Community-based intervention packages for reducing maternal and neonatal morbidity and mortality and improving neonatal outcomes

Zohra S Lassi, Zulfiqar A Bhutta, Zohra S Lassi, Zulfiqar A Bhutta

Abstract

Background: While maternal, infant and under-five child mortality rates in developing countries have declined significantly in the past two to three decades, newborn mortality rates have reduced much more slowly. While it is recognised that almost half of the newborn deaths can be prevented by scaling up evidence-based available interventions (such as tetanus toxoid immunisation to mothers, clean and skilled care at delivery, newborn resuscitation, exclusive breastfeeding, clean umbilical cord care, and/or management of infections in newborns), many require facility-based and outreach services. It has also been stated that a significant proportion of these mortalities and morbidities could also be potentially addressed by developing community-based packaged interventions which should also be supplemented by developing and strengthening linkages with the local health systems. Some of the recent community-based studies of interventions targeting women of reproductive age have shown variable impacts on maternal outcomes and hence it is uncertain if these strategies have consistent benefit across the continuum of maternal and newborn care.

Objectives: To assess the effectiveness of community-based intervention packages in reducing maternal and neonatal morbidity and mortality; and improving neonatal outcomes.

Search methods: We searched the Cochrane Pregnancy and Childbirth Group's Trials Register (31 May 2014), World Bank's JOLIS (25 May 2014), BLDS at IDS and IDEAS database of unpublished working papers (25 May 2014), Google and Google Scholar (25 May 2014).

Selection criteria: All prospective randomised, cluster-randomised and quasi-randomised trials evaluating the effectiveness of community-based intervention packages in reducing maternal and neonatal mortality and morbidities, and improving neonatal outcomes.

Data collection and analysis: Two review authors independently assessed trials for inclusion, assessed trial quality and extracted the data. Data were checked for accuracy.

Main results: The review included 26 cluster-randomised/quasi-randomised trials, covering a wide range of interventional packages, including two subsets from three trials. Assessment of risk of bias in these studies suggests concerns regarding insufficient information on sequence generation and regarding failure to adequately address incomplete outcome data, particularly from randomised controlled trials. We incorporated data from these trials using generic inverse variance method in which logarithms of risk ratio (RR) estimates were used along with the standard error of the logarithms of RR estimates.Our review showed a possible effect in terms of a reduction in maternal mortality (RR 0.80; 95% confidence interval (CI) 0.64 to 1.00, random-effects (11 studies, n = 167,311; random-effects, Tau² = 0.03, I² 20%). However, significant reduction was observed in maternal morbidity (average RR 0.75; 95% CI 0.61 to 0.92; four studies, n = 138,290; random-effects, Tau² = 0.02, I² = 28%); neonatal mortality (average RR 0.75; 95% CI 0.67 to 0.83; 21 studies, n = 302,646; random-effects, Tau² = 0.06, I² = 85%) including both early and late mortality; stillbirths (average RR 0.81; 95% CI 0.73 to 0.91; 15 studies, n = 201,181; random-effects, Tau² = 0.03, I² = 66%); and perinatal mortality (average RR 0.78; 95% CI 0.70 to 0.86; 17 studies, n = 282,327; random-effects Tau² = 0.04, I² = 88%) as a consequence of implementation of community-based interventional care packages.Community-based intervention packages also increased the uptake of tetanus immunisation by 5% (average RR 1.05; 95% CI 1.02 to 1.09; seven studies, n = 71,622; random-effects Tau² = 0.00, I² = 52%); use of clean delivery kits by 82% (average RR 1.82; 95% CI 1.10 to 3.02; four studies, n = 54,254; random-effects, Tau² = 0.23, I² = 90%); rates of institutional deliveries by 20% (average RR 1.20; 95% CI 1.04 to 1.39; 14 studies, n = 147,890; random-effects, Tau² = 0.05, I² = 80%); rates of early breastfeeding by 93% (average RR 1.93; 95% CI 1.55 to 2.39; 11 studies, n = 72,464; random-effects, Tau² = 0.14, I² = 98%), and healthcare seeking for neonatal morbidities by 42% (average RR 1.42; 95% CI 1.14 to 1.77, nine studies, n = 66,935, random-effects, Tau² = 0.09, I² = 92%). The review also showed a possible effect on increasing the uptake of iron/folic acid supplementation during pregnancy (average RR 1.47; 95% CI 0.99 to 2.17; six studies, n = 71,622; random-effects, Tau² = 0.26; I² = 99%).It has no impact on improving referrals for maternal morbidities, healthcare seeking for maternal morbidities, iron/folate supplementation, attendance of skilled birth attendance on delivery, and other neonatal care-related outcomes. We did not find studies that reported the impact of community-based intervention package on improving exclusive breastfeeding rates at six months of age. We assessed our primary outcomes for publication bias and observed slight asymmetry on the funnel plot for maternal mortality.

Authors' conclusions: Our review offers encouraging evidence that community-based intervention packages reduce morbidity for women, mortality and morbidity for babies, and improves care-related outcomes particularly in low- and middle-income countries. It has highlighted the value of integrating maternal and newborn care in community settings through a range of interventions, which can be packaged effectively for delivery through a range of community health workers and health promotion groups. While the importance of skilled delivery and facility-based services for maternal and newborn care cannot be denied, there is sufficient evidence to scale up community-based care through packages which can be delivered by a range of community-based workers.

Conflict of interest statement

Dr Zulfiqar A Bhutta is the principal investigator of two included studies evaluating community care perinatal care package in Pakistan (Bhutta 2008; Bhutta 2011) but he was not involved in assessing these trials for inclusion in this review, assessing trial quality, or data extraction. These tasks were carried out by other members of the review team who were not involved with the original studies (ZSL and BAH).