Optimal Frequency of Urinary Albumin Screening in Type 1 Diabetes

OBJECTIVE Kidney disease screening recommendations include annual urine testing for albuminuria after 5 years' duration of type 1 diabetes. We aimed to determine a simple, risk factor-based screening schedule that optimizes early detection and testing frequency. RESEARCH DESIGN AND METHODS Urinary albumin excretion measurements from 1,343 participants in the Diabetes Control and Complications Trial and its long-term follow-up were used to create piecewise-exponential incidence models assuming 6-month constant hazards. Likelihood of the onset of moderately or severely elevated albuminuria (confirmed albumin excretion rate AER >= 30 or >= 300 mg/24 h, respectively) and its risk factors were used to identify individualized screening schedules. Time with undetected albuminuria and number of tests were compared with annual screening. RESULTS The 3-year cumulative incidence of elevated albuminuria following normoalbuminuria at any time during the study was 3.2%, which was strongly associated with higher glycated hemoglobin (HbA(1c)) and AER. Personalized screening in 2 years for those with current AER <= 10 mg/24 h and HbA(1c) <= 8% (low risk [0.6% three-year cumulative incidence]), in 6 months for those with AER 21-30 mg/24 h or HbA(1c) >= 9% (high risk [8.9% three-year cumulative incidence]), and in 1 year for all others (average risk [2.4% three-year cumulative incidence]) was associated with 34.9% reduction in time with undetected albuminuria and 20.4% reduction in testing frequency as compared with annual screening. Stratification by categories of HbA(1c) or AER alone was associated with reductions of lesser magnitude. CONCLUSIONS A personalized alternative to annual screening in type 1 diabetes can substantially reduce both the time with undetected kidney disease and the frequency of urine testing. Article Highlights Kidney disease screening recommendations include annual urine testing for albuminuria after 5 years' duration of type 1 diabetes. We investigated simple screening schedules that optimize early detection and testing frequency. Personalized screening in 2 years for those with current AER <= 10 mg/24 h and HbA(1c) <= 8%, in 6 months for those with AER 21-30 mg/24 h or HbA(1c) >= 9%, and in 1 year for all others yielded 34.9% reduction in time with undetected albuminuria and 20.4% fewer evaluations compared with annual screening. A personalized alternative to annual screening in type 1 diabetes can substantially reduce both the time with undetected kidney disease and the frequency of urine testing.