Diagnostic accuracy of neonatal foot length to identify preterm and low birthweight infants: a systematic review and meta-analysis

Introduction: Eighty percent of neonatal deaths occur among babies born preterm and/or small for gestational age (SGA). In sub-Saharan Africa and South Asia, approximately 40% of births occur outside of health facilities, and gestational age (GA) and birth weight are commonly unknown. Foot length (FL) has been proposed as a simple, surrogate measurement to identify and triage small babies born in the community. We conducted a systematic review and meta-analysis of the diagnostic accuracy of newborn FL to classify preterm and low birthweight infants.

Methods: PubMed, EMBASE, Cochrane, Web of Science, POPLINE and WHO Global Health Library databases were searched. Studies of live-born infants that compared FL with GA and/or birth weight were included. Data on diagnostic accuracy were summarised, described, and pooled, as appropriate.

Results: Six hundred and two studies were identified and 41 included. Techniques for measuring FL included use of a firm plastic ruler, callipers, footprint or a measuring board. Twelve studies assessed the diagnostic accuracy of FL to identify preterm births; however, data were not pooled given heterogeneity and low quality of GA. 19 studies used FL to identify low birthweight infants (<2500 g, <2000 g). Among studies in Asia (n=3), FL <7.7 cm had pooled sensitivity and specificity of 87.6% (95% CI 61.1% to 99.0%) and 70.9% (95% CI 23.5% to 95.1%), respectively, to identify <2500 g infants. FL <7.3 cm had 82.1% (95% CI 63.7% to 92.2%) sensitivity and 82.1% (95% CI 59.2% to 90.8%) specificity for identifying <2000 g infants (n=3). In the African studies (n=3), FL <7.9 cm had pooled sensitivity and specificity of 92.0% (95% CI 85.6% to 95.7%) and 71.9% (95% CI 44.5% to 89.1%), respectively, to identify <2500 g neonates.

Conclusions: FL is a simple proxy measure that can identify babies of low birthweight with high sensitivity, though somewhat lower specificity. Additional research is needed to determine the validity of FL to identify preterm infants, and understand the programmatic impact of screening on healthcare seeking and outcomes.

Prospero registration number: CRD42015020499.