Lancet Global_Appendix2.pdf

Supplementary appendix 2 This appendix formed part of the original submission and has been peer reviewed.

We post it as supplied by the authors.

Supplement to: Kruk ME, Gage A, Arsenault C, et al. High-quality health systems in the Sustainable Development Goals era: time for a revolution. Lancet Glob Health 2018;

published online Sept 5. http://dx.doi.org/10.1016/S2214-109X(18)30386-3.

High-quality health systems in the Sustainable Development Goals era: time for a revolution

Appendix 2: Additional Results

Table of Contents

Section 2: What quality of care are people receiving in LMICs today?

1. SDG Health Conditions

2. Country-specific means

a. Adherence to evidence-based guidelines (Figure 2A)

b. Variation in provider adherence to evidence-based guidelines for antenatal and sick child care in 16 low and middle-income countries

c. Systematic assessments

d. Diagnostic accuracy (Figure 2B)

e. Appropriate treatments (Figure 3)

f. Counseling and health education

g. Counseling for chronic diseases, WHO Stepwise Approach to Surveillance survey

h. Prevention and detection

i. Continuity and timely care

j. User experience (Figure 4)

k. Morbidity

l. Confidence and trust in health systems (Figure 7)

m. Why are people satisfied with poor care? (Panel 3)

n. Primary care and referral systems quality (Figure 9)

3. Quality of primary care facilities in nine countries (Figure 9)

4. Equity in maternal and child health care quality by maternal age in 91 low- and middle-income countries

5. Availability of data on quality of care across SDG conditions

Section 3: The ethical basis of high-quality health systems

6. Patient rights word cloud: Where is competent care?

7. Towards an Ethically Acceptable Minimum Standard of Quality: The Promise of South

Africa’s National Health Insurance Policy

Section 4: Measuring health system quality

1. Bibliographic analysis of quality-related literature

2. Clinical measurement methods

3. Quality of routine data quality for priority indicators on maternal and neonatal health: a case study from Gombe State, Nigeria

4. Data quality challenges for health-system quality indicators

5. Geospatial mortality and access to quality health facilities in India

6. Geographic coverage of public and private hospitals in Karnataka

7. Development and use of a high-quality health systems dashboard including effective coverage and care cascades

Section 5: Improving health systems at scale

1. HQSS improvement typology: Examples of improvement interventions organized by health system level

2. Country examples of improvement through the four universal actions

3. Increasing provider competence through pre-service education: Mexico and Tanzania ... 58

Section 2: What quality of care are people receiving in LMICs today?

1. SDG Health Conditions

Target Condition

3.1 Maternal health

3.2 Newborn and child health

3.3 HIV/AIDS, tuberculosis, malaria, hepatitis B, neglected tropical diseases

3.4 Cardiovascular disease, cancer, diabetes, chronic respiratory disease, mental health

3.5 Substance abuse

3.6, 16.1 Injuries from road traffic accidents and violence

3.7 Sexual and reproductive health

2. Country-specific means We present country-specific means of all the quality measures calculated from microdata presented in Section 2. See Appendix 1 (Methods) for data sources, years and definitions of the indices and indicators. An interactive data portal with country specific means and downloadable results is at www.hqssqualitydata.org. Country profiles are available at www.hqsscommission.org.

a. Adherence to evidence-based guidelines (Figure 2A)

Adherence to evidence-based guidelines for maternal and child health services in 18 countries, SPA and RBF surveys, 2007-2016

Country ANC Family planning Sick child care Labor and delivery

Burkina Faso 60% 49%

Central African Republic 50% 48%

Cameroon 66% 42%

Congo 57% 49%

Democratic Republic of the Congo 55% 47%

Ethiopia 47% 34% 35%

Haiti 45% 34% 29%

Kenya 63% 48% 44% 60%

Kyrgyzstan 88%

Malawi 50% 42% 30% 64%

Namibia 65% 46% 53%

Nepal 42% 26% 27%

Nigeria 61% 45%

Rwanda 62% 56% 31%

Senegal 62% 42% 39%

Tajikistan 47%

Tanzania 55% 45% 31%

Uganda 54% 55% 47% http://www.hqssqualitydata.org/ http://www.hqsscommission.org/

b. Variation in provider adherence to evidence-based guidelines for antenatal and sick child care in 16 low and middle-income countries

Bars represent the interquartile range of provider performance based on average quality scores (average of all consultations performed by the provider). Quality scores are the proportion of recommended items fulfilled by each provider during antenatal care and sick child consultations. Data are from Service

Provision Assessment surveys in nine countries and health facility baseline surveys of Results Based

Financing impact evaluations in seven countries (marked *). Individuals items included in the quality indices are defined in the Appendix 1. Antenatal care observations were not available in Tajikistan.

c. Systematic assessments

Postpartum checkups and antenatal care assessments, DHS and MICS, 2007-2016

Country

Blood pressure, urine and blood samples taken during antenatal care with skilled provider

Postpartum checkup for mothers in a health facility after delivery and before discharge

Afghanistan 22% 66%

Albania 83% 86%

Algeria 73% 89%

Argentina 99%

Armenia 100% 98%

Bangladesh 55% 92%

Belarus 100% 100%

Belize 99% 96%

Benin 80% 69%

Bhutan 90%

Bolivia 67% 97%

Bosnia Herzegovina 98%

Burkina Faso 60% 95%

Burundi 6% 44%

Cambodia 43% 95%

Cameroon 85% 58%

Central African Republic 71%

Chad 37% 55%

Colombia 96%

Comoros 80% 73%

Congo 92% 79%

Costa Rica 95%

Cote d'Ivoire 67% 92%

Cuba 100% 100%

Democratic Republic of the

Congo 46% 56%

Dominican Republic 98% 96%

Egypt 74% 91%

El Salvador 97% 95%

Ethiopia 56% 44%

Gabon 95% 74%

Ghana 96% 95%

Guatemala 65% 95%

Guinea 52% 72%

Guinea Bissau 81% 93%

Guyana 93% 86%

Haiti 80% 78%

Honduras 89% 98%

Indonesia 27% 95%

Iraq 83%

Jamaica 99%

Jordan 94% 85%

Kazakhstan 100% 98%

Kenya 85% 76%

Kosovo 83% 86%

Kyrgyzstan 25% 90%

Laos 33% 83%

Lesotho 80% 85%

Liberia 84% 90%

Macedonia 95%

Madagascar 22% 89%

Malawi 30% 47%

Maldives 96% 95%

Mali 43% 65%

Mauritania 68%

Mexico 92% 96%

Moldova 99% 94%

Mongolia 95% 96%

Montenegro 97% 96%

Mozambique 39% 56%

Namibia 94% 88%

Nepal 43% 89%

Niger 29% 83%

Nigeria 78% 81%

Pakistan 52% 83%

Palestine 96% 91%

Panama 93% 100%

Peru 85%

Philippines 57% 96%

Rwanda 53% 44%

Saint Lucia 97% 88%

Sao Tome and Principe 96% 91%

Senegal 79% 96%

Serbia 95%

Sierra Leone 72% 93%

South Sudan 29%

Sudan 79% 78%

Suriname 95%

Swaziland 86% 27%

Tajikistan 91% 96%

Tanzania 52% 51%

Thailand 99%

Timor-Leste 9% 87%

Togo 87% 93%

Tunisia 91% 93%

Turkmenistan 99% 100%

Uganda 18% 51%

Ukraine 100% 98%

Vietnam 59% 95%

Yemen 70% 52%

Zambia 42% 88%

Zimbabwe 67% 71%

d. Diagnostic accuracy (Figure 2B)

Proportion of providers identifying correct diagnoses based on clinical vignettes administered in six Sub-Saharan African countries, SDI and SPA surveys, 2012-2014

Country Diabetes Diarrhea

Malaria and anemia

Neonatal

Asphyxia Pneumonia

Post-partum hemorrhage Tuberculosis

Ethiopia 11% 45% 58% 52%

Kenya 81% 78% 26% 61% 83% 94% 79%

Nigeria 34% 24% 12% 28% 42% 51% 55%

Tanzania 80% 48% 43% 50% 76% 87% 86%

Togo 45% 30% 0% 2% 36% 57% 82%

Uganda 52% 42% 31% 53% 49% 67% 78%

e. Appropriate treatments (Figure 3)

Antenatal care treatments, oral rehydration therapy (ORT) for child diarrhea, antibiotics for child pneumonia, antiretroviral therapy among those with HIV, minimally adequate treatment for depression, 2007-2016

Received appropriate number of tetanus vaccinations and iron supplements

(DHS)

Received

ORT when seeking care at a facility for diarrhea

(DHS)

Received antibiotics when seeking care at a facility for symptoms of pneumonia

(DHS, MICS)

Percent of people who know their

HIV status who are on

ART

(UNAIDS)

Received minimally adequate treatment among those diagnosed with MDD

(WMHS)

Afghanistan 34% 63% 57% 26%

Albania 25% 86% 64%

Algeria 41% 79% 99%

Angola 55%

Argentina 45% 47% 82%

Armenia 73% 59%

Azerbaijan 52%

Bangladesh 93% 46%

Belarus 79% 50%

Belize 44% 77%

Benin 66% 85% 37%

Bhutan 63% 57%

Bolivia 40% 64% 47%

Bosnia and

Herzegovina 85%

Botswana 99%

Bulgaria 36% 7%

Burkina Faso 80% 41% 52%

Burundi 37% 62% 48% 82%

Cambodia 72% 47%

Cameroon 64% 58% 45% 63%

Central African

Republic 77% 57%

Chad 69% 54% 36%

China 74%

Colombia 58% 5%

Comoros 33% 84% 23% 92%

Congo 62% 64% 54% 79%

Costa Rica 19% 77%

Cote d'Ivoire 58% 48% 21% 71%

Cuba 65% 66% 81%

Democratic

Republic of the

Congo

41% 65% 41% 89%

Djibouti 30%

Dominican

Republic 81% 74% 66%

Ecuador 56%

Egypt 39% 41% 48%

El Salvador 74% 49%

Ethiopia 37% 65% 23% 88%

Fiji 36%

Gabon 70% 55% 56% 80%

Gambia 86%

Georgia 74%

Ghana 64% 65% 42% 77%

Guatemala 60% 71% 55%

Guinea 79% 59% 37%

Guinea Bissau 70% 40%

Guyana 17% 79% 20% 84%

Haiti 65% 80% 94%

Honduras 29% 81% 83%

India 63%

Indonesia 45% 59% 36%

Iran 37%

Iraq 52% 77% 3%

Jamaica 31% 3% 45%

Jordan 6% 39%

Kazakhstan 42%

Kenya 50% 84% 53%

Kosovo 48%

Kyrgyzstan 72% 47% 46%

Laos 50% 74%

Lebanon 7%

Lesotho 55% 92% 74%

Liberia 87% 77% 51% 57%

Macedonia 94%

Madagascar 40% 55% 50% 76%

Malawi 67% 87% 46% 94%

Malaysia 39%

Maldives 55% 69%

Mali 51% 71%

Mauritania 55% 62%

Mexico 62% 77% 6%

Mongolia 67% 95%

Montenegro 95% 67%

Morocco 77%

Mozambique 59% 81% 13% 88%

Namibia 40% 92% 49% 84%

Nepal 80% 67% 72%

Nicaragua 51%

Niger 65% 79% 10% 90%

Nigeria 70% 65% 38% 88% 0%

Pakistan 43% 50%

Palestine 82%

Panama 82% 73%

Papua New

Guinea 64%

Paraguay 53%

Peru 51% 56% 80% 1%

Philippines 67% 82% 48%

Republic of

Moldova 89% 38%

Romania 77% 14%

Rwanda 46% 63% 56% 91%

Saint Lucia 7% 0%

Sao Tome and

Principe 66% 62%

Senegal 73% 49% 18%

Sierra Leone 87% 90% 42% 75%

South Africa 65%

South Sudan 41%

Sri Lanka 56%

Sudan 51% 27%

Suriname 32% 80% 79%

Swaziland 64% 97% 2%

Tajikistan 87% 63%

Tanzania 54% 29% 88%

Thailand 75% 47% 75%

Timor-Leste 64% 90% 44%

Togo 68% 50% 37% 81%

Tunisia 18% 68% 50%

Turkmenistan 100%

Uganda 56% 58% 46% 91%

Ukraine 66%

Uzbekistan 69%

Vietnam 64% 94% 67%

Yemen 7% 49%

Zambia 52% 89% 55% 99%

Zimbabwe 43% 91% 43% 99%

f. Counseling and health education

Family planning, antenatal care, sick-child care, chronic conditions, 2007-2016

Told about potential side effects when first prescribed contraceptives

(DHS, MICS)

Told about danger signs or where to go in case of complications during ANC with skilled provider (DHS, MICS)

Counseled on

HIV during

ANC with skilled provider

(DHS, MICS)

Stated the child diagnosis during sick child consultation

(SPA, RBF)

Doctor gave advice on exercise, diet and stress among those with at least one chronic condition

(IDB)

Afghanistan 48% 56%

Albania 63% 48%

Algeria 5%

Argentina 69%

Armenia 73% 57%

Bangladesh 46%

Belarus 84%

Belize 88%

Benin 57% 64% 77%

Bhutan 65%

Bolivia 68% 67%

Bosnia

Herzegovina 18%

Brazil 46%

Burkina Faso 73% 53% 67% 40%

Burundi 35% 36% 74%

Cambodia 80% 82% 86%

Cameroon 65% 49% 80% 52%

Central

African

Republic

89% 48%

Chad 62% 47% 57%

Colombia 59% 86% 93% 53%

Comoros 56% 33% 34%

Congo 56% 62% 62% 63%

Costa Rica 71%

Cote d'Ivoire 46% 37% 65%

Cuba 97%

Democratic

Republic of the Congo

58% 63% 59% 51%

Dominican

Republic 40% 72% 87%

Egypt 45% 46%

El Salvador 94% 60%

Ethiopia 41% 45% 62% 30%

Gabon 54% 48% 85%

Ghana 64% 84% 84%

Guatemala 50% 75% 67%

Guinea 51% 34% 25%

Guinea

Bissau 79%

Guyana 52% 68% 89%

Haiti 70% 65% 82% 10%

Honduras 43% 76% 79%

Indonesia 37% 53%

Iraq 6%

Jamaica 48%

Jordan 70% 34%

Kazakhstan 83%

Kenya 54% 58% 97% 55%

Kosovo 11%

Kyrgyzstan 76% 80% 71%

Laos 53%

Lesotho 40% 63% 99%

Liberia 75% 70% 90%

Macedonia

Madagascar 58% 50% 50%

Malawi 74% 98% 41%

Maldives 45% 52%

Mali 53% 42% 54%

Mauritania 27%

Mexico 79% 69%

Moldova 90%

Mongolia 66%

Montenegro 6%

Mozambique 57% 40% 77%

Namibia 58% 74% 98% 49%

Nepal 62% 79% 35%

Niger 40% 56% 50%

Nigeria 62% 68% 73% 56%

Pakistan 34% 51%

Panama 92% 62%

Peru 65% 90%

Philippines 68% 81%

Rwanda 65% 79% 99% 22%

Saint Lucia 97%

Sao Tome and Principe 93%

Senegal 75% 49% 7%

Serbia

Sierra Leone 78% 90% 88%

South Sudan 44%

Sudan 16%

Suriname 86%

Swaziland 64% 54% 86%

Tajikistan 82% 77% 64% 66%

Tanzania 56% 49%

Thailand 90%

Timor-Leste 61% 56%

Togo 78% 69% 89%

Tunisia 7%

Turkmenistan 91%

Uganda 56% 51% 92% 51%

Ukraine 84% 38% 86%

Vietnam 65%

Yemen 56% 50%

Zambia 80% 89% 97%

Zimbabwe 58% 97%

g. Counseling for chronic diseases, WHO Stepwise Approach to Surveillance survey

Lindsay M. Jaacks, Glennis Andall-Brereton, Mongal Singh Gurung, Brice Wilfried, Joseph

Kibachio, Velephi Nhlengetfwa Okello, David Guwatudde, Clare Flanagan, Pascal Geldsetzer, Jennifer Manne-Goehler

Tobacco use, excess weight, unhealthy diets, and physical inactivity are the leading risk factors for noncommunicable diseases, yet there is limited surveillance of health system performance in terms of providing lifestyle advice for these risk factors. To address this gap, we evaluated data from the WHO Stepwise Approach to Surveillance (STEPS) survey. No question on the WHO

STEPS survey instrument explicitly queries whether or not the respondent accessed the health system in a specified period of time. Therefore, in order to verify that the respondent had indeed accessed the health system, we calculated the proportion of individuals who reported being diagnosed by a physician or other health care provider with high blood glucose, high blood pressure, or high blood cholesterol in the past 12 months that also reported receiving lifestyle advice.

Data on these variables were available for seven countries with per capita health expenditures ranging from US$35 (Burkina Faso) to US$575 (Saint Vincent & The Grenadines). Individuals were most likely to receive advice relating to dietary change, and least likely to receive advice about tobacco use.

Summary of surveyed countries and proportion of individuals who reported being diagnosed by a physician or other health care provider with high blood glucose, high blood pressure, or high blood cholesterol in the past 12 months that also reported receiving lifestyle advice.

Survey Year of

Survey

GDP

per capita+

Health expend per capita++

Quit using tobacco or don’t start

Maintain healthy body weight or lose weight

Start or do more physical activity

Dietary change*

Bhutan 2014 2,804.0 89

42.3

(36.6, 48.2)

43.6

(38.5, 48.9)

53.3

(48.3, 58.3)

87.6

(84.3, 90.3)

Burkina Faso 2013 649.7 35 22.3

(15.9, 30.3)

25.4

(17.6, 35.2)

34.3

(27.0, 42.3)

69.4

(62.5, 75.5)

Guyana 2016 4,456.6 222

16.8

(13.7, 20.5)

52.6

(48.1, 57.2)

51.3

(46.7, 55.9)

69.5

(65.7, 73.1)

Kenya 2015 1,455.4 78 10.9

(7.6, 15.2)

22.9

(17.9, 28.8)

27.1

(21.1, 34.1)

51.5

(43.0, 59.8)

Saint Vincent

& The

Grenadines

2013 7,030.1 575 5.1

(3.2, 8.1)

45.7

(40.1, 51.4)

44.4

(40.7, 48.3)

64.6

(60.8, 68.3)

Swaziland 2014 2,775.2 248 37.8

(31.5, 44.6)

54.8

(48.6, 61.0)

61.9

(55.9, 67.6)

81.3

(76.1, 85.7)

Uganda 2014 615.3 52 21.3

(15.6, 28.2)

27.4

(21.0, 34.9)

27.7

(21.3, 35.2)

51.9

(44.8, 59.0)

All analyses were conducted in Stata and took into account the sample design. Values are weighted percent (95% confidence interval).

+ 2016 Estimates in Current US$, per The World Bank.

++ 2014 Estimates in Current US$, per The World Bank.

*Includes responding “yes” to any one of the following: reduce salt, eat ≥5 servings of fruit and vegetables a day, and reduce fat.

Proportions in Uganda may be artificially lower because “all conditions” does not include hypercholesterolemia (was not asked in STEPS survey).

h. Prevention and detection

Preventive exams and HIV screening, 2007-2017

Proportion up-to-date with preventive exams

(IDB)

Women aged 50-69 who had a mammogram in past 3 years (SAGE)

Tested for HIV during

ANC (DHS, MICS)

Algeria 3%

Argentina 87%

Belarus 97%

Belize 89%

Benin 55%

Bhutan 56%

Bosnia Herzegovina 8%

Brazil 57%

Burkina Faso 35%

Burundi 45%

Cambodia 77%

Cameroon 71%

Central African Republic 35%

Chad 29%

China 21%

Colombia 43% 94%

Comoros 20%

Congo 43%

Costa Rica 83%

Cote d'Ivoire 49%

Cuba 99%

Democratic Republic of the Congo 24%

Dominican Republic 98%

El Salvador 34% 94%

Ethiopia 59%

Gabon 84%

Ghana 2% 75%

Guatemala 57%

Guinea 13%

Guinea Bissau 47%

Guyana 85%

Haiti 69%

Honduras 80%

India 1%

Iraq 1%

Jamaica 57%

Kazakhstan 93%

Kenya 98%

Kosovo 3%

Kyrgyzstan 46%

Laos 20%

Lesotho 98%

Liberia 74%

Madagascar 20%

Malawi 97%

Mali 23%

Mauritania 27%

Mexico 60% 45% 68%

Moldova 95%

Montenegro 2%

Mozambique 69%

Namibia 98%

Niger 36%

Nigeria 51%

Panama 39% 92%

Russia 40%

Rwanda 99%

Saint Lucia 100%

Sao Tome and Principe 94%

Sierra Leone 71%

South Africa 10%

South Sudan 29%

Sudan 9%

Suriname 95%

Swaziland 56%

Tajikistan 53%

Thailand 83%

Togo 68%

Tunisia 4%

Turkmenistan 93%

Uganda 81%

Ukraine 90%

Vietnam 43%

Zambia 94%

Zimbabwe 98%

i. Continuity and timely care

Retention in antenatal care and child vaccination and timely postpartum checkup for mothers, DHS and MICS, 2007-2016

Retained in antenatal care until the 4th visit among women who had at least 1 visit

Received the third dose for DTP by one year of age, among those who had received the 1st dose

Woman was checked within 1 hour of delivery after giving birth in a facility

Afghanistan 30% 77% 58%

Albania 69% 99%

Algeria 73% 94%

Argentina 96%

Armenia 96% 97% 25%

Bangladesh 40% 96% 58%

Belarus 100%

Belize 94% 81%

Benin 70% 86% 29%

Bhutan 80%

Bolivia 80% 90%

Bosnia Herzegovina 98% 91%

Brazil

Burkina Faso 35% 95% 33%

Burundi 34% 97% 48%

Cambodia 79% 92% 64%

Cameroon 74% 81% 28%

Central African Republic 54% 48%

Chad 50% 60% 29%

Colombia 94%

Comoros 62% 85% 27%

Congo 85% 76% 19%

Costa Rica 92% 95%

Cote d'Ivoire 48% 82% 53%

Cuba 100% 97%

Democratic Republic of the Congo 54% 77% 22%

Dominican Republic 97% 93% 35%

Egypt 92% 98% 51%

El Salvador 94% 94%

Ethiopia 51% 70%

Gabon 83% 75% 18%

Ghana 90% 87% 42%

Guatemala 90% 94% 5%

Guinea 65% 66% 53%

Guinea Bissau 72% 88%

Guyana 94% 93%

Haiti 75% 75% 26%

Honduras 92% 98% 15%

Indonesia 91% 82% 43%

Iraq 67% 80%

Jamaica 95% 94%

Jordan 95% 99% 23%

Kazakhstan 99% 96%

Kenya 60% 91% 43%

Kosovo 94% 97%

Kyrgyzstan 67% 92%

Laos 66% 72%

Lesotho 79% 88% 18%

Liberia 83% 75% 46%

Macedonia 98% 96%

Madagascar 55% 89%

Malawi 52% 95%

Maldives 98% 100%

Mali 56% 79% 48%

Mauritania 63% 62%

Mexico 96% 77%

Moldova 98% 97%

Mongolia 93% 99%

Montenegro 96% 94%

Mozambique 56% 87%

Namibia 85% 85% 25%

Nepal 59% 97% 53%

Niger 38% 79% 51%

Nigeria 81% 79% 52%

Pakistan 48% 84% 72%

Palestine 96% 99%

Panama 94% 86%

Peru 95% 86%

Philippines 88% 92% 37%

Rwanda 44% 99% 49%

Saint Lucia 98%

Sao Tome and Principe 91% 95%

Senegal 57% 95% 68%

Serbia 96% 96%

Sierra Leone 89% 83% 53%

South Sudan 48% 50%

Sudan 65% 85%

Suriname 93% 79%

Swaziland 84% 94%

Tajikistan 66% 96% 48%

Tanzania 52% 92%

Thailand 97% 93%

Timor Leste 63% 85%

Togo 62% 88% 52%

Tunisia 87% 94%

Turkmenistan 97% 99%

Uganda 51% 78% 36%

Ukraine 98%

Vietnam 77% 92%

Yemen 41% 80% 60%

Zambia 57% 92% 51%

Zimbabwe 81% 94%

j. User experience (Figure 4)

Respect, communication, time spent with providers and wait time, 2007-2017

Country

Never experienced lack of attention or respect from public facility staff (AFRO)

Rated respect at last outpatient visit as good or better

(HQSS)

Regular GP explains things in a way that is easy to understand

(IDB, CWF)

Rated how the provider listened at last outpatient visit as good or better

(HQSS)

Did not have a problem with the amount of explanation received during this visit (SPA)

Regular

GP

spends enough time with you (IDB, CWF)

Rated how much time the provider spent with patient at last outpatient visit as good or better (HQSS)

Rated wait time at last outpatient visit as good or better

(HQSS)

Did not have a problem with the wait time at this visit

(SPA)

Algeria 43%

Argentina 89% 83% 76% 54%

Australia 88% 84%

Benin 52%

Botswana 59%

Brazil 67% 35%

Burkina

Faso 53%

Burundi 59%

Cameroon 30%

Canada 88% 80%

Cape Verde 46%

China 60% 63% 46% 44%

Colombia 79% 73%

Cote d'Ivoire 41%

Egypt 50%

El Salvador 72% 61%

Ethiopia 77% 68%

France 88% 81%

Germany 94% 88%

Ghana 65% 82% 80% 81% 61%

Guinea 37%

Haiti 92% 80%

India 75% 73% 71% 66%

Indonesia 76% 64% 55% 54%

Jamaica 69% 60%

Kenya 44% 79% 81% 89% 73% 59% 58%

Lebanon 75% 73% 67% 58%

Lesotho 49%

Liberia 48%

Madagascar 39%

Malawi 38% 91% 71%

Mali 30%

Mauritius 34%

Mexico 73% 79% 71% 74% 64% 55%

Morocco 40% 55% 53% 53% 46%

Mozambique 65%

Namibia 28% 90% 61%

Nepal 83% 72%

Netherlands 90% 85%

New

Zealand 91% 88%

Niger 53%

Nigeria 47% 74% 87% 77% 72%

Norway 88% 80%

Panama 76% 54%

Rwanda 84% 59%

Senegal 65% 89% 89% 99% 73% 71% 77%

Sierra Leone 25%

South Africa 45% 77% 79% 72% 63%

Sudan 50%

Swaziland 32%

Sweden 86% 79%

Switzerland 88% 85%

Tanzania 42% 93% 72%

Togo 47%

Tunisia 20%

Uganda 9% 79% 52%

United

Kingdom 94% 87%

United

States 88% 81%

Zambia 21%

Zimbabwe 36%

k. Morbidity

Obstetric fistulas and uncontrolled lung disease and depression, 2007-2016

Number of women who reported ever experiencing symptoms of a fistula among those whose last birth was attended by a skilled provider per 1,000

(DHS)

Percent of people receiving treatments for chronic lung disease who report no symptoms from the disease in the past 2 weeks (SAGE)

Percent of people receiving treatments for depression who report no symptoms from the disease in the past 2 weeks (SAGE)

Afghanistan 27.3

Benin 6.7

Burkina Faso 0.5

Cameroon 5

Chad 25.4

China 63% 16%

Comoros 15

Congo 1.7

Cote d'Ivoire 45.2

Democratic Republic of the Congo 2.6

Ethiopia 3.6

Ghana 66% 0%

Guinea 5.7

Haiti 9

India 56% 10%

Kenya 12.3

Malawi 6.3

Mali 4.4

Mexico 10% 8%

Niger 2.9

Nigeria 4.3

Pakistan 32.3

Russia 59% 5%

Senegal 1.7

Sierra Leone 8.9

South Africa 45% 2%

Tanzania 6.1

Togo 11.3

Uganda 21.6

Yemen 8.6

Zambia 4.7

l. Confidence and trust in health systems (Figure 7)

Believes the system works pretty well and only minor changes are needed (CWF, HQSS, IDB)

Thinks government handles improving basic health services well (AFRO)

Confident that if sick will receive the most effective treatment (IDB)

Confident that if got sick tomorrow, could get the care s/he needed (HQSS)

Algeria 66%

Argentina 29% 93%

Australia 48%

Benin 85%

Botswana 57%

Brazil 4% 34%

Burkina Faso 67%

Burundi 63%

Cameroon 67%

Canada 42%

Cape Verde 61%

China 27% 90%

Colombia 17% 56%

Cote d'Ivoire 67%

Egypt 52%

El Salvador 10% 55%

Ethiopia

France 40%

Germany 42%

Ghana 23% 58% 90%

Guinea 65%

Haiti

India 28% 70%

Indonesia 41% 89%

Jamaica 12% 40%

Kenya 28% 76% 87%

Lebanon 37% 85%

Lesotho 47%

Liberia 43%

Madagascar 59%

Malawi 43%

Mali 49%

Mauritius 60%

Mexico (IDB) 17% 75%

Mexico (HQSS) 24% 82%

Morocco 26% 69% 68%

Mozambique 86%

Namibia 71%

Nepal

Netherlands 51%

New Zealand 47%

Niger 56%

Nigeria 30% 46% 94%

Norway 46%

Panama 9% 57%

Rwanda

Senegal 29% 87% 67%

Sierra Leone 54%

South Africa 35% 52% 87%

Sudan 41%

Swaziland 77%

Sweden 44%

Switzerland 54%

Tanzania 85%

Togo 67%

Tunisia 16%

Uganda 31%

United Kingdom 63%

United States 25%

Zambia 34%

Zimbabwe 26%

m. Why are people satisfied with poor care? (Panel 3)

Satisfaction with services and expectations of quality, 2007-2017

Very satisfied with family planning visit

(SPA)

Very satisfied with ANC visit

(SPA)

Very satisfied with sick child visit (SPA)

Rated hypertension patient vignette as good, very good or excellent quality (HQSS)

Argentina 54%

China 47%

Ethiopia 82% 70% 61%

Ghana 56%

Haiti 97% 91% 91%

India 59%

Indonesia 55%

Kenya 86% 83% 77% 43%

Lebanon 58%

Malawi 93% 93% 81%

Mexico 49%

Morocco 42%

Namibia 80% 76% 76%

Nepal 52% 43% 40%

Nigeria 58%

Senegal 97% 99% 98% 62%

South Africa 50%

Tanzania 91% 89% 74%

n. Primary care and referral systems quality (Figure 9)

Results from the Inter-American Development Bank survey on primary care, 2013

Country

Regular PCP knew important information about their medical history

Regular PCP helps coordinate care

Used emergency room for a condition that could have been treated at the primary care level

Specialist had basic medical information about the patient from the GP

GP

subsequently had the results after the specialist visit

Brazil 43% 26% 30% 36% 33%

Colombia 74% 47% 52% 47% 53%

El Salvador 69% 43% 46% 60% 53%

Jamaica 60% 34% 59% 63% 59%

Mexico 76% 49% 37% 64% 69%

Panama 74% 38% 40% 64% 58%

3. Quality of primary care facilities in nine countries (Figure 9)

Erlyn Macarayan

Full manuscript forthcoming

Using data from Service Provision Assessment surveys, the Commission examined the performance of primary care facilities in nine countries across three domains of quality identified in our framework: evidence-based care, competent systems and two elements of user experience

(client focus and clear communication). The analysis revealed important gaps in the quality of primary care across all three domains, with a mean primary care quality score of 0.41 out of 1, ranging from 0.32 on average in Ethiopia to 0.46 in Namibia.

Quality of primary health care facilities in nine countries sorted by GDP per capita

Data are from Service Provision Assessment surveys, 2007-2016. Numbers represent mean scores for each of the three subdomains (competent systems, evidence-based care, client focus and clear communication) ranging from 0 (lowest) to 1 (highest). The x axis is shown from 0 to 1.5, with 3 being the maximum achievable score by scoring 1 on each of the three subdomains.

4. Equity in maternal and child health care quality by maternal age in 91 low- and middle-income countries

DTP Diphtheria tetanus pertussis vaccine. Data are from Demographic and Health Surveys and Multiple

Indicator Cluster Surveys (2007-2016). Age groups are pooled across countries and sampling weights are adjusted to weigh countries equally. Indicators are defined in Appendix 1.

5. Availability of data on quality of care across SDG conditions

In Section 2, we aimed to describe quality across all health conditions addressed in the SDG targets. The figure below represents availability of data on quality across health conditions found during searches performed in the context of the Commission, with lighter colors indicating the least amount of nationally representative data on quality of care. Darker boxes reflect more data, but that is still far from perfect as discussed in section 4 (measuring health system quality).

We found mismatch between the disease burden and the availability of data on quality of care in

LMICs. For example, cardiovascular diseases and diabetes together account for 20% of DALYs in LMICs, however there is currently little data on the quality of care provided for these two conditions. Injuries are the third leading cause of DALYs in LMICs which bear the majority of the world’s injury burden, yet there is virtually no nationally representative data on the quality of trauma and emergency care in LMICs.1 Similarly, cancers rank fifth in terms of DALYs and their contribution is likely to increase, but no cross-national data on quality of the cancer care currently exist in LMICs.2 Finally, data on the quality of care for mental health and substance abuse is lacking in most countries.3

A n te n at al c ar e P o st p ar tu m c ar e af te r fa ci li ty b ir th C h il d h ea lt h

0 20 40 60 80 100

Retention through 4 visits

Iron supplements

Blood pressure, urine and blood samples

Checkup within 1 hour

Any checkup

DTP retention, all 3 doses

Proportion (%)

Age Group

20−35

Section 3: The ethical basis of high-quality health systems

6. Patient rights word cloud: Where is competent care?

Supporting material for: Defining a national quality guarantee

People implicitly count on receiving good quality care when they visit a clinic. However, today this is not generally well specified. Many countries around the world have developed patients’ rights charters outlining their approach to patient care and providing an ethical basis for a national quality guarantee. While these charters contain many of the same basic principles, there is significant variation in their length, scope, and detail. The figure below depicts the most common themes found in 10 sampled patient rights charters with the size of the word indicating the frequency of its use. Many of the guarantees to patients revolve around legal and human rights frameworks, for example the right to information, privacy, consent, and confidentiality.

However, not much is promised to the patient regarding competent care. This may be because standards of care are assumed in the health system but a basic level of competent care should be promised to the patient by the government.

7. Towards an Ethically Acceptable Minimum Standard of Quality: The Promise of

South Africa’s National Health Insurance Policy Ames Dhai

Supporting material for: Defining a national quality guarantee

An ethically acceptable minimum standard of quality (EAMSQ) in South Africa must be such that it responds to the Constitutional commitment to the right to health care and it facilitates the realization of that right. The elements proposed below form not only the basis for an EAMSQ, but also the basis of the right to the highest attainable standard of care. All processes during implementation of South Africa’s National Health Insurance Policy (NHIP) must consider the 12 recommended elements. This will assist as the journey progresses towards a healthy and productive nation where all enjoy benefits. Quality in itself is to be viewed as a social investment.

1. Professionalism, integrity, trust, compassion and caring: this element is comprehensively covered in the Oaths taken by health care professionals upon graduation. They are also included in the Codes and Policies of South Africa’s Regulatory Bodies, including the

Health Profession’s Council of South Africa and the South African Nursing Council.

2. Respect for Dignity: responds to section 10 of the Bill of Rights of the South African

Constitution affirms everyone’s right to inherent dignity and to have their dignity respected and protected. This element is further borne out in the National Health Act and other similar Acts of Parliament, the National Patients’ Rights Charter, Oaths taken by graduates, Codes and Policies of the Statutory Councils.

3. The principle of need: is aligned to three principles of the NHIP. The right to access healthcare which the NHI undertakes to ensure. Based on its principle of equity, the NHI will ensure a fair and just health care system for all and timely access will be provided to health services. The appropriateness principle ensures the delivery of healthcare at appropriate levels and will be tailored to local needs.

4. Fair and impartial processes for decision-making: Responds to the NHIP principle of equity.

5. Benefit maximization: responds to several principles in the NHIP, including the equity, affordability, efficiency and effectiveness principles. Affordability as a principle warrants that health services will be procured at reasonable cost taking into account the need for sustainability within the context of South Africa’s resources. The efficiency principle undertakes to allocate and utilize health care resources in a manner that optimises value for money. Allocative and productive efficiency will be combined by maximising health outcome for a given cost while using the given resources to maximum advantage. The welfare of the community will be maximised by achieving the right mixture of healthcare programs for the entire population. The NHI’s effectiveness principle pledges that the healthcare interventions covered under NHI will result in desired and expected outcomes in everyday settings. It will ensure that the health system meets (ethically) acceptable standards of quality and achieve positive health outcomes.

6. Priority to the worse off: is aligned to social solidarity, the second principle of the NHIP, will protect against financial risk by providing for financial risk pooling to enable cross-subsidisation between the rich and poor, young and old, and healthy and sick. This element also responds to the right to access healthcare principle in the policy.

7. Equity and equality of opportunity: aligns with the equity and healthcare as a public good principles. The latter underscores health care as a social investment and will act as a safeguard against it being treated like any other commodity of trade.

8. Acceptable trade-offs: responds to the equity and right to healthcare principles. In addition, section 36 of the Bill of Rights limits rights to the extent that the limitation is reasonable and justifiable in an open and democratic society based on human dignity, equality and freedom. The limitation takes into account all relevant factors including the nature of the right; the importance of the purpose of the limitation; the nature and extent of the limitation; the relationship between the limitation and its purpose; and less restrictive means to achieve the purpose.

9. Financial risk protection: this responds to social solidarity and health care as a public good principles.

10. Robust public accountability: To ensure that the procedures in determining quality standards and distributing healthcare equitably are fair, the NHIP describes the establishment of interim structures in preparation for the NHI Fund. These include the

Ministerial Advisory Committee on Health Care Benefits for National Health Insurance and the Ministerial Advisory Committee on Health Technology Assessment for National

Health Insurance. The former Committee will be a precursor to the NHI Benefits

Advisory Committee which will recommend the process of priority setting to inform the decision-making processes when determining the benefits to be covered. There are several sub-committees described in the Policy. The role of the Office of Health

Standards Compliance which was established in 2013 is to ensure compliance with the

National Quality Standards for Health. These Standards are based on seven domains and six National Core Standards for quality. The domains are: patient rights; patient safety;

clinical governance and care; clinical support services; public health; leadership and corporate governance; operational management including financial, asset and human resources management; and facilities and infrastructure. The National Core Standards for quality measure: cleanliness; attitude of staff towards patients; infection control; safety and security of patients and staff; reduction of waiting times; and availability of medicines at facilities.

11. Transparency: According to the NHIP, communities are to be empowered – this will lead to communities demanding transparency; and the Patients’ Rights Charter is to be robustly implemented. This includes the right to participate in development of health policies.

12. Public participation mechanisms: According to the NHIP, communities are to be empowered – this will allow active and meaningful participation; and the Patients’ Rights

Charter is to be robustly implemented. This includes the right to participate in development of health policies.

Section 4: Measuring health system quality

1. Bibliographic analysis of quality-related literature Daniel McKenna, Anna D. Gage, Emilia J. Ling

Supporting material for: Why measure health system quality?

To quantify the extent of literature on quality of care in LMICs, we conducted a bibliometric analysis in PubMed, Embase and Web of Science. The search terms on quality were developed from the HQSS framework and were combined with country search terms separated by income group. We collected all titles between 2000 and 2016; titles were not screened for relevance. The figure below shows the results, which reveal that healthcare quality research is overwhelmingly located in high income countries. In LMICs, the quality literature has grown tremendously over the period, on average 46% increase per year and a sharp uptick in 2010. Low middle-income countries, which include large countries such as India and Indonesia, publish the fewest articles on quality per year, with 1,077 articles in 2016 in comparison to 2,359 articles in low income and

3,179 articles in upper middle-income countries.

Quantity of publications on quality of care, 2000-2016

2. Clinical measurement methods

Naima T. Joseph

Supporting material for Section 2 (What quality of care are people receiving in LMICs today) and Section 4 (Measuring health system quality: What is – and is not – measured in LMIC health systems today)

The purpose of this table is to synthesize the application and limitations of methods for measuring the content of clinical care, including which clinical conditions and services are relevant for each tool as well as concerns around reliability and validity. The results in report Section 2 are based on these methods; we review the methods here to illustrate overall limitations of these findings as well as provide guidance on appropriate tools for those interested in measuring clinical care content.

Method Description Applications Limitations Level and

Frequency of Data

Collection

Implementation

Considerations

Administrative data

Analysis of claims submitted by physicians for payment

Conditions requiring frequent clinical visits (i.e. asthma, hypertension or diabetes), those requiring hospitalization (i.e.

appendicitis), pharmaceutically intensive conditions (i.e.

congestive heart failure)

Longitudinal quality assessment

Reliability and validity vary across conditions, e.g. high variation for heart failure, low for joint procedures.

Coding requirements for identification of conditions are highly specific, potentially failing to capture entire population served

Does not capture clinical decision making or interpersonal elements

Patient level

Routine

Ubiquitous and inexpensive to collect when data collection system is in place

Requires implementation of electronic medical system in addition to billing system to capture coding

Chart abstraction or review of medical record

Manual: chart abstractor reviews medical records to produce a data set of physician decisions

Electronic: data from electronic medical record fields related to physician

All clinical conditions

Data quality: incomplete and inaccurate content

Burden on clinicians for clinical record keeping and data capture

Patient level

Routine

Nearly ubiquitous when data collection system is in place and able to be obtained after the patient-provider encounter.

Electronic medical record technology provides clinical decision making are automatically extracted for analysis

Does not capture interpersonal elements of clinical encounter/process improved uniformity, legibility, communication

Costly to develop reliable data collection system to reduce poor or missing records, train and monitor data abstractors

Confidentiality of patient records

Clinical vignettes

Physician surveys using simulated patient cases

Broadly applicable to conditions with clear evidence indicating an appropriate choice for the patient, but not so obvious as to fail to solicit a variation in responses.

Well suited for algorithmic approaches (e.g., IMCI) and assessing appropriate diagnosis and treatment in curative care

Valid and comprehensive method for measuring provider knowledge;

correlation with performance during real clinical encounter subject to “know-do gap”

Does not capture interpersonal elements of clinical encounter/process

Provider level

Episodic

Inexpensive

Useful for cross-national comparisons of quality of care provided in different settings

Adaptation of clinical vignette to local setting may be required

Direct clinical observation

Trained observers record health workers performance during standard clinical encounters, typically using a checklist of expected actions

Non-emergent clinical conditions in inpatient and outpatient settings.

Limited acceptability for sensitive conditions such as HIV or genitourinary examination and challenging to use for rare conditions or populations. Well studied in mental health

Can include interpersonal elements of clinical encounter / process

Hawthorne effect (bias when participant changes his or her behavior due to being evaluated)

Variation across observers

Lack of control over the type of patients who present or duration of clinic visit

Provider level

Episodic

Quality assessment varies with time/number of clinical observations

Patient confidentiality/privacy

Can be expensive, requires trained clinical observers

Standardized patients

Trained actors role-play a specified patient case and record the clinical decisions.

The physician is blinded.

Conditions which have established medical protocols with clear triage, management, and treatment checklists and

Standardized patient recall and interrater reliability

Provider

Episodic

Can be expensive, requires recruiting and training standardized patients

Physicians who agree to which pose a low risk to standardized patients from invasive examinations, i.e.

unstable angina, asthma, diarrhea.

Not applicable for conditions with physical signs of presentation i.e. obstetric care.

Allows comparisons across different types of providers in different settings for the same conditions

Captures technical as well as interpersonal elements of the process participate may not be representative

Patient registry Organized system of uniformly collected data

(clinical and other) to evaluate specified outcomes for a population defined by a particular disease, condition, or exposure, and that serves a predetermined scientific, clinical, or policy purpose

Wide range of clinical conditions, chronic or self-limited, rare or common

Allows collection of patient reported, experience and outcome measurement

Data collected are uniform, purpose driven, and reflective of the clinical status of the patient

(i.e. derived from history, physical examination, laboratory data, and health outcomes)

Major purposes: describing the natural history of disease, determining clinical and/or cost-effectiveness, assessing safety or harm, or quality improvement

Potential for data entry error and missing data generalizability accrual

Routine

Patient level

Costs

Training of competent registry users and continued audit/data management

Patient interviews

Population Surveys-representative surveys that assess health risks, behaviors, and status of the population

Patient exit Interviews

Wide range of clinical conditions, chronic or self-limited, rare or common

Self-reported data allow for patient-reported experience and outcome data

Demographic information for

Recall errors and reporting bias

Information asymmetry:

patients do not know full content of health care

Cross sectional data

Superficial

Population

Episodic assessing equity

Sources4-17

3. Quality of routine data quality for priority indicators on maternal and neonatal health: a case study from Gombe State, Nigeria

Antoinette Bhattacharya, Tanya Marchant

Full manuscript forthcoming. Supporting material for Section 4: Data quality

Study setting

Gombe State, with a population of approximately 2.9 million, is located within northeastern

Nigeria where maternal and neonatal mortality is estimated to be considerably higher than the national estimates of 814 maternal deaths per 100,000 live births and 34 neonatal deaths per

1,000 live births. Gombe State has 615 facilities across 11 local government areas (LGAs), including 587 primary facilities and 28 referral facilities (secondary or higher).

Data sources

Electronic versions of a facility’s monthly summary report are available from DHIS2 for 615 facilities in Gombe. For all Gombe facilities included in DHIS2, data for July 2013-June 2017

(48 months) were downloaded at one time; this included 75 data elements related to the provision of antenatal care, labor and delivery, and postnatal/postpartum care.

Two facility-based surveys were conducted by an external research team to assess readiness to provide MNH services in Gombe. These surveys included extracting data from the antenatal and postnatal care register and the labor and delivery register, which reflect the delivery of services for the six calendar months immediately prior to the time of the survey: August 2016-January

2017 and January 2017-June 2017. The surveys took place in 115 facilities across Gombe: 97 primary health facilities and 18 referral facilities.

Data quality metrics assessed

The RHIS in Gombe State was assessed for completeness and internal consistency of data elements for 22 core indicators defined by the Ending Preventable Maternal Mortality (EPMM) and Every Newborn Action Plan (ENAP) initiatives. The metrics were assessed for the reference year July 2016-June 2017. Assessment followed the WHO data quality review for health facilities and metrics were measured for the state-level and disaggregated by primary and referral facilities, as follows:

• Completeness of data element: presence of non-zero value of data element for expected facility-months

• Consistency over time: ratio of value of data element for the reference year to the mean of preceding three years

• Outliers in reference year: presence of moderate and extreme outliers (+2 to 3 standard deviations and +3 standard deviations from the mean, respectively) of monthly values

• Consistency between related data elements: ratio of related data elements values that have a predictable relationship

Consistency between original facility registers and reported data: agreement between original facility register and facility monthly summary report

Results

Twelve of the 16 EPMM-ENAP indicators that can be assessed using facility-based information are currently collected; these comprise only one third of the maternal and newborn health data documented in facility registers. Basic coverage data for the EPMM-ENAP indicators were incomplete on over 40% of expected monthly reports, with lower completion rates for complicated data elements such as mortality or complications. Comparison of indicators over time and across sources identified substantial inconsistencies in over half of the data elements assessed, with over-reporting by primary facilities and under-reporting by referral facilities.

Application of thresholds for adequate consistency from global guidelines for data quality assessment18 suggest that at present relatively few of the EPMM-ENAP indicators are suitable to use in Gombe State.

4. Data quality challenges for health-system quality indicators

Mike English, Svetlana V. Doubova

Supporting material for Section 4: Data quality

Countries at all levels of development struggle with data systems and platforms, whether these are nascent or broadly implemented. We profile two additional such systems as sources for consideration of global and national strategies to address data quality challenges.

Kenya

A review of routine data from 272 hospitals in Kenya found that discharge data for maternal and pediatric patients were complete for only one-third of facilities, and that no hospital had a complete accounting of surgical discharges per year.19 Monitoring in-facility mortality for health-system sensitive conditions would provide critical insight on safety and quality, especially with basic adjustment factors such as illness severity available,20 but represents a major advance over current capacity in many countries.

Mexico

Mexico was the first Latin American country to introduce an EHR on a large-scale at the

Mexican Institute of Social Security (IMSS),21 which delivers healthcare to approximately 62 million people.22 The IMSS EHR is a uniform national system for primary care clinics; it is fairly comprehensive and easy to use. However, a high percentage of data are missing and up to 15% of clinical values have implausible values. Use of the EHR is largely confined to physicians; its use has not been clearly incorporated into the responsibilities of other health team members.

Patient identifiers differ between the EHR and the laboratory databases, limiting information exchange; health information is also not exchanged electronically among primary care clinics, emergency rooms, and hospital EHR databases.

Improving data quality and use

Action is required at multiple levels. Globally, investment in RHIS should include greater internal capacity to promote data quality, including automatic checks within and across indicators and distinguishing between missing and zero data elements. Actors at the national and sub-national levels should:

• Prioritize complete capture of data elements required for core indicators (e.g., EPMM-

ENAP) while reducing total number of data elements collected

• Strengthen capacity for numeracy throughout health system, including in reporting tallies and summary statistics

• Provide a centralized approach across the health system, building in interconnectivity for efficient information exchange

• Incorporate consistency checks for a core set of indicators to identify those within suggested WHO thresholds as adequate for interpretation and use

• Support national and sub-national use of priority indicators to encourage ownership and uptake of quality-relevant indicators

5. Geospatial mortality and access to quality health facilities in India

Anna J Dare, Patrycja Kolpak, Sze Hang Fu, Peter Rodriguez, Prabhat Jha. For the Million Death

Study collaborators

Full manuscript forthcoming. Supporting material for Section 4: Moving forward: three opportunities to measure better, Opportunity 1

This analysis examines the relationships between spatial access to health care at the district hospital (first-level hospital), quality of care and premature mortality (death at ages <70 years) from select time-critical conditions that are relevant to health-related global goals, namely maternal conditions, infections, select surgical conditions, and acute myocardial infarction.

Methods

Data sources

• The Million Death Study (MDS) is a nationally representative, population based, household survey of premature mortality in India, conducted in collaboration with the Registrar General of

India. The first phase of the study, reported here, took place between 2001-3. During this time, sample households were continuously monitored for vital events. Where a death occurred, verbal autopsy was undertaken using a validated tool including a half page local language narrative. 2 of

400 physicians then independently assigned ICD10 codes to classify causes of death. A full description of the MDS methods appears elsewhere.23,24

• The District Level Household and Facility Survey (DLHS-3), undertaken in 2007-08, provides details on the level and availability of resources for each of 565 district hospitals. The

Government of India designates community health centres and district hospitals to provide first level care for time-critical conditions (defined as conditions requiring definitive treatment within

24h of symptom onset, as recognised by a lay person, in order to prevent death).25 For surgical and comprehensive emergency obstetric conditions, district hospitals are the main facilities providing definitive surgical care, comprehensive emergency obstetric care (EmOC), as well as general intensive medical care as most community health centres have restricted capacity.26 We classified district hospitals as being either any district hospital (regardless of resource level) or well-resourced for each condition group (see Table 2 for definitions). A definition of comprehensive EmOC was adapted for the survey variables from that outlined by the UNFPA.

Classification of…