Abstract
IMPORTANCE: The persistent shortage of dermatologists in the United States affects access to care and patient outcomes. OBJECTIVE: To characterize the effect of geographic variations in dermatologist density on the provision of dermatology procedures within Medicare. DESIGN, SETTING, AND PARTICIPANTS: This was a cross-sectional study using the 2013 Medicare Provider Utilization and Payment Database. Dermatology-related procedures were defined by the top 50 billing codes accounting for more than 95% of procedures billed by dermatologists. Billing codes corresponding to evaluation and monitoring visits and dermatopathology were excluded. Total costs were estimated from the Centers for Medicare & Medicaid Services physician fee schedule, based on the nonfacility national payment amount with no modifiers. Nationally representative administrative database that includes 100% of charges billed by noninstitutional clinicians covered under Medicare Part B. A total of 10 391 dermatologists practicing within the 50 states and Washington, DC, were included. The Medicare-eligible population was defined as all persons 65 years or older. EXPOSURES: Density of dermatologists, categorized into first (5.3 per 100 000 persons ≥65 years) through fifth (54.8 per 100 000 persons ≥65 years) quintiles. MAIN OUTCOMES AND MEASURES: Utilization of dermatology procedures (mean volume per 100 000 persons ≥65 years) and total cost (mean amount billed per person ≥65 years) by clinician type across quintiles of dermatologist density. RESULTS: In 2013, dermatologists billed Medicare for 28 million procedures costing $2.21 billion. Mean billed amount by dermatologists per person 65 years or older was $15.87 in the lowest-density quintile vs $92.02 in the highest-density quintile. This trend suggests that each interval increase of 10 dermatologists per 100 000 persons 65 years or older is correlated with a $14.81 increase in Medicare spending on dermatology procedures (95% CI, 8.28-21.34; P = .005). Utilization of these procedures differed among clinician types, with dermatologists largely performing destruction of premalignant lesions and PCPs primarily doing injections. CONCLUSIONS AND RELEVANCE: There is evidence of supply-sensitive variation in the provision of dermatology procedures for the Medicare-eligible population; higher dermatologist density is correlated with increased utilization of dermatology procedures and subsequent billed charges to Medicare. Further research is needed to determine the effect of such variations on outcomes and whether incentives can better align dermatologists with areas of clinical need.