Validation of a novel mobile phone application for type 2 diabetes screening following gestational diabetes mellitus

Abstract

Background: We sought to determine if using fasting blood glucose (FBG) through text-based care is an effective screening tool for type 2 diabetes in the postpartum period compared to in-person, 2-hour oral glucose tolerance testing (2hr OGTT). Methods: This was a single-center interventional study that included individuals diagnosed with gestational diabetes. Patients were enrolled in standard, office-based 2hr OGTT in combination with text-based remote diabetes screening. Study participants were instructed to record FBG for 3 consecutive days using a mobile application. We assessed agreement with 2hr OGTT using sensitivity, specificity, positive and negative predictive value with exact binomial 95% confidence intervals. Results: A total of 446 individuals diagnosed with gestational diabetes met inclusion criteria, 239 of which were enrolled in standard office-based screening and 207 were enrolled in dual screening using standard 2hr OGTT testing combined with text-based remote FBG screening. A FBG value less than 100 mg/dL had 100% sensitivity (86-100%), 86% specificity (77-93%) with a 100% (94-100%) negative predictive value and 71% (54-85%) positive predictive value. Follow-up was significantly higher among individuals enrolled in remote text-based screening compared to standard in-office screening (48% vs. 25%, respectively; P<0.001). Conclusions: Text-based screening may be a feasible alternative to in-office screening. A mobile-based system using FBG successfully screened all patients with type 2 diabetes in the postpartum period with 100% sensitivity and negative predictive value. Remote telehealth screening significantly increased follow-up with type 2 diabetes screening.