Detection and risk stratification of women at high risk of preterm birth in rural communities near Nagpur, India

Abstract

Abstract Background Presently, preterm birth is globally the leading cause of neonatal mortality. Prompt community based identification of women at high risk for preterm births (HRPB) can either help to avert preterm births or avail effective interventions to reduce neonatal mortality due to preterm births. We evaluated the performance of a package to train community workers to detect the presence of signs or symptoms of HRPB. Methods Pregnant women enrolled in the intervention arm of a cluster randomized trial of Antenatal Corticosteroids (ACT Trial) conducted at Nagpur, India were informed about 4 directly observable signs and symptoms of preterm labor. Community health workers actively monitored these women from 24 to 36 weeks of gestation for these signs or symptoms. If they were present (HRPB positive) the identified women were brought to government health facilities for assessment and management. HRPB positive could also be determined by the provider if the woman presented directly to the facility. Risk stratification was based on the number of signs or symptoms present. The outcome of preterm birth was based on the clinical assessment of gestational age < 37 weeks at delivery or a birth weight of <2000 g. Results Between July 1, 2012 and 30 November, 2013, 686 of 7050 (9.7%) pregnant women studied, delivered preterm. 732 (10.4%) women were HRPB positive, of whom 333 (45.5%) delivered preterm. Of the remaining 6318(89.6%) HRPB negative women 353 (5.6%) delivered preterm. The likelihood ratio (LR) of a preterm birth in the HRPB positives was 8.14 (95% confidence interval 7.16–9.26). The LR of a preterm birth increased in women who had more signs or symptoms of HRBP (p < 0.00001). More signs or symptoms of HRPB were also associated with a shorter time to delivery, lower birth weight and higher rates of stillbirths, neonatal deaths and postnatal complications. Addition of risk stratification improved the prediction of preterm delivery (Integrated Discrimination Improvement 17% (95% CI 15–19%)). Conclusions The package for detection of signs and symptoms of HRPB is feasible, promising and likely to improve management of preterm labor. Trial registration NCT01073475 on February 21, 2010 and NCT01084096 on March 9, 2010.

Keywords

• Preterm birth
• Risk stratification
• Diagnostic accuracy
• Community workers