Improved reproducibility of the 'large-window' method of assessing transcapillary and interstitial fluorescein diffusion in the skin in healthy subjects and in subjects with insulin-dependent diabetes mellitus

The transcapillary and interstitial diffusion of intravenously administered sodium fluorescein is used as a marker for capillary permeability. Fluorescein diffusion has been expressed by different parameters with reported coefficients of variation of 14-20%. Aim of the present study is to select a parameter which combines excellent reproducibility with the potential for discriminating insulin-dependent diabetic patients from healthy subjects. We performed three experiments to assess day-to-day reproducibility: 5 healthy subjects were measured twice, 1 healthy subject was measured 6 times and 1 subject with insulin-dependent diabetes mellitus was measured 5 times. We averaged the relative fluorescence light intensity [I-REL(t)] from dye arrival until a certain time point [I-AV(t)], instead of using the relative intensity at one time point. I-AV(7 min) showed markedly improved reproducibility, expressed as geometric mean of the coefficients of variation of the three separate experiments: 10%. In addition, a group of 12 insulin-dependent diabetic subjects was compared with 12 healthy control subjects. Median I-AV(7 min) was 69.5% (95% CI: 65.3-78.1%) in the diabetic subjects and 54.9% (95% CI: 52.1-60.0%) in the control subjects (p < 0.001). Since I-AV(7 min) combines excellent reproductibility with a good discriminating power, we advise its use in further studies.