[Validity of urinary glucose test for diabetes screening in workplace regular medical checkups]

It has been acknowledged that urinary glucose level varies depending on the time after meal which may hence affect validity of urinary glucose test as a screening test for diabetes. However, sample collection time has not been standardized for urinary glucose test in workplace regular medical checkups. In this study, we investigated the effect of timing of urine sampling on results of urinary glucose screening tests, and using data obtained from the 75 g oral glucose tolerance tests (OGTT) evaluated the validity of urinary glucose test in workplace medical checkups. Also, we made a survey among industrial physicians about the timing of sample collection. Between 1991 and 1996, 455 males and 116 females who participated in a 2-day health examination including the OGTT in a hospital in Chiba were used as study subjects. These did not include an additional 22 subjects who had already been diagnosed as having diabetes but received the OGTT. The examinees observed a strict fast after supper the previous evening and urine and blood samples were collected before and 2 hours after glucose intake for the OGTT. Diagnosis of diabetes was made following the criteria of WHO (1985) and screening test results were defined positive when urinary glucose level was equal to or more than 40 mg/dl. Sensitivity and specificity were calculated for diabetes and impaired glucose tolerance separately for male and female. Industrial physicians who participated in a lecture course for qualification were surveyed about the timing of urine sample collection for urinary glucose test in workplace regular medical checkups. For the urine samples collected 2 hours after glucose intake, sensitivity was very high (male 84%, female 100%), in addition to a relatively high specificity (male 71%, female 92%). However the samples collected before glucose intake showed extremely poor sensitivity (male 11%, female 0%). Similar results were obtained when screening was made for impaired glucose tolerance and both diabetes and impaired glucose tolerance combined. The survey among the industrial physician revealed that in the majority (58%) of workplace medical checkups, the urine sample had been collected when examinees were fasting. The largest reason for this was that the urinary glucose test was performed together with upper gastro-intestinal examination or measurement of serum triglyceride both of which require fasting of examinees. The results of the present study showed that urinary glucose tests have been frequently performed at a timing that produces low validity. The procedures of workplace regular medical checkups should be evaluated with the concept of Evidence Based Medicine. The standard of the timing of urine sample collection for urinary glucose test should be carefully examined.