The District Health BarometerA tool to monitor progress and support improvement of

equitable provision of primary health care

Fiorenza Monticelli, HST Conference, Indaba Conference Centre, JHB, 10&11 Oct 2007

What is the purpose of the DHB?

To function as a TOOL to monitor progress and support improvement of equitable provision of primary health care by: – Illustrating important aspects of the health system at

district level through analysis of indicators.– Ranking , classifying and analysing health districts (in

various groupings eg. metros, provinces, ISRDP sites), by indicators

– Comparing these indicators over time. – Portraying data as is, in order to highlight

performance and / or data quality issues. Thus, if the data look implausible - it merely highlights data quality issues at source which need attention.

Background• One of the main goals of the South African health

system is to provide equitable access to and quality of health care.

• Oversupply of data coexists with need for objective and transparent monitoring information.

• In order to meet this need HST successfully piloted a District Health Barometer (DHB) in 2005, in collaboration with the National Department of Health.

• Guided by an advisory committee made up of DOH managers at national, provincial and district level, including experts and stakeholders from the academic and research arenas.

• Two reports published: DHB year 1, covering 2004 data and DHB 2005/06. DHB 2006/07 is due January 2008

• Data sources :– DHIS, – StatsSA, – Treasury (BAS)* data , – National TB register – Directly from the district / province. (seldom)

• No data values received are changed (corrected)• Averages have been calculated, ie for metro and ISRDP

districts and PCE values are calculated from the data received from Treasury.

• The DHIS data is extracted at the end of June, once the official data set has been signed off by the districts and provinces and sent to Treasury and the NDOH.

• Data is illustrated in graphs, maps and tables to make it easier to read for comparative purposes.

• Where the data is not publicly available, such as with the DHIS and Treasury data, HST have asked for and received written permission to use the data for each DHB published.

• Data illustration examples follow• * North West (Walker System)

Methodology

Socio-economic quintile 5 (best)

Socio-economic quintile 4

Socio-economic quintile 3

Socio-economic quintile 2

Socio-economic quintile 1 (worst)

Deprivation index 2001 and

Socio-economic quintiles

Ugu

Uthukela*

Vhembe*, Sisonke*

Gr Sekukhune

Zululand

Umkhanyakude

Umzinyathi

Alfred Nzo

O.R.Tambo

* Not ISRDP

Deprivation index 2001

Northern KZN

Eastern Cape

Limpopo

% Household access

to piped water, 2001

Ilembe

C. Hani

Ukhahlamba,

•Ugu

•Uthukela

•Vhembe,

•Sisonke

•Gr Sekukhune

•Zululand

•Umkhanyakude

•Umzinyathi

•Alfred Nzo

•O.R.Tambo

There is a correlation between the districts which have less than 70% of households with access to piped

water, and those districts grouped into

the lowest socio-economic quintile

Per capita expenditure, 2005/06

0 50 100 150 200 250 300 350 400 450

Gr Sekhukhune DMMetsw eding DM

Siyanda DMG Sibande DMBohlabela DM

West Rand DMAmajuba DMNkangala DM

Waterberg DMCapricorn DM

Lejw eleputsw a DMA Nzo DM

Uthukela DMCacadu DM

Tshw ane MMN Mandela MMOverberg DM

Frances Baard DMUkhahlamba DMCape Winelands

Fezile Dabi DMO Tambo DM

iLembe DMEhlanzeni DM

Umzinyathi DMuMgungundlovu

Zululand DMBojanala Platinum

Sedibeng DMUgu DM

Mopani DMT Mofutsanyane

Motheo DMVhembe DM

Pixley ka Seme DMUthungulu DMAmathole DM

Eden DMJohannesburg MM

C Hani DMKgalagadi DMSouthern DM

eThekw ini MMEkurhuleni MM

Sisonke DMCentral DM

West Coast DMUmkhanyakude DM

Central Karoo DMCape Tow n MM

Xhariep DMNamakw a DMBophirima DMSouth Africa

Rand

EC

FS

GP

KZN

LP

MP

NC

NW

WC

SA

Per Capita Expenditure 2005/06

Sisonke R271

Umkhanyakude R309

Vhembe R237

Zululand R222

R232

Uthukela R195

Gr Sekhukune R115

Alfred Nzo R188

O.R Tambo R213

Change in per capita expenditure 2001/02 to 2005/06

-150 -100 -50 0 50 100 150 200 250

Ekurhuleni MMJohannesburg MM

Metsweding DMOverberg DM

Eden DMCape Town MM

Central Karoo DMWest Rand DM

Cape Winelands DMTshwane MMAmajuba DMSiyanda DM

uMgungundlovu DMGr Sekhukhune DM

West Coast DMBophirima DM

Uthukela DMSouthern DM

Waterberg DMUgu DM

Frances Baard DMZululand DM

N Mandela MMSedibeng DM

Pixley ka Seme DMBojanala Platinum

Ehlanzeni DMeThekwini MM

Umzinyathi DMMopani DM

Lejweleputswa DMUthungulu DMBohlabela DM

iLembe DMG Sibande DM

Sisonke DMCapricorn DM

Motheo DMAmathole DMVhembe DM

A Nzo DMKgalagadi DM

Fezile Dabi DMO Tambo DM

Cacadu DMNkangala DM

Central DMUmkhanyakude DM

Namakwa DMC Hani DM

T Mofutsanyane DMUkhahlamba DM

Xhariep DMSouth Africa

Rand (change)

EC

FS

GP

KZN

LP

MP

NC

NW

WC

SA

Absolute change in per capita

expenditure 2001/2 to 2005/6

CT, JHB, EKH

Overall the trends show there is a move towards

greater equity in health funding in

primary health care.

Average length of stay in a

district hospital 2005/6

Patients in many of the ISRDP districts

have a longer average length of stay in a district

hospital than their counterparts in

other better resourced districts.

R307

R319Immunisation Coverage 2005/6

• average immunisation coverage in SA = 90%

• ranges from a low of 66% for Metsweding to a high of 120% for Chris Hani

•undercount of population <1yr affects indicator

Do urban and rural children get vaccinated at the same levels?

Immunisation coverage averages 2003/4-2005/6

0

20

40

60

80

100

120

2003/04 2004/05 2005/06

Per

cen

tag

e ISRDP

Metro

SA

Immunisation drop out rate (DTP1-3), 2005/06

-5 0 5 10 15

O.R. TamboBohlabelaKgalagadi

UmkhanyakudeZululand

UkhahlambaThabo Mofutsanyane

Alfred NzoGreater Sekhukhune

UmzinyathiChris Hani

Central KarooUgu

ISRDP averageSouth Africa

Percentage

EC

FS

GP

KZN

LP

MP

NC

NW

WC

SA

Immunisation drop-out rate ( DTP1-3) 2005/6 in the ISRDP districts

TB Smear conversion rate 2004-2006 and TB cure rate 2003-2005 for districts in the lowest socio-economic quintile

DistrictSocio-economic quintile Smear conversion rate TB cure rate

2001 2004 2005 2006*Rank 2006*

change 2004-06* 2003 2004 2005*

Rank 2005*

change 2003-05*

Alfred Nzo (EC) 1 38.8 40.8 39.2 48 0.4 42.5 35.9 48.4 42 5.9

O.R. Tambo (EC) 1 34.9 40.4 48.4 36 13.5 35.3 35.3 64.0 26 28.7

Sisonke (KZN) 1 42.1 40.3 36.2 50 -5.9 23.7 50.6 49.2 41 25.5

Ugu (KZN) 1 41.0 35.4 42.0 43 1.0 37.6 33.4 33.8 50 -3.8

Umkhanyakude (KZN) 1 47.1 44.0 41.4 45 -5.7 30.0 34.9 41.8 46 11.8

Umzinyathi (KZN) 1 57.1 62.9 69.0 11 11.9 53.6 55.1 65.8 19 12.2

Uthukela (KZN) 1 35.5 42.0 40.9 47 5.4 36.1 40.2 44.5 44 8.4

Zululand (KZN) 1 48.0 47.5 58.0 22 10.0 40.4 51.3 66.3 18 25.9

Greater Sekhukhune[1] (LP) 1 53.7 40.1 41.4 45 -12.3 49.1 54.8 54.0 36 4.9

Vhembe (LP)12 1 65.0 73.1 76.8 2 11.8 63.4 75.1 72.0 5 8.6

South Africa 46.6 50.5 55.8 9.2 56.7 50.8 57.6 0.9

Children <5 not gaining weight rate, ISRDP and Metro districts compared 2005/06

Not gaining weight rate, 2005/06

0 2 4 6

Thabo Mofutsanyane

Zululand

Ugu

Kgalagadi

Chris Hani

Greater Sekhukhune

Ukhahlamba

Alfred Nzo

Umkhanyakude

Bohlabela

Central Karoo

Umzinyathi

O.R. Tambo

ISRDP average

South Africa

Percentage

EC

FS

GP

KZN

LP

MP

NC

NW

WC

SA

Not gaining weight rate, 2005/06

0 2 4 6

City of Tshwane

eThekwini

Ekurhuleni

Nelson Mandela Bay Metro

City of Cape Town

City of Johannesburg

Metro average

South Africa

Percentage

EC

FS

GP

KZN

LP

MP

NC

NW

WC

SA

ISRDP METRO

Not gaining weight rate, 2005/06

0 2 4 6

NamakwaUthukela

LejweleputswaSiyanda

Pixley ka SemeThabo Mofutsanyane

Fezile DabiXhariep

ZululandBophirima

EdenCacaduCentral

UguSisonke

KgalagadiChris Hani

Greater SekhukhuneFrances Baard

MotheoBojanala

UkhahlambaiLembe

Alfred NzoUmkhanyakude

Gert SibandeBohlabela

City of TshwaneCentral Karoo

WaterbergSouthern

eThekwiniNkangala

UmzinyathiSedibeng

EkurhuleniAmatholeEhlanzeniUthungulu

AmajubaUMgungundlovu

Nelson Mandela BayWest Rand

O.R. TamboOverberg

MopaniMetsweding

CapricornCape Winelands

VhembeCity of Cape Town

West CoastCity of Johannesburg

South Africa

Percentage

EC

FS

GP

KZN

LP

MP

NC

NW

WC

SA

Children <5 not gaining weight rate, all districts, 2005/06

Xhariep (FS)

Fezile Dabi (FS)

Thabo Mofutsanyane (FS)

Pixley ka Seme (NC)

Siyanda (NC)

Lejweleputswa (FS)

Namakwa (NC)

Zululand & Uthukela (KZN)

Stillbirth rate, 2005/06ISRDP and Metro districts

Stillbirth rate, 2005/06

0 20 40 60 80

South AfricaISRDP average

KgalagadiCentral Karoo

BohlabelaUmkhanyakude

Chris HaniUmzinyathi

Greater SekhukhuneUgu

Thabo MofutsanyaneUkhahlamba

Alfred NzoO.R. Tambo

Zululand

Stillbirths per 1000 births

EC

FS

GP

KZN

LP

MP

NC

NW

WC

SA

Stillbirth rate, 2005/06

0 20 40 60 80

South Africa

Metro average

City of Cape Town

City of Johannesburg

City of Tshwane

Ekurhuleni

eThekwini

Nelson Mandela Bay Metro

Stillbirths per 1000 births

EC

FS

GP

KZN

LP

MP

NC

NW

WC

SA

Why do Zululand, OR Tambo, A Nzo (ISRDP, rural), and

Nelson Mandela metro (urban) have such unacceptably

high stillbirth rates?

24.8

32.2

23.8•Developed countries 5 deaths per 1000

•Developing countries 30 deaths per 1000

District Profiles

Spider Graphs – a different way to view district performance

The closer to the centre the better

Further investigation

• Are the overall trends which show that there is a move towards greater equity in health funding in primary health care continuing and are they having the desired results in that the more disadvantaged areas are able to use these funds appropriately?

• Why do patients in many of the ISRDP and socio-economically disadvantaged districts, have a longer average length of stay in a district hospital than their counterparts in other better resourced districts?

• If it is true that many of the rural and disadvantaged areas are indeed managing to provide an effective immunisation service, why are other health areas in these districts not achieving the same results and what can be learnt from the immunisation programme that can positively influence other programmes?

Further investigationAre Vhembe’s results with respect to TB a true

reflection? Which lessons learnt and successful systems implemented in this district can be shared, particularly with those districts with populations that are similarly disadvantaged?

Why do Zululand, OR Tambo, A Nzo (ISRDP, rural), and Nelson Mandela metro (urban) have such unacceptably high stillbirth rates? Is it due to data quality or is there another reason?

When comparing two similarly socio-economically disadvantaged districts, such as Vhembe and Uthukela, why does their performance in health delivery differ so vastly, what can be learnt from Vhembe and what can be done to assist Uthukela district?

CONCLUSIONS• Routine service level data colleted is transformed into

information that leads to action

• Inequities between rural and urban districts and districts with populations of differing socio-economic levels are highlighted

• Improved feedback

• Tool for M&E, strategic planning at province & national level

• Improved transparency of performance of health sector

• Leads to continuous improvement of data quality of DHIS

THANK YOU

We acknowledge the National Department of Health for access to and use of their data for this publication and Atlantic

Philanthropies for funding the project.