Publications

Objectives

Current practice for people with diabetes includes annual monitoring for microalbuminuria, so that
preventive treatments such as ACE inhibitors can be given to those showing early signs of diabetic
nephropathy. Urine albumin tests can be highly variable even from day-to-day within an individual, and
so a positive test for microalbuminuria may occur in someone who in general has normal urine albumin
levels. We sought to estimate the frequency of such “false-positive” tests in individuals with type 1
diabetes under annual monitoring according to NICE guidelines.

Method

We used data on 483 individuals with type 1 diabetes aged 5-25 years from the Oxford Regional
Prospective Study (ORPS) to fit a statistical model for trends and variation in albumin creatinine ratio
(ACR), to estimate rates of positive tests in a monitoring programme based on NICE guidance..

Results

During up to 15 years follow-up the average increase in log-ACR per year was 0.021 (95% CI 0.012-
0.030), corresponding to 2.1% per year increase in ACR. The standard deviation of day-to-day
variation was 0.49. Using these estimates, if a population similar to ORPS were screened annually for
ACR>2.5 (males) or ACR>3.5 (females), there would be approximately 1 false-positive test per 1 truepositive
test over the first five years. There would also be approximately 1 false-negative test per 2
true positive tests.

Conclusions

The high number of false-positives (people measuring above thresholds who usually have normal
albuminuria) suggests that urine albumin monitoring based on NICE guidance in people with type 1
diabetes fails to target treatments appropriately.