A systematic review of inequalities in the use of maternal health care in developing countries: examining the scale of the problem and the importance of context
Lale Say, Rosalind Raine
Volume 85, Number 10, October 2007, 812-819
Annex 1. Characteristics and quality assessment of studies
| Study | Setting, year | Methods | Participants | Comparison groups | Outcome measures | Qualitya |
|---|---|---|---|---|---|---|
| Addai (1998) |
Ghana(national),1993 |
Cross-sectional household survey – retrospective analysis of DHS; multistage cluster sampling |
Women delivered within 5 years of the survey |
Urban-rural,age |
Most recent delivery in a medical setting |
– |
| Anson (2004) |
China(province),1996–99 |
Cross-sectional household survey – retrospective analysis from a larger study; stratified sampling of 288 villages |
Women delivered (live birth) at least once |
Age,economic (per-capita family income) |
Most recent delivery in a medical setting |
– |
| Barbhuiya (2001) |
Bangladesh(sub-district), 1997 |
Prospective cross-sectional household survey; all eligible women from randomly selected villages |
Women pregnant (last trimester) or delivered during past 3 months |
Economic (family income) |
Most recent delivery in a medical setting |
– |
| Bhatia (1995) |
India(Karnataka, sub-district), 1993 |
Cross-sectional household survey – analysis of data from a child survival study; all eligible women living in town and villages with more than 500 inhabitants |
Women younger than 35 years with at least one child younger than 5 years (for antenatal care outcome: women attended antenatal care - 90% of sample) |
Urban-rural, age, economic (financial value of household assets) |
Antenatal care during the first trimester for most recent delivery; most recent delivery in a medical setting |
+ |
| Bloom (2001) |
India(Varanasi, Uttar Pradesh),1995–96 |
Prospective cross-sectional household survey; two-stage cluster sampling; urban setting (poor/middle income households within 15 minutes’ walking distance from health facility) |
Women delivered within 3 years of the survey, and either Muslim or Hindu (97% of population); youngest eligible woman from household |
Economic (type of building walls: cement [high] or other material [low]) |
Most recent delivery by a skilled attendant (with formal medical training, regardless of delivery site) |
+ |
| Bolam (1998) |
Nepal (Kathmandu),1994–95 |
Case-control study;all eligible women in selected urban and peri-urban areas |
Pregnant women in urban and peri-urban areas |
Urban/peri-urban |
Index delivery in a medical setting |
– |
| Burgard (2004) |
South Africa (national), 1998;Brazil (national), 1996 |
Cross-sectional household survey – retrospective analysis of DHS; multistage cluster sampling |
Deliveries within 5 years of survey |
Age, economic (household assets) |
Antenatal care during the first trimester |
+ |
| Celik (2002) |
Turkey(national), 1993 |
Cross-sectional household survey – retrospective analysis of DHS; multistage cluster sampling |
Ever-married women delivered within 3 years of survey |
Urban-rural, age |
Most recent delivery by a skilled attendant |
– |
| Duong (2004) |
Viet Nam(district),2000 |
Case-control study;five stratified areas from 41 communes; cases (hospital delivery) and controls (home delivery) randomly selected from immunization and antenatal care records |
Women delivered at a hospital setting or at home within 3 months of survey |
Age, economic (income) |
Index delivery in a medical setting |
– |
| Eggleston (2000) |
Ecuador(national), 1994 |
Cross-sectional household survey – retrospective analysis of DHS; multistage cluster sampling |
Women who delivered within 2 years of survey and who had at least one antenatal visit (76% had antenatal care) |
Urban-rural, age, economic (household assets) |
Antenatal care during the first trimester for most recent delivery |
– |
| Falkingham (2003) |
Tajikistan(national),1999 |
Cross-sectional household survey – retrospective analysis of living standards survey; two-stage cluster sampling |
Ever-married women who delivered at least once |
Urban-rural residence,economic (per capita household expenditure) |
Most recent delivery by a skilled attendant; most recent delivery in a medical setting |
+ |
| Gertler (1995) |
Jamaica(national); 1989 |
Cross-sectional household survey – retrospective analysis of Jamaican Survey of Living Conditions; multistage cluster sampling |
Women aged 15–45 years who had experienced a pregnancy lasting longer than 7 months within 5 years of survey (one randomly chosen woman from each household) |
Age, economic (per capita household expenditure) |
Antenatal care during the first trimester for most recent pregnancy; most recent delivery in a medical setting |
++ |
| Glei (1999) |
Guatemala (four departments; rural communities),1995 |
Cross-sectional household survey – Guatemalan Survey of Family Health; 15 communities in each department selected (probability proportional to size), 50 women in each selected community were interviewed |
Births to women aged 18–35 years within 5 years of survey (up to two births per women) |
Age, economic (household food consumption) |
Births in a medical setting |
+ |
| Hodgkin (1996) |
Kenya(district),1989 |
Prospective cross-sectional household survey; randomly selected 60 villages – nine households per village |
Women delivered within 1 year of survey |
Economic (income) |
Most recent delivery in a medical setting |
– |
| Hotchkiss (2001) |
Nepal(national),2001 |
Cross-sectional household survey – retrospective analysis of Nepal Living Standards Survey; two stage cluster sampling |
Ever-married women delivered within 3 years of survey |
Urban-rural, age, economic (per capita household expenditure) |
Most recent delivery by a skilled attendant (an auxillary nurse, midwife, maternal/child health worker, nurse or a doctor) |
+ |
| Hotchkiss (2003) |
Morocco(national),1995 |
Cross-sectional household survey – retrospective analysis of DHS; multistage cluster sampling; linked service availability module |
Women who gave birth at least once within 5 years of survey |
Urban-rural, age, economic (household assets and characteristics) |
Most recent delivery in a medical setting |
+ |
| Kavitha (1997) |
India(two villages),1995 |
Prospective cross-sectional survey; all eligible women |
Births within the 4 years of survey (liveborn and living at the time of survey) |
Economic (household income) |
Most recent delivery in a medical setting |
– |
| Letamo (2003) |
Botswana(national),1996 |
Cross-sectional household survey – retrospective analysis of BotswanaFamily Health Survey (DHS); multistage cluster sampling |
Women had at least one pregnancy within 5 years of survey |
Age, economic (household assets and characteristics) |
Most recent delivery not by a skilled attendant, most recent delivery not in a medical setting |
– |
| Li (2004) |
China(district), 1994 |
Cross-sectional household survey – retrospective analysis of an infant health survey; administrative villages were selected by stratifying by township, per capita income, and population size; selected from list of infants |
Women delivered within 3 years of survey |
Economic (household income) |
Most recent delivery by a skilled attendant |
+ |
| Magadi (2000) |
Kenya(national),1993 |
Cross-sectional household survey – retrospective analysis of DHS; multistage cluster sampling |
Births within 5 years of the survey |
Urban-rural residence, economic (household assets) |
Most recent delivery not in a medical setting |
– |
| McCaw Binns (1995) |
Jamaica(national),1986–87 |
Prospective cohort study |
All women delivered during a 2-month period and attended antenatal care (96% had antenatal care) |
Urban-rural, age, economic (food expenditure) |
Antenatal care during first trimester for index delivery |
++ |
| Mekonnen (2003) |
Ethiopia(national),2000 |
Cross-sectional household survey – retrospective analysis of DHS; multistage cluster sampling |
Women delivered at least once within 5 years of survey |
Urban-rural, age |
Most recent delivery by a skilled attendant |
– |
| Navaneetham (2002) |
India(four states),1992–93 |
Cross-sectional household survey – retrospective analysis of National Family and Health Survey; multistage cluster sampling |
Ever-married women who delivered (live birth) within 4 years of survey |
Age, urban-rural, economic (living standard index from household assets and environmental conditions) |
Antenatal care during first trimester for most recent delivery; most recent delivery in a medical setting; most recent delivery by a skilled attendant |
+ |
| Paul (2002) |
Bangladesh(rural area),1995–97 |
Prospective cross-sectional survey; all villages in selected district |
Women delivered within 2 years of survey |
Age, economic (landholding size) |
Most recent delivery by a skilled attendant (includes trained traditional attendants) |
++ |
| Phoxay (2001) |
Lao People’s Democratic Republic(district),1999 |
Prospective cross-sectional survey; randomly selected villages; 10–20% of reproductive-age women per village |
Women who have children younger than 5 years and who lived in the area during most recent pregnancy and birth |
Economic (income) |
Most recent delivery by a skilled attendant |
– |
| Stupp (1994) |
Belize(national), 1991 |
Retrospective cross-sectional household survey; two-stage cluster random sample |
Married or in union women 15–44 years age who delivered within 5 years of survey (one woman from each household) |
Urban-rural |
Most recent delivery in a medical setting |
– |
| Toan (1996) |
Viet Nam(district),1994 |
Prospective cross-sectional household survey; multistage cluster sampling |
Women delivered during the 5 years before survey |
Age, economic (quality of housing: roof) |
Most recent delivery in a medical setting |
– |
| Tsui (2002) |
Paraguay (four departments), 1998;Uganda (nine of 39 districts), 1997;United Republic of Tanzania (national – except Zanzibar), 1996;India (state – Uttar Pradesh), 1995 |
Retrospective analysis of cross-sectional multistage cluster surveys linked with probability sample of health facilities |
Women delivered within 5 years of surveys in Paraguay and United Republic of Tanzania and within 3 years of surveys in Uganda and India |
Urban-rural, age, economic (household assets) |
Most recent delivery in a medical setting; most recent delivery by a skilled attendant; contraceptive use |
– |
| Van der Heuvel (1999) |
Zimbabwe(district),1996 |
Prospective cross-sectional household survey; two-stage cluster sampling |
Women delivered within 3 years of survey and lived in catchment area of the hospital during most recent delivery and at time of survey |
Economic (household assets and characteristics) |
Most recent delivery in a medical setting |
– |
| Wagle (2004) | Nepal, 2001 | Prospective cross-sectional household survey | Women delivered within 45 days of survey | Age, economic (household assets) | Index delivery in a medical setting | – |
DHS, demographic and health survey.a Based on the SIGN methodological quality checklists. Code ++ (high-quality case-control, cohort or cross-sectional studies with a very low risk of confounding, bias or chance and a high probability that the association is causal), + (well-conducted case-control, cohort or cross-sectional studies with a low risk of confounding, bias or chance and a moderate probability that the association is causal), – (case-control, cohort or cross-sectional studies with a high risk of confounding, bias or chance and a significant risk that the association is not causal).