Prevalence of gestational diabetes mellitus based on various screening strategies in western Kenya: a prospective comparison of point of care diagnostic methods

Abstract

Background: Early diagnosis of gestational diabetes mellitus (GDM) is crucial to prevent short term delivery risks and long term effects such as cardiovascular and metabolic diseases in the mother and infant. Diagnosing GDM in Sub-Saharan Africa (SSA) however, remains sub-optimal due to associated logistical and cost barriers for resource-constrained populations. A cost-effective strategy to screen for GDM in such settings are therefore urgently required. We conducted this study to determine the prevalence of gestational diabetes mellitus (GDM) and assess utility of various GDM point of care (POC) screening strategies in a resource-constrained setting.

Methods: Eligible women aged ≥18 years, and between 24 and 32 weeks of a singleton pregnancy, prospectively underwent testing over two days. On day 1, a POC 1-h 50 g glucose challenge test (GCT) and a POC glycated hemoglobin (HbA1c) was assessed. On day 2, fasting blood glucose, 1-h and 2-h 75 g oral glucose tolerance test (OGTT) were determined using both venous and POC tests, along with a venous HbA1c. The International Association of Diabetes in Pregnancy Study Group (IADPSG) criteria was used to diagnose GDM. GDM prevalence was reported with 95% confidence interval (CI). Specificity, sensitivity, positive predictive value, and negative predictive value of the various POC testing strategies were determined using IADPSG testing as the standard reference.

Results: Six hundred-sixteen eligible women completed testing procedures. GDM was diagnosed in 18 women, a prevalence of 2.9% (95% CI, 1.57% - 4.23%). Compared to IADPSG testing, POC IADPSG had a sensitivity and specificity of 55.6% and 90.6% respectively while that of POC 1-h 50 g GCT (using a diagnostic cut-off of ≥7.2 mmol/L [129.6 mg/dL]) was 55.6% and 63.9%. All other POC tests assessed showed poor sensitivity.

Conclusions: POC screening strategies though feasible, showed poor sensitivity for GDM detection in our resource-constrained population of low GDM prevalence. Studies to identify sensitive and specific POC GDM screening strategies using adverse pregnancy outcomes as end points are required.

Trials registration: Clinical trials.gov : NCT02978807 , Registered 29 November 2016.

Keywords: Diagnosis; Gestational diabetes mellitus; Low middle income; Prevalence; Screening.

Fig. 1

Study schema and patient flow

Fig. 2

Bland Altman Plots of the Venous and POC Testing Strategies Utilized in this Study. a - Fasting Blood Glucose, Venous vs. POC, Correlation coefficient 0.56, p < 0.001. b - 1 h post 75 g glucose load, Venous vs. POC, correlation coeffiecient = 0.72, p < 0.001. c - 2 h post 75 g glucose load, Venous vs. POC, correlation coefficient = 0.6910, p < 0.001. d - HbA1c, Venous vs. POC, correlation coefficient = 0.49, p < 0.001