Skip to main content

Table 3 Summary of effects, affected fraction and baseline coverage of included interventions for stillbirths in the Lives Saved Tool

From: Modelling stillbirth mortality reduction with the Lives Saved Tool

Intervention

Effectiveness Estimatea

(95% CI)

Source of effectiveness data

Affected fractionb

Input data sources for calculating the affected fraction estimate

Source for baseline coverage

Micronutrient supplementation

RR 0.92

(0.86 to 0.99)

Haider et al. [30]

15 RCTs

98,808 women

All stillbirths

-

Zero as default

Malaria prevention with ITp or ITN

RR 0.67

(0.47 to 0.97)

Gamble et al. [45], Ishaque et al [18]

3 cRCTs

Stillbirths attributable to falciparum malaria

Proportion of pregnant women exposed to falciparum malaria [47]

Prevalence of placental malaria in those exposed to falciparum malaria (27.8%) [87]c

Risk of stillbirth with placental malaria:

OR=2.19 (1.49 – 3.22) [88]

Latest DHS/MICS estimate for “% pregnant women receiving 2+ doses of Sp/Fansidar during pregnancy” or “% pregnant women sleeping under an insecticide-treated bednet (ITN)” as a proxy if above NA

Balanced Energy supplementation

RR 0.60

(0.39 to 0.94)

Ota et al. [42]

5 RCTs

3,408 women

Stillbirths occurring in food-insecure households

% pop living <$1.90/day from World Bank [43] is used as a proxy

Zero as default

Syphilis Detection and Treatment

RR 0.18,

(0.10 – 0.33)

Blencowe et al. [50]

8 studies

3,931 births

Stillbirths attributable to syphilis

Prevalence data from Newman et al [89]

Risk of stillbirth with active syphilis:

RR=10.89 (95% CI 6.61 – 17.93) [90]

Defaults based upon ANC4+ coveraged:

If ANC4+ <40%- assume 20%*ANC4+

If ANC4+ 40-74%-assume 50%*ANC4+

IF ANC4+ 75-95%-assume 70%*ANC4+

If ANC4+>95%-assume 100%*ANC4+

Diabetes screening and management

10% reduction

(IQR -5 – 30% for APSB)

(IQR 3.5 – 25% for IPSB)

Syed et al. [17]

Expert opinion from 31 experts from 6 WHO regions

Stillbirths attributable to diabetes

Prevalence data from

International Diabetes Federation Atlas [91]

Risk of stillbirth with diabetes:

RR=3.38 [92]

Defaults based upon ANC4+ coveraged:

Assumed to be 5%*ANC4+

Detection and management of hypertensive disease of pregnancy (including treatment with magnesium sulphate)

20% reduction

(IQR -10 – 30% for APSB)

(IQR 10 – 40% for IPSB)

Jabeen et al. [16]

Expert opinion from 33 experts from 6 WHO regions and a range of disciplines

Stillbirths attributable to hypertensive disease of pregnancy

Prevalence data from Dolea et al 2003 [93]

Risk of stillbirth with hypertensive disease of pregnancy:

RR=2.1e

Defaults based upon ANC4+ coveraged:

Assumed to be 5%*ANC4+

Induction of labour for pregnancies lasting >41 weeks

RR 0.31

(0.12 – 0.88)

Gulmezoglu et al. and Hussain et al. [13, 57]

17 trials

7407 women

Stillbirths attributable to prolonged pregnancy

Prevalence: 7.5% of all pregnancies are estimated to progress post term if no policy to induce at post-term [94]

Risk: 1.8 [59]

Default assumption is that 100% of CEmOC deliveries have access to induction of labor for post-term pregnancies, if needed.

(Only available for births in CEmOC facilities)

Skilled attendance outside BEmOC or CEmOC facilities

RR 0.77

(0.69 – 0.85)

Yakoob et al. [19]

2 studies

All intrapartum stillbirths

NA

From DHS/ MICS and other nationally representative surveys

Childbirth care in BEmOC facility

45%

(IQR 30 – 70%)

Yakoob et al. [19]

Expert opinion from 27 experts from 6 WHO regions and a range of disciplines

All intrapartum stillbirths

NA

Defaults based upon facility delivery ratesf

If Facility delivery:

<30% assume 0% BEmOC/ 10% CEmOC

30 – 50% assume 30% BEmOC/ 20% CEmOC

50 – 95% assume 15% BEmOC/ 60% CEmOC

>95% assume 0% BEmOC/ 100% CEmOC

Childbirth care in CEmOC facility

75%

(IQR 50 – 87%)

Yakoob et al. [19]

Expert opinion from 27 experts from 6 WHO regions and a range of disciplines

All intrapartum stillbirths

NA

  1. a Further details of quality of evidence for estimate of effectiveness are presented in additional file 2
  2. b The affected fraction is the proportion of the time-specific mortality, here antepartum or intrapartum stillbirths, that is considered susceptible to that intervention
  3. c This is based on rates for primigravida in one study. The rate is 20.8% for women in second pregnancy, and 15.6% for higher order pregnancies, and hence may overestimate attributable fraction.
  4. d ANC4+ coverage from household survey data (DHS/MICS). Proportions attending ANC4+ receiving intervention assumptions based on opinion of two experts (Professors Zulfi Bhutta and Joy Lawn)
  5. e Reference not available for source used for approximation of risk. Studies from high income countries suggest aOR 1.3, 1.6 and 2.2 for pregnancy induced hypertension, pre-eclampsia and eclampsia respectively [22]. With similar orders of magnitude in low- and middle-income countries (LMIC) studies [23].
  6. f Facility delivery rates from nationally representative household survey data (DHS/MICS). Proportions receiving BEmOC/ CEmOC assumptions based on expert opinion