Keywords
How to Cite
Abstract
Background. Gestational diabetes mellitus (GDM) is becoming more common in low- and middle-income countries, including Pakistan. A simple, affordable and easy-to-use risk prediction model for early identification of high-risk women is therefore needed. This study set out to develop and validate a non-invasive risk prediction model, the Non-INvaSive Pregnancy RIsk ScoRE (INSPIRE), which uses risk factors already reported in the literature to identify Pakistani pregnant women at high risk of GDM.
Methods. In a cross-sectional study, 550 pregnant women attending antenatal clinics at one tertiary and two secondary hospitals in islamabad were enrolled between weeks 29 and 33 of pregnancy. The data were randomly split into a derivation set (422 women, 80%) and a validation set (94 women, 20%). Researchers interviewed the women about sociodemographic details and family history of diabetes, measured mid-upper arm circumference (MUAC), and took obstetric history and oral glucose tolerance test (OGTT) results from hospital records. Multivariable logistic regression on the derivation set produced the predictor weights. Calibration was tested with Pearson's chi-square goodness-of-fit test, and discrimination was measured with the area under the curve (AUC), both in the validation set.
Results. GDM affected 32 % of the women (136 of 550). The final score uses six predictors: maternal age, MUAC, family history of diabetes, previous GDM, a previous bad obstetric outcome, and previous macrosomia. In the validation set the score was well calibrated (chi-square 29.55, p = 0.08) and showed acceptable discrimination (AUC 0.721, 95% CI 0.61 to 0.83), with 74.1% sensitivity and 59.4% specificity.
Conclusion. INSPIRE is a non-invasive risk prediction model that allows early identification of Pakistani pregnant women at high risk of GDM and separates them from those at low risk, which can spare many women an unnecessary OGTT.