Abstract: PD32-07
Sources of Funding: UCSF Teaching to Choose Wisely Grant, 2015-2016

Introduction

Cost-control initiatives targeting physicians are currently underway to improve cost-efficient clinical care. Little is known regarding surgeon-specific cost choices. We sought to characterize surgeon periorperative spending to evaluate the accuracy of a singular score summarizing surgeon cost-efficiency.

Methods

All surgeons operating at UCSF Medical Center from July 2012-October 2015 across all specialties were included. Mean OR cost was calculated combining surgical supplies and time in the OR ($/minute). Cases were included if ≥ 4 surgeons performed said operation and all included surgeons completed ≥ 20 operations. Z-scores, defined as [(cost of surgeon&[prime]s case) - (median cost for said case across all surgeons)] / [standard deviation (SD) of all surgeons&[prime] costs for said case], were calculated for each surgeon and surgical case performed. Estimated mean z-scores were then calculated using mixed effects modeling. Surgeons&[prime] z-score variability was evaluated using calculated SDs.

Results

257 surgeons across 17 surgical subspecialties were included in the analysis. The mean raw z-score was -0.12 ranging from -1.85 to 3.4 with SD of 0.78. The estimated z-score variability across the surgeons ranged from -0.5 to 0.5 SDs from the average estimated z-score (Figure 1). When comparing surgical subspecialties, the variability of z-scores ranged from 0.06 to -0.07. In individual subspecialties, the variability (SD) ranged from 0.25 to 0.02.

Conclusions

The z-score variability across all surgeons was small, supporting consistent surgeon-spending behavior across diverse cases. This supports using a surgeon&[prime]s estimated mean z-score as a singular and accurate metric characterizing surgeons&[prime] cost behavior.

Funding

UCSF Teaching to Choose Wisely Grant, 2015-2016