



GBP Inception Implement-

ation Extension Total

Current

Amendment

Revised

Total

Programme

Spend up to

Dec 2018

Total Fees &

Expenses

(output

based)

710,709 8,438,280 1,644,598 10,793,587 2,164,116 12,957,703 12,483,434

Total CICF

and vouchers

(reimbursable

)

576,364 11,408,794 1,250,000 13,235,158 -3,283,324 9,951,834 9,507,261

Of which

vouchers

(DSF)

884,703

Total Cost 1,287,073 19,847,074 2,894,598 24,028,745 -1,119,208 22,909,537 21,990,695

Table 9. MANI expenditure per Consortium Partner

GBP and % Fees Expenses Total % to CICF

% to

Bungoma

County

% for other

KPMG 2,087,108 134,110 2,221,218 100%

Population Council 402,191 87,555 489,746 80% 20%

IHPMR 159,127 71,946 231,073 100%

Mannion Daniels 806,790 121,449 928,239 100%

AMREF 357,566 432,531 790,097 100%

CARE 357,911 392,197 750,108 100%

Internews 563,209 563,209 40% 40% 20%

Additional data and assumptions provided by Options and or collected/elaborated by the evaluation team

were necessary to do an annual apportionment of expenditure between the following components: HSS

Bungoma, overall CICF, and CICF in Bungoma.

• Expenditure per year: Inception phase = Q1 and Q2 2015; implementation phase : from Q3 2015 to

Q4 2017 (thus, 10 Quarters); Extension phase (and current amendment): 2018.

• Apportionment between phases: 6% for inception phase, 76% for implementation phase, 18% for

extension phase. Flat apportionment per Quarter inside each phase.

• We have grouped LT and ST input days, management days, travel costs and living costs under a single

category of cost “Input”.

• The cost of KPMG represents 70% of the total CICF management cost (according to Options), the

remaining being equally shared between Options and Population Council (15% each). As the cost of

KPMG is GPB million 2.2 (and KPMG being 100% on CICF), we deduce that the total CICF

management cost represents GBP 3,412,249 (including also the CICF cost of Internews).

The apportionment is presented in the tables below (Bungoma HSS and CICF).



Table 10. MANI-HSS Bungoma expenditure per year (GBP and %)

GBP 2015 2016 2017 2018 Total %

Inputs (LT & ST days,

management days, travel

costs and living costs)

1,072,708 1,536,738 1,536,738 909,040 5,055,224 51%

Equipment 91,132 130,554 130,554 77,227 429,466 4%

Activity 426,360 610,794 610,794 361,308 2,009,256 20%

Performance base

financing (PBF) 149,473 214,132 214,132 126,667 704,405 7%

Demand side financing

(DSF) 187,732 268,941 268,941 159,089 884,703 9%

Other 185,214 265,333 265,333 156,955 872,835 9%

Total 2,112,619 3,026,492 3,026,492 1,790,286 9,955,889 100%

Table 11. MANI-CICF expenditure 2015-2018 (GBP and %) 2015 2016 2017 2018 Total

Management cost 740,458 1,173,814 1,173,814 324,164 3,412,249

KPMG 482,004 764,099 764,099 211,016 2,221,218

Options 104,784 166,108 166,108 45,873 482,873

Population Council 104,784 166,108 166,108 45,873 482,873

Internews 48,887 77,498 77,498 21,402 225,284

Expenditure on CICF Fund 1,871,095 2,966,160 2,966,160 819,143 8,622,558

Total 2,611,553 4,139,973 4,139,973 1,143,307 12,034,806

% Management cost 28% / 39.5%

3.2 CICF AND CICF BUNGOMA

The table below shows the list of CICF grants by round, category (innovation or scaling up), with the duration

in months and the total expenditure (up to December 2018).

Table 12. CICF Grants: Category, Duration and budget (GBP)

Round I/S Grant Duration

(months)

Total

(Contract

Amount)

Unexpended

funds (e.g.

grant close

or realised

savings)

Total

(Adjusted

Project

Amounts)

1 I Health E-net (CICF-INN-R1-GA-

001) 01-01-2016 / 30-06-2018 29 249,851 2,174 247,677

1 S IRC (CICF-SCL-R1-GA-002) 01-

12-2015 / 30-10-2017 22 1,000,000 26,838 973,162

1 I

Malteser International (CICF-

INN-R1-GA-003) 15-10-2015 /

30-04-2018

30 249,875 - 249,875



1 I MKU (CICF-INN-R1-GA-004) 01-

11-2015 / 30-04-2019 41 338,575 - 338,575

1 S

Pathfinder International (CICF-

SCL-R1-GA-005) 15-10-2015 /

31-12-2017

26 999,858 79,033 920,825

1 I RCTP-FACES (CICF-INN-R1-GA-

006) 15-10-2015 / 30-04-2018 30 249,884 30,964 218,920

1 I PATH (CICF-INN-R1-GA-007) 01-

01-2016 / 31-03-2019 38 665,000 609 664,391

2 I

Gertrude Children's' Hospital

(CICF-INN-R2-GA-001) 01-03-

2016 / 30-06-2018

27 242,496 142 242,354

2 S

Afya Research Africa (CICF-SCL-

R2-GA-002) 01-03-2016 / 31-

03-2019

36 938,642 - 938,642

2 I APHRC (CICF-INN-R2-GA-003)

15-02-2016 / 31-05-2018 27 246,507 25,333 221,174

2 I RCTP-FACES (CICF-INN-R2-GA-

004) 01-03-2016 / 31-03-2019 36 368,885 - 368,885

2 S

Save the Children International

(CICF-SCL-R2-GA-005) 01-04-

2016 / 30-04-2018

24 822,271 - 822,271

2 S MGH (CICF-SCL-R2-006) 01-06-

2016 / 31-03-2019 33 821,086 - 821,086

3 I

Center for Public Health and

Development (CICF-INN-R3-GA-

002) 01-07-2017 / 31-03-2019

20 425,608 - 425,608

3 I

Afya Research Africa (CICF-INN-

R3-GA-003) 01-07-2017 / 31-

03-2019

20 399,954 - 399,954

3 I

Lexlink Consulting (CICF-INN-R3-

GA-004) 01-07-2017 / 31-03-

2019

20 170,346 - 170,346

3 I

Management Sciences for

Health, Inc. (CICF-INN-R3-GA-

005) 01-08-2017 / 30-04-2019

20 421,903 - 421,903

3 I APHRC (CICF-INN-R3-GA-006)

01-08-2017 / 30-04-2019 20 355,626 - 355,626

Total 8,966,366.43 165,093.33 8,801,273.10



Table 13. CICF: Expenditure per type of grants

# Grants Total months Contract

Amount

Adjusted Project

Amounts %

Innovation 13 358 4,384,510 4,325,287 49%

Scaling 5 141 4,581,857 4,475,986 51%

Total 18 499 8,966,366 8,801,273 100%

The CICF grants for Bungoma are listed in Tables 14 and 15. As the MGH project (UBT) covers 4 counties, we

divided the actual expenditure by 4 to get the share of Bungoma (there was no other available information

to use another method of apportionment).

Table 14. CICF projects in Bungoma (budgets)

Rd. Grantee Project County Budget(£)

1 (I) Mount Kenya

University (MKU)

Collaborative new-born support project (new-

born care units and training/mentoring) Bungoma 338,575

2

(S/I)

Save the Children

(SCI)

1. Increasing access to quality care for preterm

and low birth weight babies through kangaroo

mother care

2. Improving the monitoring of labour through

the use of a digital partograph

Bungoma 822,271

2 (S)

Massachusetts

General Hospital

(MGH)

Every second matters for mothers and babies -

uterine balloon tamponade 4 counties

821,086 (total

budget for the 4

counties: so, a

theoretical

budget of

205,271 for

Bungoma)

Total

1,981,932 (or

1,366117 if MGH

limited to

Bungoma)

Table 15. CICF Bungoma: Actual Expenditure per year (GBP)

2016 2017 2018 Total

MGH 23,893 95,370 40,000 159,262

MKU 97,816 97,816 97,816 293,448

SCI 221,255 295,007 295,007 811,270

Total 342,964 488,193 432,823 1,263,980

Management Cost (OPTIONS, KPMG, POP.

COUNCIL, INTERNEWS) (28% of total or 39.5% of

direct expenditure)

135,471 192,836 170,965 499,272

Grand Total 478,435 681,030 603,788 1,763,252



We decided to consider all CICF projects in Bungoma (whether innovation or scaling projects) as contributing

to Bungoma HSS13 (quoted from the CICF evaluation report):

• “The Collaborative Newborn Support Project was an innovation project designed as an

implementation research study to address the high rate of neonatal mortality in Bungoma County

which was estimated at 32/1,000 in 2015. The project was implemented by Mount Kenya University

(MKU), a private university located in Thika, Kiambu County and the Fountain Africa Trust (FAT), a

local NGO based in Webuye, Bungoma County in collaboration with the Bungoma County Ministry of

Health and the Kenya Paediatric Association (KPA). The interventions included: (1) Upgrading and

refurbishing eight new-born care units; (2) Training of staff in the eight hospitals using a new-born

care training package developed by the KPA; including the provision of job-aids; (3) Radio spots, a

weekly radio information and call-in programme, and community theatre skits with information

about new-born care; (4) Establishment of a telehealth platform for NBU staff to consult with

paediatricians in Bungoma (there is one paediatrician in Bungoma) and at MKU; (5) An individualised

telephone follow-up service with advice and reminders on new-born care for all mothers who

delivered in the eight hospitals”. All these interventions can be considered as having supported HHS

in Bungoma County with potential impact on MNH outcomes.

• KMC: “In April 2016, Save the Children International (SCI) received a Round 2 CICF scaling grant to

accelerate the implementation of KMC in 18 health facilities in Bungoma county, to support the MoH

in finalising and disseminating national guidelines for KMC and to increase the use of KMC through

awareness raising in the mass media. The first two objectives were pursued in cooperation with the

MoH at county and national level; for the third component, SCI worked in collaboration with BBC

Media Action, an international organisation affiliated with the British Broadcasting Corporation. The

grant was shared with an innovation project implemented by SCI for the piloting of an electronic

partograph in selected hospitals of Bungoma County. The project ended after a cost extension in

December 2018. Funding for the support of KMC in Bungoma County and for the piloting of the e-

partograph had already ended earlier, but the extension was granted for SCI to provide technical

support to the MoH and to work with partners on the revision of national guidelines for quality

obstetric and perinatal care. The CICF KMC grant to SCI contributed to the improvement of new-born

care in Bungoma County, together with the CICF grant to MKU and the health systems support

provided by the MANI programme. The contribution to national policies and guidelines was also

confirmed in interviews, although all MoH documents that mention KMC available on the MoH

website predate the initiation of the project, and several other large organisations, including UNICEF,

were active in the promotion of KMC”. All these interventions can be considered as having supported

HHS in Bungoma County with potential impact on MNH outcomes (even the training on using

partographs).

• UBT: “In June 2016, the MGH was awarded a CICF scaling grant to introduce and promote the uptake

of the ESM-UBT kit in maternity care facilities initially in Turkana, Garissa, Kakamega and Bungoma

Counties, later scaled up to 12 counties. The grant ended after a no-cost and a cost extension in

March 2019 with a total grant budget of GBP 821,086. The grant-funded project was implemented by

the Kisumu Medical and Educational Trust (KMET), an organisation that specialised in promoting

reproductive health with a focus on adolescent girls and young women in partnership with the MoH

and medical training institutions. Early collaboration with the MoH resulted in a draft communiqué

issued by the MoH and UNFPA in 2014 entitled ‘Ending preventable maternal and mortality in Kenya:

13 See the CICF evaluation report under a separate cover.



Accelerating use of uterine balloon tamponade to improve management of post-partum

haemorrhage in the 47 counties in Kenya.’ The Kenya Obstetrical and Gynaecological Society

endorsed the use of UBT in 2016 and recommended that all counties ‘operationalise training,

procurement and provision of the technology in all health facilities where it is needed’. Interviewed

staff reported that according to data extracted from facility registers, 654 kits had been used at the

time of the evaluation. They estimated that the rate of under-reporting is high, because inventory

monitoring indicated that the number was likely to be above 1,000. Introduction of the kits in the

eight counties that are additional to the CICF portfolio was funded by other international partners.

The Kenya Medical Supplies Authority (KEMSA) has included ESM-UBT kits in its range of products

and several counties have included the procurement of the kits in their health budgets. The CICF-

funded project had a research arm that included the follow-up of women treated with UBT after

discharge from the maternity ward. The scope of the in-service training provided by KMET in

collaboration with MoH has also increased to comprehensively cover the management of PPH, and

negotiations with medical training colleges are on-going to include this as a pre-service training

module. In meetings with the evaluation team CHMT officials in Kakamega and Bungoma County

confirmed that the counties were committed to rolling out UBT to all health facilities providing

maternity services and have included the ESM-UBT kits in their procurement plan. Facilities visited by

the evaluation team in November 2018 and in April 2019 reported that kits were in stock.

Uncontrolled PPH is not a common event, so the reported number of times the kits had been used

were generally in the low one-digit range. The project to scale-up UBT in Kenya was highly successful,

providing timely input to bring an on-going effort by multiple national and international partners to

scale. The technology is appropriate and has the potential to contribute to a reduction of maternal

mortality in Kenya. The portfolio of interventions supported by CICF which included health worker

training in collaboration with the MoH, advocacy at central and county level, support to budget

planning at county level, and business planning for the production and commercialisation of the

ESM-UBT kits triangulated well with programmes of other international partners to generate an

outcome that has a high probability of being sustainable. No need to say that ESM-UBT should be

considered as having supported MNH-HSS In Bungoma with potential impact on MNH outcomes.

3.3 LSTM – MIH (OVERALL EXPENDITURE)

The figures below show the MiH Quarterly (actual) expenditure by category of cost since Q2 2014 until Q4

2018 in GBP and in percentage. It illustrates that the “Start-up and Office running costs” have increased

until Q4 2016 (almost proportionally to the level of in-service training activities) but did not decrease

accordingly when those activities were finished: it is thus difficult to consider those costs as pure fixed cost.

The “Start-up and Office running cost” + “Management fee” (MF) represent 42% of the total expenditure for

the whole duration of the programme, but varies from 15% (Q3 2014) to 70% (Q4 2017): that efficiency rate

(and the variations between Quarters) can be considered as normal taking stock of the nature of the

programme. However, the relatively high absolute (and relative) costs of the “start-up and office running

costs” in the last six quarters of the programme, while main training activities had mostly closed, may require

some further explanation by LSTM. One assumption is that human resource investment for pre-service

training and operational research is partly or fully captured under this category.

The expenditure per category of cost also shows that the MiH moved from a programme that was initially

more focused on training, to one that put more emphasis on generating evidence and learning from

implementation (hence 7% of the original budget was allocated to M&E, supervision and operational



research (OR)). Indeed, once training targets were achieved the focus shifted to generating evidence, with

the cost category “M&E and OR” becoming more substantial14. That could also explain the observation that

the “Start up and Office running cost” don’t really decrease after Q2 2017 (when the training activities were

almost finished).

Once training targets were reached LSTM submitted a request to deliver more training from efficiency

savings in April 2018, this was approved in October 2018 by which time LSTM had started reducing staff

numbers and eventually restructured the team to get a better balance between activities and LOE in

April/May 2019. Most M&E, support supervision and OR activities were completed towards the end of 2018.

There is nothing to contradict the main Value for Money (VfM) results from the LSTM Quarterly reports, e.g.

“Value for money (VfM) is being achieved by minimising input costs (economy) and maximising the efficiency

and effectiveness of the project, through the following measures:

• Quarterly spot-checks of cost items claimed by CMNH-LSTM Kenya (alongside continuing use of the

quarterly financial tracker which allows scrutiny of each transaction made under the programme);

• Programme management and financial internal control spot-check visits are made to the CMNH-

LSTM Kenya office on a regular basis;

• Continual review of UK-based staff and volunteer travel, meaning activities maximised in-country for

each flight made;

• Volunteer UK faculty continue to be used to deliver EmONC training courses in-country, giving their

time and expertise pro bono;

• The model of phasing out the use of UK faculty and increasing use of in-country faculty as the

programme progresses means that fewer costs are incurred for international travel; and

• Building the capacity of in-country teams has continued through training, coaching and mentoring,

resulting in even greater competence and reducing the need for UK-based staff to undertake

international travel to address issues in-country”.

14 The « M&E » budget line also includes the cost of supportive supervision that started in 2016.



Figure 2. MiH Expenditure per Quarter (GBP)

Figure 3. MiH Expenditure per Quarter (in percentage)



Error! Reference source not found.4 below shows the direct cost per trainee (in-service): around GBP 500

(USD 63615). The investment cost (equipment, set-up of the training systems, etc.) explains the high cost

during the two first Quarters (as well as the low number of trainees at the start of the programme). This

(direct) unit cost per trainee comprises mainly of Accommodation and Subsistence for trainers and trainees

as well as trainers fees. In terms of international comparisons, it compares favourably with e.g. the cost of

short courses (Supply Chain Management) provided by the Regional Centre of Excellence for Vaccines,

Immunization and Health Supply Chain Management (RCE-VIHSCM, University of Rwanda): full cost of

US$ 1,716 per trainee, of which US$1,000 for tuition fee (5 days course, 2 trainers and 2 facilitators).

Figure 4. Direct Unit Cost per (direct) trainee ( in -service training) (GBP)

3.4 LSTM-MIH IN BUNGOMA

289 staff16 from Bungoma county have been trained by LSTM (from 2014 to 2016), of which 255 ( 88% )

through EmONC in service-training modalities (direct training), 14 in MPDSR/QI, 6 in QA and Data

Management, and 14 through ToT. According to LSTM, the direct expenditure17 for Bungoma training

activities represented GBP 88,70318 (0.9% of the total expenditure 2014-2018, or 2.4 % of total expenditure

for in-service training). The (direct) unit cost per trainee (in-service training) ranges from GBP 296 to 358,

which also compares favourably with other regional training and research organisations (see previous

section).

15 www.xe.com (27 May 2019) 16 Figures vary between several files received, but some files were including trainees from other counties (trainings were organised for staff coming from different sub-counties (within and outside Bungoma county). 17 Cost of trainers and trainees, transportation, subsistence and accommodation, etc. 18 2014: GBP 40,438; 2015: GBP 15,410; 2016: GBP 32,854

3,108

1,470

430 492572

356504

350 379 370

773

466 410

1,433

513

0

500

1,000

1,500

2,000

2,500

3,000

3,500

Q2 2014 Q3 2014 Q4 2014 Q1 2015 Q2 2015 Q3 2015 Q4 2015 Q1 2016 Q2 2016 Q3 2016 Q4 2016 Q1 2017 Q2 2017 Q3 2017 Total

GB

P



The (uncomplete and not 100% reliable) breakdown by sub county (provided by LSTM) shows that the sub-

counties supported by MANI have been privileged compared to the other sub counties, e.g. 11 trainees

(EmONC in service -training) in Mount Elgon vs. 32 trainees in Sirisia. It may be due to the difference of total

staff between sub-counties, but it is also certainly since MANI asked LSTM to train additional staff (because

some already trained staff had left). The evaluation team estimated a loss of in-service EmONC training

between 25% and 50% over a period of 1-2 years, based on the Bungoma data.

Table 16. MiH expenditure in Bungoma County

MiH Bungoma (GBP) 2014 2015 2016 Total

In-service 39,501 15,005 30,116 84,622

TOT 937 937

MPDSR/QI training 405 185 590

M&E and OR 2,553 2,553

Total 40,439 15,410 32,854 88,703

Start-up and Office running cost &

Management fee as % on activity costs

(based on quarterly financial reports)

20% 33% 38% 29%

Start-up and Office running cost &

Management fee cost 8,088 5,085 12,485 25,658

Grand total 48,526 20,496 45,339 114,361

The table below summarises the DFID MNH support to Bungoma county from the different components:

MANI HSS Bungoma; MiH Bungoma; and CICF Bungoma.

Table 17. Total DFID MNH Bungoma (GBP)

GBP 2014 GBP 2015 GBP 2016 GBP 2017 GBP 2018 GBP Total

MANI HSS Bungoma 2,112,619 3,026,492 3,026,492 1,790,286 9,955,889

MiH Bungoma 48,526 20,496 45,339 114,361

CICF Bungoma 478,435 681,030 603,788 1,763,252

Total 48,526 2,133,115 3,550,266 3,707,522 2,394,074 11,833,502

Population (6 SC) 1,041,954 1,038,944 1,111,753 1,164,325 1,070,236

Per capita (GBP) 0.05 2.05 3.19 3.18 2.24



4 COST-EFFECTIVENESS ANALYSIS

The basic principle of the Incremental Cost-Effectiveness Ratio (ICER) method is to divide the DFID MNH

programme expenditure by the incremental number of DALYs averted. The number of DALYs averted is

computed based on the increase of coverage rates. The objective of the cost-effectiveness analysis is to

provide comparisons between ICER as calculated in the VfM report of the formative evaluation and the

present VfM report (summative evaluation)19: some improvement is expected since the effects of additional

funding in HSS on health outcomes are likely to be lagged.

4.1 ADDITIONAL DFID MNH EXPENDITURE VS. DOMESTIC MNH EXPENDITURE IN BUNGOMA

The table below summarises the substantial support from DFID to MNH in Bungoma county, especially from

2015 to 2018 (between 51% to 65% of total annual MNH expenditure). See Sections above for detailed

calculations and methods. It is to be noted that (a) the county budget is for all the SC, (b) the MANI HSS

budget is for 6 SC, (c) the CICF in Bungoma covered all SC, and (d) the MiH covered all the SC, but with a

special focus on “MANI SC”. It proved totally impossible to calculate the exact expenditure from all

components (MiH, CICF, and specially the SC budget) for the 6 SC supported by MANI HSS. We therefore

must accept a certain margin of error in the cost-effectiveness calculations.

Such assistance for one county in a rather limited area (MNH) can give hope for a substantial impact.

Table 18. Bungoma County Health Expenditure 2014 - 2018

GBP 2014 GBP 2015 GBP 2016 GBP 2017 GBP 2018

County Health Expenditure

Total Health 12,394,705 12,313,604 15,817,314 18,306,700 19,376,322

MNH (12.1%) 1,499,759 1,489,946 1,913,895 2,215,111 2,344,535

DFID MNH Expenditure

MANI HSS Bungoma (6SC) 2,112,619 3,026,492 3,026,492 1,790,286

MiH Bungoma 48,526 20,496 45,339

CICF Bungoma 478,435 681,030 603,788

Total DFIF MNH Bungoma 48,526 2,133,115 3,550,266 3,707,522 2,394,074

Total DFID + County MNH 1,548,285 3,623,061 5,464,161 5,922,633 4,738,609

Share of DFID on total 3% 59% 65% 63% 51%

4.2 BASELINE

1. MATERNAL MORTALITY RATIO

The Maternal Mortality Ratio (MMR) (maternal deaths per 100,000 live births) varies considerably between

sources/models:

• The KDHS2014 provides an MMR of 360, which is for the period 2007-2014

• The UN modelled estimate is 525 in 2014 (WDI)

19 Therefore, contrary to appraisal exercises, the purpose of the cost-effectiveness analysis is not to enable choices between concurrent options.



• The UNFPA MMR for 2010 is 495, with estimates per Region and County20 (sole source of MMR data

per County):

Nairobi: 212

Western Region: 319

Bungoma: 259

We decided to use a tentative national baseline MMR (year 2014) of 400 (an in-between the UN modelled

estimate of 525, UNFPA 2009/10 figures (495) and KDHS 2007-2014 figures (362), and to compute the MMR

per County based on the variances from UNFPA 2010 data but taking the Western Region ratio of 319 instead

of the Bungoma County ratio (259) that seems rather inconsistent with the other MNH indicators of that

County compared to other similar counties. We will thus use a MMR of 258 for Bungoma (baseline 2014).

Table 19. National and Bungoma MMR in 2014 (baseline)

UNFPA MMR (Based on national

average of 495/100,000 in 2009)

MMR computed on national

average of 400 in 2014 (baseline)

National 495 400

Bungoma 319 258

2. NEONATAL MORTALITY RATE

NMR data are more consistent among data sources/models:

• The World Bank WDI NMR21 data:

Year 2010 2011 2012 2013 2014 2015

NMR 25.9 25.1 24.3 23.6 22.9 22.2

• KDHS2014 comprises of NMR22 (national level and per region) per Region:

National: 22

Western Region: 19

We decided to use the KDHS2014 NMR rate of the the Western Region and to apply it for Bungoma for our

baseline 2014: 19/1000

4.3 COST-EFFECTIVENESS RATIOS

This section presents the cost-effectiveness analysis for the DFID MNH programme in Bungoma county. It is

based on specific county BoD, variation of coverage trends (deliveries in HFs), and incremental health

expenditure (i.e. total DFID MNH support to Bungoma County (MANI HSS, MANI CICF , and LSTM/MiH).

The most common approach for cost-effectiveness, promoted by the World Health Organization’s Choosing

Interventions that are Cost–Effective (WHO-CHOICE) project, involves the use of thresholds based on per

capita gross domestic product (GDP) : an intervention that, per disability-adjusted life-year (DALY) avoided,

20 The Policy Brief No. 38 (October 2013 -Population Studies and Research Institute, National Council for Population and Development, UNFPA) 21 NMR: number of deaths of infants 0-27 days per 1,000 live births. 22 KDHS2014 NMR are for the period 2009-2014



costs less than three times the national annual GDP per capita is considered cost–effective, whereas one

that costs less than once the national annual GDP per capita is considered highly cost–effective.

The table below shows the deaths and DALYs related to maternal and neonatal disorders in Kenya from 2014

to 2018 (data 2014 to 2017 are from IHME accessed in August 2019, and 2018 data are calculated on the

variation 2016-2017). The first part of the table shows absolute figures (total deaths and DALYS for the whole

country) while the second part of the table shows the figures per 100,000 population (which thus considers

the demographic increase). We will use these figures to translate the number of deaths averted into DALYs

averted.

Table 20. MNH Burden of Disease 2014-2018 (IHME accessed in August 2019 ) Maternal disorders (absolute numbers)

Deaths DALYs Quotient DALY/Death Annual change

(Deaths)

Annual change

(DALYs)

2014 3,966.43 243,297.14 61.34

2015 3,962.46 243,062.54 61.34 -0.10% -0.10%

2016 3,993.02 244,948.66 61.34 0.77% 0.78%

2017 3,990.42 245,141.62 61.43 -0.07% 0.08%

2018 3,987.82 245,334.73 61.52 -0.07% 0.08%

Neonatal disorders (absolute numbers)


(Deaths)

Annual change

(DALYs)

2014 17,579.37 1,764,853.55 100.39

2015 17,505.17 1,771,833.52 101.22 -0.42% 0.40%

2016 17,256.60 1,764,455.73 102.25 -1.42% -0.42%

2017 16,978.27 1,755,556.10 103.40 -1.61% -0.50%

2018 16,704.43 1,746,701.36 104.57 -1.61% -0.50%

Maternal disorders (figures per 100,000 pop)


(Deaths)

Annual change

(DALYs)

2014 8.81 540.44 61.34

2015 8.59 527.09 61.36 -2.50% -2.47%

2016 8.46 518.81 61.33 -1.51% -1.57%

2017 8.26 507.26 61.41 -2.36% -2.23%

2018 8.06 495.97 61.50 -2.36% -2.23%

Neonatal disorders (figures per 100,000 pop)


(Deaths)

Annual change

(DALYs)

2014 39.05 3,920.33 100.39

2015 37.96 3,842.32 101.22 -2.79% -1.99%

2016 36.55 3,737.18 102.25 -3.71% -2.74%

2017 35.13 3,632.38 103.40 -3.89% -2.80%

2018 33.77 3,530.52 104.56 -3.89% -2.80%



The table below23 shows the cost per DALY averted: GBP 328 for the whole period of intervention (2015-

2018), ranging from GBP 11424 in 2018 to GBP 741 in 2017 (all “highly cost-effective” if compared to the GDP

per capita – GBP 1,286 in 2018-25). The year 2014 is used as baseline to reflect the fact that the substantial

support to Bungoma county started in 2015: thus all 2014 initial values and rates/ratios are kept the same

in the following years. The coverage increase in HF deliveries is then translated into maternal and neonatal

deaths averted, and then into DALYs averted.

Table 21. CEA Bungoma County

BUNGOMA 2014

(baseline) 2015 2016 2017 2018

Total 2015/18

Population 1,041,954 1,038,944 1,111,753 1,164,325 1,070,236

Expected deliveries 41,195 38,597 41,135 44,823 41,740

BoD Baseline (what if expected deaths and DALYs remained at the 2014 level?) in the 6 MANI sub-counties

MMR (258/100,000)

258 258 258 258 258

NMR (19/1000) 19.0 19.0 19.0 19.0 19.0

Expected maternal deaths

106 100 106 116 108 429

Quotient maternal DALY/Death

61.34 61.34 61.34 61.34 61.34

Expected maternal DALYs

6,519 6,108 6,510 7,094 6,606 26,317

Expected neonatal deaths

783 733 782 852 793 3,160

Quotient neonatal DALY/Death

100.39 100.39 100.39 100.39 100.39

Expected neonatal DALYs

78,576 73,620 78,461 85,496 79,615 317,193

Actual data Bungoma County : Coverage increase (6 MANI SC)

Delivery in health facility coverage rate

55.4% 59.6% 63.0% 60.8% 79.7%

Coverage increase (on 2014 baseline)

4.3% 7.6% 5.4% 24.4%

Impact of coverage increase (100% efficacy) (6 MANI SC)

Maternal deaths averted

4.25 8.12 6.25 26.23 45

Neonatal deaths averted

31.3 59.8 46.0 193.1 330

Total deaths averted

35.5 67.9 52.2 219.4 375

Maternal DALYS averted

260 498 383 1,609 2,750

Neonatal DALYs averted

3,139 6,001 4,618 19,389 33,147

Total DALYs averted

3,399 6,499 5,001 20,998 35,897

Cost-effectiveness (GBP) (100% attributability

DFID MNH expenditure Budget (GBP) (MiH and CICF in

2,133,115 3,550,266 3,707,522 2,394,074 11,784,977

23 Population, expected deliveries, and coverage rates are from Kenya DHIS2 (accessed during July-August 2019). The DHIS 2 figures are for the 6 sub-counties supported by MANI in Bungoma. Population figures don’t seem very consistent (100,000 inhabitants less in 2018 than in 2017 !), but it is the most robust source of data that we could find . 24 The DFID MNH business case was based on a CER of GBP 100 per DALY averted. 25 But based on efficacy and attributability rates of 100% (see below)



BUNGOMA 2014

(baseline) 2015 2016 2017 2018

Total 2015/18

to DFID additional funding)

Bungoma County and MANI in the 6 SC)

Total MNH expenditure (DFID + County)(GBP)

3,623,061 5,464,161 5,922,633 4,738,609 19,748,464

DFID MNH expenditure/Total MNH expenditure

59% 65% 63% 51% 60%

Cost per DALY averted (GBP)

628 546 741 114 328

Cost per death averted

60,065 52,290 70,965 10,914 31,425

The tables below show the sensitivity analyses both to the efficacy rate (from 25% to 100%) and to the

attributability rate (from 25% to 100%), with a RAG rating system based on WHO thresholds and on the

Kenya GDP per capita 2018 being US$ 1,710 (current US$26 = GBP 1,286 (2018): green if the cost per DALY

averted is less than the GDP per capita (highly cost-effective), amber if the cost per DALY averted is less than

3 times the GDP per capita (cost-effective), red if the cost per DALY averted is greater than 3 times the GDP

per capita (not cost-effective).

• The efficacy rate reflects the direct impact of HF deliveries compared to home deliveries: an efficacy

rate of 100% would mean that facility-based deliveries imply 100% less morbidity/mortality than

home deliveries, or there is no maternal/neonatal death/DALY anymore among additional deliveries

in health facility ( which is very unlikely ). An efficacy rate of 0% would mean that the increase of HF

deliveries has no impact at all on maternal deaths/DALYs.

• The attributability rate reflects the percentage of the outcomes that can be attributed to the

additional funding coming from the DFID MNH programme. An attributability rate of 100% would

mean that each additional death/DALY averted is due to the incremental funding coming from the

programme (and not to any other intervention or domestic funding).

The simulation shows that CE ratios remain “highly effective” or “effective” except in the case where both

rates are lower or equal to 25% (i.e. the additional DFID budget does not contribute to more than 25% of

the increased coverage and outcomes (the rest being attributed to other causes), and the efficacy rate of

HF deliveries is lower than 25% meaning that the benefits of HF deliveries compared to home deliveries are

much lower than expected). It is reasonable to assume that both rates are in the region of 50% which would

place the value for money assessment of the MNH programme in Bugoma County in the region between

cost effectiveness and highly cost effectiveness.

Table 22. CEA Bungoma County (GBP , 2015-2018)

Efficacy Attributability

25% 50% 75% 100%

25% 5,253 2,626 1,751 1,313

50% 2,626 1,313 875 657

75% 1,751 875 584 438

100% 1,313 657 438 328

26 WB data, accessed July 2019



The Bungoma CE ratios are much better than the ones calculated for the formative evaluation VfM report

where there were no “highly cost-effective” ratios whatever the efficacy and attributability rates were (see

Table below from the Formative Evaluation report):

Table 23. CEA Bungoma County (GBP, 2014-2016, Formative Evaluation) Cost per DALY averted

(US$) Attributability

Efficacy 18% 36% 54%

25% 17,570 8,785 5,857

50% 8,785 4,392 2,928

75% 5,857 2,928 1,952

100% 4,392 2,196 1,464

This improvement may reflect (1) the importance attached to VfM (for all the “Es”) from the different

implementers and (2) the fact that the effects of some investments in HSS on health outcomes are likely

lagged (e.g. infrastructure, green energy, PBF which may produce their full impact several months or years

after the start of their implementation).

It must be noted however that WHO thresholds as commonly used to rate the cost-effectiveness ratios are

rather “generous” in the sense that most of the MNCH projects/interventions (well implemented) are likely

to be rated as at least cost-effective (if not “highly cost-effective”). Some recent studies question those WHO

thresholds (e.g. one study published in 201527 suggests “ that cost-effectiveness thresholds representing

likely health opportunity costs tend to be below the lower bound suggested by WHO of 1x GDP per capita.

Hence, many previous and existing recommendations about which interventions are cost-effective that are

based on the WHO threshold are likely to do more harm than good… This suggests that current interventions

acceptable at a 1x GDP per capita threshold (or even below it) may be displacing more health than they

generate”.

Anyway, the Bungoma CE ratios 2015-18 now compare favourably28 with most of the 3MDGF 2015 CE ratios

in the different Regions/Townships in Myanmar, especially if we consider that the 3MDGF expenditure

includes all INGOs costs but does not include the overall management of the programme (UNOPS cost),

while the analysis of the DFID MNH in Bungoma includes all costs/expenditures. The 3MDG, a multi-donor

fund (8 DPs including DFID) managed by UNOPS (fund holder), comprised of 3 components: (1) MNCH, (2)

HSS, and (3) support to HIV/AIDS, Malaria and TB vertical programmes. The MNCH component budget

amounted to US$ 200 million for 5 years, covering around 4 million people in most vulnerable/remote/hard

to reach Townships (=districts). NGOs received grants (on a competitive process) from the fund holder

(UNOPS) to implement the programme in their respective Townships. The 3MDG conditional cash transfer

(demand side financing scheme) sub-component represented more than 25% of the overall MNCH

expenditure. The BoD in the targeted Townships is comparable to the BoD of Bungoma.

The table below shows the CE ratios per Region (several Townships are covered in each Region). The

attribution to the impact (DALYs avoided) of the 3MDGF has been set to 75%. The 201529 CE ratio (cost per

27 Jessica Ochalek, James Lomas, Karl Claxton. Cost Per DALY Averted Thresholds for Low- and Middle-Income Countries : Evidence from Cross-Country Data. University of York. CHE Research Paper 122. December 2015. 28 Contrary to the results of the VFM analysis in the Formative Evaluation Report 29 2015=third year of implementation



DALY averted) ranged from US$ 257 (Magway Region) to US$ 1,832 (Chin Region - specific terrain

(mountainous area and very low population density).

Table 24. CEA 3MDGF (US$, 2015, Myanmar)

Region

Additional

DALYs averted

(2015)

Contribution of

3MDGF to

additional

DALYs averted

(75%)

3MDGF Actual

Expenditure

2015 (US$)

Cost per DALY

averted US$

(100% efficacy)

Cost per DALY

averted US$

(50% efficacy)

Chin 8,337 6,253 5,727,029 916 1,832

Magway 29,109 21,832 2,807,040 129 257

Kayah 12,156 9,117 2,141,302 235 470

Ayeyarwady 16,149 12,112 7,183,514 593 1,186

Total 65,751 49,314 17,858,886 362 724



5 ADDITIONAL NOTES ON EFFICIENCY AND SUSTAINABILITY

5.1 MIH / LSTM

Efficiency issues have already been addressed in the Section 3.3. above:

• The LSTM “Start-up and Office running cost” increased until Q4 2016 (almost proportionally to the

level of in-service training activities) but did not decrease accordingly when those activities were

finished: they then move from a nature of variable cost to a nature of fixed cost.

• The “Start-up and Office running cost” + “Management fee” (MF) represent 42% of the total

expenditure for the whole duration of the programme, but varies from 15% (Q3 2014) to 70% (Q4

2017): that efficiency rate (and the variations between Quarters) can be considered as normal taking

stock of the nature of the programme. However, the relatively high absolute (and relative) costs of

the “start-up and office running costs” in the last six quarters of the programme, while main training

activities had mostly closed, may require some further explanation by LSTM. One assumption is that

human resource investment for pre-service training and operational research is partly or fully

captured under this category.

• Once training targets were reached LSTM submitted a request to deliver more training from

efficiency savings in April 2018; this was approved in October 2018 by which time LSTM had started

reducing staff numbers and eventually restructured the team to get a better balance between

activities and LOE in April/May 2019. Most M&E, support supervision and OR activities were

completed towards the end of 2018.

Our main conclusion was that there is nothing to contradict the main Value for Money (VfM) results from

the LSTM Quarterly reports, e.g. “Value for money (VfM) is being achieved by minimising input costs

(economy) and maximizing the efficiency and effectiveness of the project, through a series of measures”.

However, it seems that Kenya, whether at national or county level, is not yet able (financially and

institutionally) to take over EmONC in-service and pre-service training in their own institutions with their

own staff, organisations and domestic funding. The building of the capacity of in-country teams supported

by LSTM through training, coaching and mentoring, resulting in even greater competence and reducing the

need for UK-based staff to undertake international travel to address issues in-country was therefore not

enough. After almost 10 years of implementation (in Kenya and in other countries), the MiH programme was

not able to produce a sustainable training system in Kenya that could be implemented locally and financed

with domestic resources.

5.2 MANI HSS BUNGOMA

The Section 3.1. above showed that the “management cost”30 of MANI HSS Bungoma represented 51% of

total expenditure. We could find exactly the same ratios during the 3MDGF in Myanmar (from 2013 to 2017)

where the INGOs Programme Management Budget (PMB) (for HSS at Township level) represented exactly

half of the total expenditure (PMB + activities). Thus, there is no specific efficiency issue for the MANI HSS

Bungoma component of the DFID MNH programme.

The strong evidence that the MANI project in Bungoma County was effective, cost-effective and generated

value for money requires, however, a more contextualised analysis. The evaluation confirmed that a large

30 Inputs (LT & ST days, management days, travel costs and living costs)



investment in a project that comprehensively supported health systems with large financial investments and

intensive on-site technical and policy assistance can effectively and verifiably improve maternal and

neonatal health. But women in Bungoma County die in childbirth every week, they did so before the start of

the MANI project and they continue to do so after the project closed. An analysis of the effectiveness and

cost-effectiveness of the DFID investment in MNH in Bungoma County can therefore not be restricted to the

short period from 2015 to 2018. It also has to consider the sustained changes that were initiated with MANI

support.

The sustainability analysis that is presented in the main evaluation report confirms that many of the systems

and processes introduced in the county with MANI support continue and are likely to continue having an

effect on improved maternal and neonatal health. But many did not, because of the reduced fiscal space for

MNH programming in Bungoma County after the end of the project31, and in some cases because they were

not sufficiently appropriated to continue without the technical support by the MANI project team.

New financing sources for MNH opened for the county after the end of the MANI project, for instance

through Linda Mama and through the World-Bank managed THS-UHC programme. But it was already evident

at the time of the evaluation mission that effective low-cost activities such as reimbursing boda-boda

transport to health facilities for delivery, telephone follow-up for missed ANC appointments and

performance-based stipends for CHCs and birth companions only continued in some health facilities,

presumably on initiative of engaged health staff who, as is common practice, will soon rotate to other

assignments in other facilities. Key activities like the monthly or quarterly meetings of health facility

managers to review and discuss MNH outcome data, which contributed greatly to catapulting Bungoma

County to the head of the national MPDSR table already ceased. Performance-based financing, which

contributed to a major extent to improved productivity and quality of MNH care, was discontinued and may

or may not be revived with revenue generated by facilities from the Linda Mama scheme.

The important lesson learnt by the MANI project is that comprehensive HSS delivered with on-site technical

support is effective and cost effective. There may be a more painful lesson in store, namely that three years

of this type of support is not sufficient to generate sustained gains in maternal and neonatal health. It is too

early to come to such a conclusion. For now we have to rely on hope rather than evidence that it will not be

reached.

5.3 CICF

With a management cost of less than 40 percent which, according to an estimate provided by Options and

KPMG, was used to about 50 percent for technical assistance and capacity building, management of the CICF

programme can be considered as highly cost-efficient. This assessment still holds when the management

costs incurred by the grantees are added. They were generally budgeted in the range of 15 percent of the

grant value plus the cost of project personnel which, depending on the type of grant, can be allocated in

differing proportions to management or activity costs.

Cost-efficiency in management is, however, not the same a cost-effectiveness. In order to assess the cost-

effectiveness of the CICF programme, a single metric would be required for aggregating the results of all

grants, from a digitalised supervision tool in Nairobi to motorcycle ambulances in Turkana. It would also

have to include the lessons learnt from piloting innovative solutions that were found to be not feasible or

practicable. Such a single metric does, of course, not exist.

31 Between 2015 and 2018, the DFID MNH programme covered between 51% and 65% of MNH expenditures in Bungoma County (See Table 18)



As documented in the CICF programme evaluation report, the grants awarded by the programme generated

many solutions for improved MNH, some of which such as KMC, UBT or human milk banking are well on

their way of being fully integrated in national MNH policies and programmes. Others, such as improved

newborn care in Bungoma County, public/private/community partnerships for primary care in remote areas

of Homa Bay or solar-powered CEmONC facilities in Turkana have filled important gaps in MNH service

coverage in these counties.

The question about cost-effectiveness can therefore only be answered by considering the counter-factual,

i.e. the costs incurred by the support of these contributions to MNH in Kenya through a funding mechanism

that differed from a challenge fund.

Challenge funds, according to a definition shared by a number of international agencies including DFID are

‘competitive mechanisms to allocate financial support to innovative projects, to improve market outcomes

with social returns that are higher/more assured than private benefits, but with the potential for commercial

viability’32. This definition applies only to a small proportion of grants issued under the CICF. Market

outcomes with social returns could be, for instance, generated by KMET, the organisation in Kisumu that

markets condom-based uterine balloon devices to the public and private sector, by Lexlink, a consulting

company in Nairobi that is developing a digitalised supervision tool for MNH services which has raised the

interest of county health departments, and Afya Research Africa, a private sector health care provider

operating a network of rural health facilities in collaboration with communities and the public sector.

Supporting the expansion of these initiatives can be clearly attributed to the challenge funding approach.

For most of the other grant-funded projects, there are clear counterfactual options. Supporting international

NGOs to roll-out KMC in Kenya, to establish a human milk bank in a high-volume maternity hospital in Nairobi

or to implement a number of supply- and demand-side interventions to improve maternity services in

Turkana does not meet any criteria of challenge funding. Several of these grant-funded interventions were

already integrated in the UNICEF-component of the initial MNH programme or could have readily been

integrated in the MANI project. In hindsight, it is impossible to determine whether this approach would have

been as effective or as cost-effective as programming under CICF.

In the final analysis, the cost-effectiveness of a challenge fund depends on the emergence of at least one

brilliant idea that takes the county, region or world in storm. But brilliant ideas are rare and something like

the Grameen Bank or the M-Pesa system can only be invented once. The majority of grants issued by the

CICF did not have the necessary profile to generate such brilliant ideas which may have been related to a

reluctance in risk-taking buy the Grant Selection Committee. The final assessment of the cost effectiveness

of the CICF programme is therefore mixed: An efficiently managed portfolio of grants that generated a

number of positive results and many lessons learnt, but rather moderate innovation outcomes.

32 Ompa C (2013). Understanding challenge funds. ODI Report



ANNEX 1. LINDA MAMA ANALYSIS BY OPTIONS

From Options, May 201833:

“In April 2017, the Ministry of Health’s Free Maternity Service (FMS) was transferred to the NHIF under the

brand name ‘Linda Mama: Boresha Jamii’. Phase 1 (from April 2017) commenced with the faith-based and

private sectors; from July 2017, under Phase 2, the public sector was added. From March 2018, Linda Mama

introduced Phase 3 to include antenatal care (ANC) and post-natal care (PNC). For all Linda Mama services,

benefits are ‘portable’ (e.g. mothers do not need to receive services from the same provider/site). Currently,

it is estimated that 502 ‘low cost’ private and faith-based facilities are contracted while approximately 4,000

public sector facilities nationally are reported to be accredited as part of the scheme. Reimbursement rates

under Linda Mama are tiered to level of care and by sector. Table 1 presents the tariffs for ANC, delivery

(normal and caesarean) and PNC. In comparison, normal delivery under the NHIF national scheme is

reimbursed at KES 10,000 (US$97). Public-sector tariffs are lower given that these are additional to line-item

budgets such as salaries and consumables.

As of March 2018, approximately 40% of the health facilities in Bungoma county were accredited with Linda

Mama. The NHIF has a target of accrediting all public health facilities (currently 120 are accredited with 47

pending). For the private sector there is no target, with only 12 private facilities accredited to date (with one

in process). Accreditation is simple for public health facilities and entails a written request by the County

Director of Health to the NHIF branch office, listing the public health facilities to be included. For private

providers, they must be registered with the NHIF and comply with various norms and standards.[3]

Performance-to-date under Linda Mama in Bungoma county has significant potential for growth. It was

reported in the April 7th, 2018 breakfast meeting that 9,776 women have registered (as of the end of

March 2018), with 4,906 women delivering normally and 639 delivering through cesarean section.

Approximately Ksh 12 million had been paid with Ksh 6.8 million pending. A county manager summed

up Linda Mama growth in the breakfast meeting, “by the look of things [the public sector] have woken

up.”

There is some market segmentation with private providers having more mixed clientele than their public

counterparts. The private providers estimated that 75–90% of their patients were on the NHIF SupaCover

with the remainder covered through Linda Mama and private insurance, with very few fee-paying. One

private provider noted that some women on Linda Mama can afford to be on the NHIF SupaCover and should

33 https://medium.com/@options_int/implementing-linda-mama-in-bungoma-county-7b27136d30a9



be directed to do so, “[we need to] help those people that can pay to realise that they should be paying.” In

the public sector the number of NHIF clients was reported to be much less but growing.

In Bungoma county, all facilities — both public and private — can receive reimbursements directly to

their bank account. This is a significant enabler for the programme in the public sector as facilities can

benefit directly from the ‘fruit’ of Linda Mama. Bungoma county is one of the few counties in Kenya to

allow for this as, in many other counties, reimbursements from the NHIF and Linda Mama are pooled in

the county revenue account (CRF) and may get delayed or diverted.

Linda Mama has benefitted from the MANI project in Bungoma County. This has included facility-based

performance-based financing (PBF) and transportation vouchers for pregnant women to reach health

facilities. Both interventions have laid a solid foundation for Linda Mama as facilities have forged stronger

community linkages and health workers have developed skills and experience in autonomous decision making

on PBF, channeling reimbursements towards initiatives that promote quality and access to MNH services.

While the MANI project plans to exit Bungoma County at the end of the year, supporting community

awareness and facility readiness for uptake of Linda Mama is a key part of the programme’s exit strategy.

As such, the project team sees Linda Mama as “…the beginning of a journey toward UHC.”

Bungoma County is well poised to deliver on the promise of Linda Mama. Linda Mama represents a

tremendous opportunity for expectant mothers, who previously may not have been able to deliver in a

facility (one such mother was met as part of the case study). As expectant mothers can designate a

facility, healthy competition between providers also represents potential for driving up clinical quality

and client-centred care within facilities and across providers. As public providers have some autonomy

and authority over how reimbursements are spent, they can benefit from the Linda Mama ‘fruit’. As a

flagship programme for the county, the DoH leadership is also poised to deliver on the promise of Linda

Mama. The passionate support of the County First Lady has also provided foundation for success. This

is bearing results as Bungoma County is reported to be leading on Linda Mama in the country.

While a strong foundation has been laid, there is still much that can be done to deliver on the Linda Mama

promise. These are not unique to Bungoma and are of relevance to other counties, the Ministry of Health and

the NHIF.

Develop strategies to encourage mothers to transit from Linda Mama to NHIF as part of plans for UHC. Like

the Linda Mama model facility, this may include subsidy for indigent households while encouraging those

who can afford to pay, to get on SupaCover.

Exploit opportunities to integrate sensitization and procedural orientation on Linda Mama into existing and

planned health worker and community engagement. Linda Mama should not be viewed as separate, but

integral to other reproductive, maternal and neo-natal initiatives. Consider additional guidance on referrals,

treatment of other delivery complications and mother or infant illness which is not currently catered for in

the reimbursement schedule.

Provide procedural clarity to health workers and mothers on ANC and PNC packages, so that these are

delivered consistently and completely. Reinforce the inclusion of post-partum family planning (PPFP) within

the PNC package. Consider a separate reimbursement, or a top up to one of the PNC visits, to cater for this

service. The effective inclusion of modern PPFP in Linda Mama would allow Kenya a viable strategy for

improving its uptake, which is estimated at only 16% at six months, despite an estimated 64% of women

delivering in health facilities.[5] This suggests a missed opportunity for post-partum mothers.



Remove or reduce bottlenecks for the utilization of Linda Mama reimbursements in the public sector. While

funds are received directly into public facility back accounts, all AIE must be approved by the Chief Officer.

Consider allowing hospital chief executives and sub-county medical officers AIE as a means of reducing

reliance on one approval point.

·Strengthen leadership and management of Linda Mama by the County DoH. Consider regular Linda Mama

management review meetings that include the NHIF, county health managers and the private sector so that

problems are identified, and remedial actions implemented. Build county level leadership that seeks to more

firmly define the role of Linda Mama and the transition of women from the programme on to SupeCover as

part of the county’s UHC plans.”

Documents

MNH PROGRAMME SUMMATIVE EVALUATION