# Physician Specialty, Experience, Region, Gender, and the Socioeconomic Status of Practice Location Predict Cost of Skin Cancer Procedures Performed in the Medicare Population

## TL;DR

This study finds that factors like physician specialty, gender, and location influence the cost of skin cancer procedures for Medicare patients.

## Contribution

The study identifies physician and practice characteristics that predict cost differences in skin cancer procedures among Medicare patients.

## Key findings

- Dermatologists initially billed higher for removals, but clustering revealed four distinct cost patterns.
- Male physicians, experienced practitioners, and those in affluent regions had higher costs.
- Cost differences between clusters reached up to $1,735 per procedure.

## Abstract

Introduction: Skin cancer affects approximately five million individuals annually in the United States, with nonmelanoma skin cancers (NMSC) and melanoma amounting to $8.9 billion in annual healthcare costs.

Objective: Given the growing financial burden on the Medicare population, the present study aims to identify cost predictors to help detect outlier physician behavior and reduce healthcare costs.

Materials and methods: Using Medicare payments from 2017-2021, we analyzed malignant destructions and excisions (collectively, “removals”) and cutaneous closures on the trunk, arms, and legs. We included physicians from seven specialties and analyzed physician gender, years of experience, region, and socioeconomic status indicators. We subdivided physicians into three specialty subgroups, dermatologists, surgeons, and generalists, and characterized their overall case volume and costs. Then, adjusting for procedure type and volume, we calculated differences between actual and database-wide expected procedural costs and used k-means clustering to group physician cost patterns.

Results: Prior to adjustments, dermatologists billed higher-than-expected for removals, while surgeons had higher-than-expected total closure costs. However, with subsequent weighted adjustments, cluster analysis identified four unique physician cost phenotypes. Clustering demonstrated male physician gender, more experience, and practice in the Northeast, West, and affluent areas predicted high costs, while lower costs were associated with dermatologists and Midwestern practitioners. Between the highest and lowest cost clusters, the summed cost for a removal and closure differed by up to $1,735.

Conclusion: Physician and practice characteristics predict substantial cost differences, highlighting the need to understand these variations to optimize cost-effective skin cancer care.