In a population-based U.K. study, screening all middle-aged high-risk patients did not help.

Early diagnosis of type 2 diabetes can improve clinical outcomes, and mathematical models have suggested that widespread screening might lower diabetes-related mortality. To test this hypothesis, researchers cluster-randomized 33 general practices in eastern England to offer either a single round of formal screening to all middle-aged patients (age range, 40–69) considered to be at high risk for diabetes according to specified clinical characteristics (28 practices with approximately 16,000 high-risk patients) or to continue diagnosing diabetes in such patients through routine clinical assessment (5 practices with approximately 4000 high-risk patients).

Overall, 73% of invited patients participated in screening, and 3% of patients in both groups received diabetes diagnoses. After a median 9.6 years of follow-up, overall mortality among high-risk patients was similar in the screening and no-screening practices (10.5 and 9.9 deaths per 1000 person-years, respectively). Rates of cardiovascular- and cancer-related mortality also were similar in the two groups.

Comment: The study population was of above-average socioeconomic status; these results might not be generalizable to less-affluent populations in which the prevalence of undiagnosed diabetes might be higher. In addition, multiple rounds of screening and longer follow-up might be necessary to detect mortality differences attributable to screening. Furthermore, important outcomes other than mortality might be modifiable by screening. Regardless, these data suggest that population screening for diabetes might be less effective than expected and should be evaluated carefully before widespread implementation.

Source: Journal Watch General Medicine