Table 3 Random-effects meta-analysis of pooled prevalence of incompleteness and sensitivity of maternal deaths recording stratified by individual covariates (excluding two studies for high risk of bias score)

Overall sensitivity

4

61% (37–82)

I^2 = 92.7

H^2 = 13.7

(P < 0.001)

Total incompleteness

16

34% (28–48)

I^2 = 91.2%

H^2 = 12.3

(P < 0.001)

Incompleteness by cause of maternal death

 Direct

6

22% (4 – 48)

I^2 = 96.1%

H^2 = 25.3

(P = 0.006)

 Indirect

6

42% (10–76)

I^2 = 87.9%

H^2 = 8.3

(P < 0.001)

Incompleteness by place of death

 Home

3

75% (20–100)

I^2 = 96.4%

H^2 = 27.6

(P < 0.001)

 Facility

3

27% (6–58)

I^2 = 96.0%

H^2 = 24.8

(P < 0.001)

Incompleteness by time of death relative to pregnancy

 During delivery or within 24 h

3

25 (16–33)

I^2 = 0%

H^2 = 1

0.0 (P = 0.738)

 During pregnancy

3

52 (39–64)

I^2 = 0.0

H^2 = 1

0.0 (P = 0.422)

 More than 24 h post-partum

3

52 (40–63)

I^2 = 36.4%

H^2 = 1.7

(P = 0.239)