To evaluate self-monitoring of blood glucose, which is considered an important practice for patients with diabetes. However, little is known about the frequency or determinants of this technique.


A detailed questionnaire on diabetes was administered to a representative sample of 2405 diabetic subjects ≥ 18 yr of age in the U.S. population in the 1989 National Health Interview Survey.


Among subjects with IDDM, 40% monitored their blood glucose at least 1 time/day. Among subjects with NIDDM treated with insulin, 26% monitored at least 1 time/day and among NIDDM subjects not treated with insulin, the percentage was 5%. When stratified by age, little difference was observed between IDDM subjects and insulin-treated NIDDM subjects in the percentage testing at least 1 time/day. By multivariate analysis, age and insulin use were the major determinants of whether diabetic subjects tested their blood glucose. Race and education were also independently related to self-monitoring of blood glucose. Blacks were 60% less likely to test their blood glucose at least 1 time/day compared with non-Hispanic whites and Mexican Americans. Those with college education were 80% more likely to test their blood glucose compared with those with lower education levels. Having had a patient education class in diabetes management and frequent physician visits for diabetes care were positively related to self-testing. Self-monitoring was not related to higher income or having health insurance.


A large proportion of patients with diabetes do not test their blood glucose. Financial barriers associated with income and health insurance do not appear to impede the practice of self-monitoring. Because of the importance of blood glucose control in the prevention of diabetes complications and the role of self-monitoring in achieving blood glucose control, it may be prudent for physicians and their patients to make greater use of this technique. Special attention should be directed to the subgroups of patients (blacks, patients not treated with insulin, those with less education, and those with no education in diabetes) in which the frequency of self-monitoring is particularly low.

This content is only available via PDF.