Low Accuracy for Prediabetes Screening Tests

This study surprised me a bit, but after reading some of my previous blogs, I realized that the study does have value and is probably correct.

The accuracy of the tests used to detect prediabetes in screening programs is low. The diagnostic accuracy of the current tests used to detect prediabetes in screening programs is low, according to a study published in BMJ. The results showed that the fasting glucose screening test is specific but not sensitive, and the HbA1c test is neither sensitive nor specific.

Eleanor Barry, MBBS, BSc, MRCP, MRCGP, from the Nuffield Department of Primary Care Health Sciences at the University of Oxford, and colleagues assessed the diagnostic accuracy of screening tests for prediabetes and the efficacy of lifestyle or metformin interventions in preventing the onset of type 2 diabetes.

The researchers performed one meta-analysis to summarize the accuracy of screening tests for identification of prediabetes, with the oral glucose tolerance test as the standard, and they conducted an additional meta-analysis that assessed relative risk of progression to type 2 diabetes after lifestyle intervention or treatment with metformin.

The investigators included 49 studies of screening tests and 50 intervention trials in their final analysis. They examined empirical studies examining the accuracy of tests for the identification of prediabetes, randomized trials, and interventional studies.

“As the prevalence of type 2 diabetes rises inexorably in high, middle, and low income countries alike, controversy continues to surround the questions of who is ‘at risk' and what preventive interventions to offer them,” the study authors wrote. “A screen and treat policy will be effective only if a test exists that correctly identifies those at high risk (sensitivity) while also excluding those at low risk (specificity); and an intervention exists that is acceptable to, and also efficacious in, those at high risk.”

The HbA1c screening test had a mean sensitivity of 0.49 and specificity of 0.79 for the identification of prediabetes, although different studies used different cut-off values. In addition, fasting plasma glucose had a mean sensitivity of 0.25 and a specificity of 0.94.

Lifestyle interventions were associated with a 36% reduction in relative risk of type 2 diabetes during the course of 6 months to 6 years, which attenuated to 20% at follow-up in the period after the trials.

The authors note that future studies should focus on the pragmatic real world effectiveness and cost effectiveness of interventions for prediabetes that have already shown to be efficacious in trials. Further research should also be conducted to evaluate population level or health system interventions.