Analyzing funding for HIV and TB in South Africa to assess allocative efficiency

(2011/12-2013/14)

Funded by The Global Fund

On behalf of the Investment Case Steering Committee (NDOH, SANAC

and UNAIDS)

16 July 2016

Teresa Guthrie, Arjun Vasan, Kavya Ghai & Tess Ryckman

Outline of this presentation

Purpose and methodology

Key Findings:

National analysis

Provincial breakdown

District breakdown

Key Messages

2 | R4D.org

Purpose of this analysis

The national, provincial, and district-level analysis of spending on

HIV and TB in South Africa was intended to supplement the cost

estimates undertaken for the Investment Case, and attempted to

answer the following questions:

1. How much money was spent on HIV and TB by the South African

Government (SAG), PEPFAR, and the Global Fund in the years

2011/12 to 2013/14 – are there issues affecting sustainability?

2. How was this money spent across interventions and geographies,

and by funding source – were these the ‘optimal’ mix/ balance?

3. What financial and epidemiological data challenges limit our

understanding – what needs to be improved to ensure that

geographical targeting (hotspots) has the impact we hope for?

3 | R4D.org

Basic Methodology

4 | R4D.org

• Collected HIV and TB

spending from PEPFAR

EA, Global Fund EFRs,

and SAG BAS & NT

reports

• SAG = DOH, DOE,

DSD, DCS, SAPS, DOD

and other ES (if labelled

HIV (BAS records & NT

records)

• TB out-patient estimated

spending was calculated

based on provincial

patient # and unit costs

• 3 datasets consolidated

into one (common

structure based on BAS –

country suitability)

• 6 sets of spending

categories were

“crosswalked” or

reconciled

• Analysis of Excel

consolidated database

• Matrices and graphs

generated

• Draft figures were

approved by NDOH, NT,

PEPFAR and GF

Collect Inputs Crosswalk Analysis

1 2 3

Six sets of spending categories were “crosswalked” (or reconciled) to the public BAS ‘common codes’

6 | R4D.org

Spending Category Example

South Africa IC Categories “Treatment & Care”

NASA Categories

“ASC.02.01.03.98

Antiretroviral therapy not

disaggregated neither by age

nor by line of treatment”

SHA Categories“HC.1.3.1 General outpatient

curative care”

Global Fund SDA

Categories *

“Care and support for the

chronically ill”

PEPFAR EA Program Areas

**“FBTCS”

BAS – 300 different names

for 35 core programs -

consolidated

“ART”

BAS

Common

Codes

* Global Fund EFR data were not split by district. Estimates were used based on

indicative split

** PEPFAR data not split by district

1

2

3

4

5

6

Outline

Purpose and methodology

Key Findings:

National analysis

Provincial breakdown

District breakdown

Key Messages

9 | R4D.org

National Analysis – SAG, GF, PEPFAR

10 | R4D.org

The SAG contributes the majority of funding, which mainly goes to HIV

11 | R4D.org

5

20

10

15

0

25

2013

17.8

22.1

3.7

3.9

0.2

0.719.217.4

13.3

0.4

2011

14.9

3.9

2012

GF

PEPFAR

SAG

Total HIV and TB spending by source

ZAR Billions

25

20

15

10

5

0

15.7

2011

17.4

3.30.2

13.8

0.23.4

3.7

2013

22.1

0.4

18.0

2012

19.2

Total HIV and TB spending by focus

ZAR Billions

TB

HIV/TB

HIV

US$ 2.3 billion in 2013/14SAG = 80% (16% inc ann.av)PEFPAR = 17% (5% dec)GF = 3% (77% increase)

NB. GF TB expenditure could not be separated within the SDAs. Excludes USAID additional contributions to TB (not reported in the EA data): USG FY11: $13,972,000, USG FY12: $12,000,000, USG FY13: $12,008,901.

Spending by thematic split for HIV has changed over time, but what should be the optimal mix?

18

16

10

6

4

2

0

8

14

12

7

2008 (NASA)

1

2012 (IC)

2010 (est)

2009 (NASA)

2007 (NASA)

2011 (IC)

2013 (IC)

4

12

Billions ZAR

Enabling Env

Social ProtectionPrevention OVC Support

Treatment & Care Prog, Mgmt & Sys Strgthening

Research

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

2010 (est)

2012 (IC)

2011 (IC)

72.8%

2009 (NASA)

2008 (NASA)

13.0%

2007 (NASA)

2013 (IC)

65.8%

20.0%

Historical Thematic Spending Historical Thematic Spending (%)

Comparing 2013/14 spending with 2014/15 Resource Needs Estimates could help answer adequacy and allocative efficiency questions

10

24

6

4

26

2

20

16

8

0

22

14

18

12

Non SA IC plus not disagg.TB

Billions ZAR

Social behaviour change communication

Programme & Social Enablers

PMTCT

Other biomedical prevention

26

Care and treatment

17%

RNE 2014/15

49%

12%24%

15%

40%

Actual 2013/14

22

18%

HCT

Comprehensive condom programming

Medical male circumcision

Key populations

Proportional contribution by funding source shows potential program vulnerabilities (2013/14)

16 | R4D.org

10% 40%0% 30%20% 70% 90%60% 100%50% 80%

TB

Social Enablers

Medical Male Circumcision

PMTCT

Key Populations

HCT

Non IC

HIV not Disaggregated

Comprehensive Condom Programming

Other Biomedical Prevention

Programme Enablers

Care and Treatment

Social Behavior Change Communication

Global Fund SAGPEPFAR

What is the optimal balance from funding sources for each program?

At what point would a program be considered vulnerable?

Provincial Analysis – SAG, GF, PEPFAR

17 | R4D.org

Does variance in spending per HIV+ person imply technical and/or allocative efficiency? (2012/13, ZAR)

5,271

2,401

3,437

2,469

1,900

3,258

0

500

1,000

1,500

2,000

2,500

3,000

3,500

4,000

4,500

5,000

5,500

3.0

2.5

1.0

1.5

2.0

0.5

0.0

3.5

4.0

Ab

ove

Nat

ion

al

Nat

ion

al

HIV

Spe

nd

pe

r HIV

+ Pe

rson

(ZAR

)To

tal P

rovi

nci

al S

pe

nd

(B

illio

ns

ZAR

)

2,560

KwaZ

ulu

Nat

al

Gau

ten

g

1,9531,788

East

ern

Cap

e

No

rth

ern

Cap

e

Free

Sta

te

No

rth

Wes

t

Nat

& P

rov

no

t d

isag

greg

Mp

um

alan

ga

Lim

po

po

Wes

tern

Cap

e

SAGPEPFARGlobal Fund

National Avg. Spend per HIV+ Person

(2,480)

Provincial proportional HIV spending by intervention – by all 3 sources (2013/14). What is the best mix?

• Treatment & Care = on average 48% [26% in NW – 66% in GP]• HCT = average 6% [2% NW – 10% KZN]• MMC = average 3% [1% in EC, NW, LP, NC – 8% in MP]• PMTCT = average 3% [0% in NW – 6% in WC]• Prog. Enablers = average 7% [1% in KZN – 52% in NW]• National level = 40% not disagg.,15% prog.enablers, 12% SBCC, 9% treatment

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

KwaZuluNatal

Gauteng

EasternCape

WesternCape

Mpumalanga

NorthWest

FreeState

Limpopo

NorthernCape

Nat&Provnot

Na onal

AboveNa onal

Averageofprovinces

Careandtreatment

Compr.condomprogramming

HCT

HIVnotdisaggregated

Keypopula ons

Medicalmalecircumcision

NonSAIC

Otherbiomedicalpreven on

PMTCT

ProgrammeEnablers

Socialbehaviourchangecommunica on

SocialEnablers

IC % Priority Split

District Analysis: Gauteng, preliminary data example

(noting a lack of district specific HIV-prevalence data limiting comparison – pending)

21 | R4D.org

Gauteng DOH spending: total and per HIV+ person, per year (2013/14, ZAR)

1,0321,072

1,685

1,8191,861

0

500

1,000

1,500

2,000

ZAR

Tsh

wan

e

Joh

ann

esb

urg

Eku

rhu

len

i

Sed

iben

g

Wes

t R

and

Provincial Average (1,419)

1.0

0.7

0.8

0.9

0.4

0.5

0.6

0.1

0.3

0.0

0.2

Joh

ann

esb

urg

Wes

t R

and

Tsh

wan

e

No

t D

isag

g.

Pro

vin

cial

Lev

el

Sed

iben

g

0.20.2

HIV

Sp

en

din

g B

illio

ns

ZAR

Eku

rhu

len

i

0.4

0.0

1.0

0.1

0.5

DOH Voted

DOH HIV CG

Total spending Spending per HIV + person

Gauteng DOH District HIV Spending by intervention (2013/14)

• District split mostly similar• Provincial level mostly training• MMC and key population spending was not

disaggregated by district – so cannot see what was the % of total prevention spending per district – but appears to be less than 20%

0.0

0.1

0.5

0.2

0.7

0.4

0.9

0.6

0.3

0.8

1.0

Eku

rhu

len

i

Sed

iben

g

Job

urg

Tsh

wan

e

Bill

ion

s ZA

R

Pro

v. L

evel

Wes

t R

and

ND

Key Pops

Medical Male Circumcision

Other biomedical prevention

Prog Enablers HCT

Condom Programming

Non IC

Care and Trx

30%

0%

90%

20%

60%

50%

10%

70%

80%

100%

40%

Tsh

wan

e

Eku

rhu

len

i

Job

urg

Wes

t R

and

ND

Sed

iben

g

Pro

v. L

evel

Gauteng District Spending (%)Gauteng District Spending

Key Messages

• South Africa appears to be achieving, or moving towards, allocativeefficiency – at national level. Spending on mostly priority interventions with proven cost effectiveness according to Investment Case.

• The non-IC spending (15% of ’13/’14 total spending) was on interventions with limited evidence of efficacy. But are these the supportive (programme enablers) that are essential for greatest impact?

• Sub-national analysis shows variations warranting further examination, mostly explained by poor financial management systems.

• Important to align and consolidate expenditure data from key funding sources – make routine in-country (aligned to the public financial reporting structure & SHA for sustainability/ routine monitoring eg. Botswana, Zimbabwe, Kenya).

• Difficult to ascertain the optimal mix of services, and the most efficient allocation of resources – without an Investment Case or Optima.

• Lack of financial and prevalence data at the district-level hinders district comparisons. This is critical going forward to ascertain the best targeted use of resources.

• TB: limited cost, spending, and impact data.

Thanks to:

July 2016

• Global Fund• Investment Case Steering Committee: • NDOH, SANAC & UNAIDS• Nat Treasury• Gesine Meyer Rath• The R4D Team

Teresa Guthrie (Independent Consultant)guthriehealthfinancingconsult@gmail.comArjun Vasan (Results for Development Institute)avasan@r4d.org