Validation of a screening tool to identify the nutritionally at-risk pregnancy

To assess the efficacy of a screening tool to be used by nurses to determine which economically disadvantaged pregnant women are most likely to benefit from a dietitian's intensive intervention.

The 26-variable screening tool was used by 53 nurses in 17 Montreal primary care facilities at first contact with 300 subjects who were eligible for prenatal programs for low-income women. Among these, 259 subjects were subsequently interviewed by a Montreal Diet Dispensary (MDD) dietitian using the Higgins Method, established as the gold standard for the detection of a nutritionally at-risk pregnancy. Comparisons of the results obtained by dietitians using the Higgins Method with scores obtained by nurses using the screening tool to determine risk levels and individual variables were analyzed using the Pearson correlation coefficient, receiver operating characteristic (ROC) curves, and predictive indices such as sensitivity, specificity, and predictive values.

Because of low agreement between the findings of MDD dietitians and nurses using the screening tool, 10 variables were excluded. The resulting 16-variable tool showed a correlation coefficient of 0.70. The ROC cut-off for this simplified tool was 8, meaning that a woman would be considered at nutritional risk if her score was 8 or more. This score optimized sensitivity (85%) with rather low specificity (50%), but retained a positive predictive value of 87% and a negative predictive value of 55%.

Revisions of the screening tool to identify nutritionally at-risk pregnant women led to the development of a simplified screening tool with an acceptable measure of nutritional risk in pregnancy. Hence, the use of this tool by any health professional will identify the majority of the nutritionally at-risk pregnant women most likely to benefit from a dietitian's intensive intervention.