Comparative Accuracy of Clinical and Formula-Based Estimation of Resection Weight in Reduction Mammoplasty and Implications for Insurance Preapproval and Payment.

INTRODUCTION:Insurance companies use minimum resection weight, sometimes based on body surface area (Schnur sliding scale), as a criterion for preapproval and ultimately coverage of reduction mammoplasty. The purpose of this study is to compare the accuracy of subjective resection estimates and estimates calculated by published formulae versus measured resection weights, and to explore the impact of these estimates on insurance preauthorization and payment. METHODS:A retrospective chart review of bilateral reduction mammaplasties performed at a single academic medical center by seven plastic surgeons from January 2011 to December 2017 was performed. Patients undergoing oncoplastic reduction, simultaneous additional body-contouring procedures, or lacking complete data were excluded. A total of 762 patients were reviewed. Absolute and relative errors between preoperative estimate and actual resection weights were calculated. A subset of patients with requisite breast measurements (n = 579) was examined to compare formula-based with clinical estimates of resection weights. RESULTS:Median error was 105 g (14% normalized by resection weight). Frequency of underestimation (40.5%) and overestimation (55.7%) were similar. In 19% (n = 291) of reduced breasts, resection estimate was less than the Schnur requirement. For 5 (2.8%) of these patients, insurers denied coverage explicitly for this reason. Our surgeons' positive predictive value of estimate > Schnur was 86.6%. In 23% (n = 352) of breasts, resection was < Schnur requirement. No insurance claim was denied a posteriori due to resection weight less than Schnur. The formula proposed by Appel et al. produced the most accurate estimates, and is the most likely to produce an estimate < Schnur in nonobese women. Correlations between each surgeon's relative errors and years of faculty experience (r < 0.07) and number of reduced breasts (r = 0.0275) were very weak. CONCLUSIONS:Resection estimate accuracy varies among surgeons and does not appear to be affected by experience. Because insurers use resection estimates to determine preauthorization, this could be problematic, particularly for surgeons tending to underestimate. However, insurers are inconsistent in application of the Schnur requirement once surgery has been preapproved and its validity as a determinant of medical necessity is in question.