Background

In Ghana, 32% of deliveries take place outside a health facility, and birth weight is not measured. Low birth weight (LBW) newborns who are at increased risk of death and disability, are not identified; 13%–14% of newborns in Ghana are LBW. We aimed at determining whether alternative anthropometrics could be used to identify LBW newborns when weighing scales are not available to measure birth weight.

Methods

We studied 973 mother and newborn pairs at the Komfo Anokye Teaching and the Suntreso Government hospitals between November 2011 and October 2012. We used standard techniques to record anthropometric measurements of newborns within 24 hours of birth; low birth weight was defined as birth weight <2.5kg. Pearson's correlation coefficient and the area under the curve were used to determine the best predictors of low birth weight. The sensitivity, specificity and predictive values were reported with 95% confidence intervals at generated cut-off values.

Results

One-fifth (21.7%) of newborns weighed less than 2.5 kg. Among LBW newborns, the following measurements had the highest correlations with birth weight: chest circumference (r = 0.69), mid-upper arm circumference (r = 0.68) and calf circumference (r = 0.66); the areas under the curves of these three measurements demonstrated the highest accuracy in determining LBW newborns. Chest, mid-upper arm and calf circumferences at cut-off values of ≤29.8 cm, ≤9.4 cm and ≤9.5 cm respectively, had the best combination of maximum sensitivity, specificity and predictive values for identifying newborns with LBW.

Conclusions

Anthropometric measurements, such as the chest circumference, mid-upper arm circumference and calf circumference, offer an opportunity for the identification of and subsequent support for LBW newborns in settings in Ghana, where birth weights are not measured by standardized weighing scales.