Comparison of glucose levels in dermal interstitial fluid and finger capillary blood.

Abstract

Alternative methods for self-monitoring of blood glucose have been pursued by many researchers, largely in response to evidence gathered in several long-term studies of patients with diabetes mellitus. These studies suggest that long-term complications of the disease may be mitigated if the disease is intensively managed, a component of which is increased monitoring. Many of the alternative methods utilize interstitial fluid (ISF) as the diagnostic fluid, rather than finger blood. A time lag in the distribution of glucose from blood to the interstitium has been observed by many, with estimates of lag time varying from none to 45 min. Dermal ISF was sampled from diabetic subjects in two tests and compared to finger blood glucose. In the first test, data were collected over time in a manner that allowed a cross-correlation analysis to predict an average lag time. Information from this test was then used as input to a data collection format for a method comparison test of 691 patients with diabetes in which ISF data were collected immediately after the finger blood reference and 15 min after the reference. An average lag time of about 25 min was determined from the cross-correlation analysis, with the correlation error reduced by three-fourths within a 15-min lag time. In the method comparison test, the correlation coefficient between finger blood glucose and ISF glucose improved from 0.923 to 0.951, and the percentage of data in the A zone of the Clarke Error Grid rose from 80.2% to 90.6% for the ISF glucose data collected at no lag and 15-min lag, respectively. Dermal ISF glucose measurement might be a reasonable alternative to blood glucose measurement for patients routinely monitoring ambient glycemia, although more testing in the sensitive hypoglycemic range is needed to clarify what might happen in cases of rapidly changing glucose.