Abstract
BACKGROUND: When monitoring long-term conditions, both over- and undertesting can risk patient harm and increase healthcare costs. AIM: To evaluate the evidence base for type 2 diabetes mellitus (T2DM) monitoring tests and develop methods for creating evidence-based testing strategies. DESIGN AND SETTING: Rapid reviews were conducted and a consensus process then used to evaluate the evidence base within primary care settings. METHOD: The authors identified tests that are recommended or used commonly to monitor T2DM. Filtering questions were created to examine the rationale for use of each test, which were answered by stepwise rapid reviews of evidence cited by guidelines, systematic reviews, and individual studies. A consensus group of patient representatives and clinicians voted whether tests should be included or excluded based on the evidence or whether further evidence was needed. RESULTS: Of 15 tests, only haemoglobin A1c, to monitor disease progression and treatment response, and estimated glomerular filtration rate, to detect chronic kidney disease, have a strong evidence base. Based on available evidence and consensus group feedback, routinely testing for fructosamine to monitor disease progression; thyroid function, vitamin B12, ferritin, folate, clotting, bone profile, C-reactive protein, erythrocyte sedimentation rate, and B-type natriuretic peptide; and liver function for adverse treatment effects of metformin was deemed unnecessary. The study found insufficient evidence for lipids and haemoglobin to screen for secondary conditions, and for vitamin B12 to screen for adverse effects in those taking metformin. CONCLUSION: The study found that the evidence base for most T2DM monitoring tests is weak or absent. Clinicians should avoid non-evidence-based tests unless there are additional clinical indications for testing. Standardised evidence-based testing panels for T2DM and other long-term conditions could reduce unnecessary testing.