The Inverse Research Law – the least evidence from settings where need is greatest. An overview of reviews of interventions to reduce the risk of preterm births and their generalisability to low- and middle-income settings




Oral session: Global health and equity (3)


Thursday 24 October 2019 - 16:00 to 17:30


All authors in correct order:

Campbell F1, Jayasooriya SM2, Soma-Pillay P3, Balen J2, Gillespie B4, Mitchell C4, Anumba D4
1 School of Health and Related Research (ScHARR), UK
3 University of Pretoria, South Africa
4 University of Sheffield, UK
Presenting author and contact person

Presenting author:

Fiona Campbell

Contact person:

Abstract text
Background: preterm birth (PTB at < 37 weeks' gestation) is the leading cause of death worldwide in children under five years. Approximately 10.6% of live births (14.8 million infants) each year are premature, with 81% of preterm births in Asia and sub-Saharan Africa. Developing clinical guidelines to prevent PTB in low- and middle-income (LMIC) settings, where the preterm birth rates are highest, requires drawing on existing evidence that is largely reliant on research data gathered in high-income settings. We sought to explore the extent to which systematic review evidence considers generalizability of findings to LMIC settings, and the settings in which primary data are gathered.

- identify existing systematic reviews that have explored interventions to reduce the risk of preterm birth;
- describe the proportion of primary studies included in the reviews undertaken in LMICs;
- describe the effectiveness of the interventions;
- explore the methods used in systematic reviews to consider generalizability of findings across diverse contexts.

Methods: a mapping review described the evidence across a range of interventions, implemented during pregnancy to prevent spontaneous PTB. Systematic searches of electronic databases identified 188 systematic reviews for inclusion. We abstracted data and subjected findings to a narrative synthesis. If reported, we classified the countries where the primary studies were undertaken using World Bank categories.

Results: we included 188 systematic reviews addressing the following interventions: healthcare delivery and health systems (n = 18), predicting risk of PTB (n = 18), lifestyle changes (n = 15), nutritional interventions (n = 36) periodontal disease treatment (n = 23), influenza vaccination (n = 2), infection screening and treatment (18 reviews), cerclage (n = 19), corticosteroids (n = 10), cervical pessary (n = 6) and progesterone (n = 23).

We found a lack of consistency between reviews on the effects on PTB and an absence of strong evidence supporting existing interventions in reducing PTB. Twenty-eight per cent of reviews did not report the country where the primary studies were undertaken suggesting a poor understanding of the influence of context on PTB outcomes. The proportion of studies undertaken in LMICs varied across intervention types. We found that there is very little data from low-income countries, with most LMIC data derived from lower middle- and middle-income countries.

Conclusions: this review describes the current evidence published over the last decade, on interventions to reduce the risk of PTB. We suggest an inverse care pattern of evidence with the least amount of evaluation occurring in those contexts where PTB rates are highest. The extent to which findings can be generalized to different contexts has been largely unexplored in the existing evidence reviews. The lack of data to support effective interventions in LMIC contexts has implications for the development of contextually relevant clinical guidelines.