Techniques of Monitoring Buried Fasciocutaneous Free Flaps

Sir: In their article titled “Techniques of Monitoring Buried Fasciocutaneous Free Flaps,” the authors demonstrate some persistent problems in the analysis of postoperative monitor performance in microsurgical flap cases.1 I have contended that there are three categories of data that usefully depict flap performance.2,3 In ascending order of importance, these categories are as follows: False-positives: How often does the monitor incorrectly indicate vascular occlusion and lead to unnecessary interventions? True-positives: How many flaps with vascular complications are actually detected by the monitor? Flap salvage: Of flaps with circulatory complications detected by the monitor, how many are salvaged? The third category may represent the best rationale for using a monitor. If a monitor does not lead to improved salvage of complicated flaps, why use it? To date, studies of hand-held Doppler monitoring, quantitative fluorometry, laser Doppler monitoring, and implantable Doppler probes have shown improved flap salvages rates compared with clinical observation in specific practice situations.2,4–8 Idiosyncrasies of practice situations and the monitors themselves have prevented even these systems from being universally used. A common error in analyses of monitor performance has been to equate a monitor's performance with its apparent ability to detect uncomplicated flaps. If, in a series of 100 flaps, a monitor is not falsely triggered for 95 uncomplicated flaps, one could naively say that the monitor was “95 percent effective.” The best that could really be said for the monitor is that it did not lead to unnecessary responses. If there were five complicated flaps that the monitor missed or did not detect in such a way as to lead to flap salvage, the monitor would have failed to achieve any measurable purpose. One way of analyzing monitor data is to hypothesize a nonfunctioning monitor and imagine the performance of this entity in comparison with the actual monitoring system. I have previously proposed the “green bean test” to serve in this capacity.2 If a green bean is placed on a flap, with the contingency that the bean will turn red if flap circulation has failed, a nonfunctioning monitor has been introduced, because the bean will never turn red. In the article by Ferguson and colleagues, how do their monitor data compare with the imagined performance of the green bean (or, more simply, no monitor at all)? The comparison is summarized in Table 1.Table 1: Comparative Monitor PerformancesThis analysis shows that, in the series presented, the green bean performed better than the implantable Doppler probe (a previously reported observation).2 The green bean performed as well as the hand-held Doppler and the monitor island. The green bean appears to be a superior monitoring system compared with all the others combined. One conclusion that could be drawn from this analysis is that the authors could consider discontinuing all monitoring with a neutral or possibly beneficial effect on their clinical management. A more general conclusion is that investigators, reviewers, and readers should concentrate on specific performance data when evaluating monitoring strategies used in microsurgical cases. False-positives, complication detection rates and, most importantly, flap salvage rates define the utility of monitors. William Lineaweaver, M.D. Rankin Plastic Surgery Center 348 Crossgates Boulevard, Suite 2400 Brandon, Miss. 39042 [email protected]