Population-based studies and analyses of maternal mortality in Malawi
| Study | Year | Method | Case definition | Location | Maternal deaths | MMR (95% CI)* |
|---|---|---|---|---|---|---|
| Chiphangwi et al91 | 1977–1979 | Indirect sisterhood† | Deaths of sisters who died “during pregnancy, childbirth or within 6 weeks of giving birth” | Southern region (Thyolo district) | 150 | 409 (349 to 480) |
| Malawi DHS 1992 (reanalysis)22 | 1977–1983 | Direct sisterhood | Deaths during pregnancy, childbirth and up to 6 weeks afterwards | Malawi (all regions) | 269 (??) | |
| Malawi DHS 1992 (reanalysis)22 | 1979–1985 | Direct sisterhood | Deaths during pregnancy, childbirth and up to 6 weeks afterwards | Malawi (all regions) | 42 | 408 (242 to 575) |
| McDermot et al92 | 1987–1989 | Prospective cohort | Deaths during pregnancy, childbirth and up to 6 weeks afterwards | Southern region (Mangochi district) | 15 | 398 (241 to 656) |
| Malawi DHS 1992 ‡3 | 1986–1992 | Direct sisterhood | Deaths during pregnancy, childbirth and up to 6 weeks afterwards | Malawi (all regions) | 71 | 620§ (410 to 830) |
| Malawi DHS 1992 (reanalysis)22 | 1986–1992 | Direct sisterhood | Deaths during pregnancy, childbirth and up to 6 weeks afterwards | Malawi (all regions) | 82 | 752 (497 to 1006) |
| Beltman et al93 | 1994–1996 | Indirect sisterhood† | Deaths of sisters who died “during pregnancy, childbirth or within 6 weeks of giving birth” | Southern region (Thyolo district) | 84 | 558 (260 to 820) |
| Malawi DHS 2000‡75 | 1994–2000 | Direct sisterhood | Deaths during pregnancy, childbirth and up to 2 months afterwards | Malawi (all regions) | 344 | 1120 (950 to 1288) |
| Malawi DHS 2004‡45 | 1998–2004 | Direct sisterhood | Deaths during pregnancy, childbirth and up to 2 months afterwards | Malawi (all regions) | 240 | 984 (804 to 1164) |
| MICS 2006‡23 | 2000–2006 | Direct sisterhood | Deaths during pregnancy, childbirth and up to 2 months afterwards | Northern region | 33 | 543 (325 to 761) |
| Central region | 190 | 678 (529 to 828) | ||||
| Southern region | 246 | 1029 (840 to 1217) | ||||
| Urban Malawi | 77 | 861 (492 to 1230) | ||||
| Rural Malawi | 392 | 802 (689 to 915) | ||||
| Malawi (all regions) | 469 | 807 (696 to 918) | ||||
| van den Broek et al94 | 2002 | Household survey | “The death of a woman associated with childbirth”. Timing not stated | Southern region (rural) | 9 | 413 (144 to 682) |
| Malawi DHS 2010‡1 | 2004–2010 | Direct sisterhood | Deaths during pregnancy, childbirth and up to 2 months afterwards | Malawi (all regions) | 331 | 675 (570 to 780) |
| MaiMwana (control arm)9 | 2006–2009 | Surveillance (prospective) | WHO ICD10 maternal death (see Background section) All 29 maternal deaths were verified by verbal autopsy | Central region (Mchinji district, rural) | 29 | 585 (407 to 838) |
| MaiKhanda (total¶)10 25 | 2007–2010 | Surveillance (prospective) | WHO ICD10 maternal death (see Background section), 51/102 (50%) verified by verbal autopsy, the rest verified by call-backs to community | Central region (Kasungu, Lilongwe and Salima districts, rural) | 102 | 299 (247 to 363) |
*Calculated as (100 000/MMR)×maternal deaths; for sisterhood studies calculated as ((maternal deaths/exposure years)/general fertility rate)×100 000.
†Should not be compared with direct sisterhood method as although the reference period is on average 12 years before the survey it includes recent deaths, which will bias the 12-year-old estimate upwards given that the MMR in Malawi increased in the 1990s.
‡The lower 95% CI of the MMR is calculated from the upper 95% CI of the GFR (less deaths/more births) and visa versa. Fertility is only reported for the whole sample in the MICS survey, therefore only the whole sample MMR (Malawi (all regions)) could be recalculated.
§In the 1992 DHS, the GFR used to calculate MMR is stated as 0.220 (table 11.4, p. 123). However in chapter 3 on fertility the total GFR is stated as 223/1000 women (or 0.223; table 3.1, p. 19) but this is for women aged 15–44 only. Using the raw data on number of women interviewed weighted by population of each cluster (district) so that the sample is representative of the whole of Malawi (table 2.8.2, p. 15) results in a total GFR/woman aged 15–49 years of 0.208 which is different to the 0.220 used to arrive at the MMR of 620 produced in the report.
¶Given the MaiKhanda trial showed no effect of either of the interventions on maternal mortality.
DHS, Demographic and Health Survey; GFR, general fertility rate; MICS, Multiple Indicator Cluster Survey; MMR, maternal mortality ratio.