Point-of-care assessment of microvascular blood flow in critically ill patients.

Abstract

Sublingual microvascular videomicroscopy can assess tissue perfusion in critically ill patients; however, data analysis is currently limited to delayed off-line evaluation. We hypothesized that a real-time point-of-care (POC) determination of the microcirculatory flow index (MFI), an established metric for assessing microcirculatory perfusion, agrees well with the conventional off-line analysis. Prospective observational study. Urban academic intensive care unit. A heterogeneous population of critically ill patients. A single operator performed side stream darkfield videomicroscopy of the sublingual microcirculation and made a POC determination of MFI in real-time on a portable bedside monitor by assigning a score (0 = no flow to 3 = normal) to each quadrant of the image and averaging the four values. After image processing, de-identification and randomization, the same operator, blinded to the previous interpretation, repeated the MFI assessment by viewing an AVI-formatted image sequence on a 94 cm 1,080 pixel LCD monitor (reference standard). There were 205 paired measurements in 18 subjects. The POC and reference standard MFI had good agreement by Bland-Altman analysis [mean difference of -0.031, SD = 0.198 (95% CI, -0.43 to 0.37)]. The POC assessment was 94% sensitive and 92% specific for detecting impaired microvascular flow (defined a priori as an MFI < 2.5 based on previously published data). A POC determination of MFI had good agreement with conventional off-line analysis, and was highly sensitive and specific for detecting impaired microvascular flow. This real-time technique may be useful in future clinical trials targeting impaired microcirculatory perfusion in critically ill patients.