Geographic Variation in Medicare Fee-for-Service Health Care Expenditures Before and After the Passage of the Affordable Care Act

Question Which categories of spending were associated with reductions in geographic variation of Medicare per-beneficiary spending across the US after the passage of the Affordable Care Act? Findings In this cross-sectional study of Medicare enrollees aged 65 years or older, geographic variation in Medicare fee-for-service spending per beneficiary was stable from 2007 to 2011 and then declined steadily from 2012 to 2018. A key factor associated with reduced geographic variation in spending was reduced variation in postacute care spending, specifically home health spending. Meaning These findings suggest that antifraud enforcement efforts and payment reforms that were instituted as part of the Affordable Care Act may have reduced geographic variation in Medicare fee-for-service per-beneficiary spending, although significant geographic variation remains. Importance Geographic variation in Medicare spending is often used as a measure of wasteful spending. A 2013 Institute of Medicine report found that postacute care was a key contributor of geographic variation from 2007 to 2009. However, payment reforms and antifraud efforts implemented after the passage of the Affordable Care Act (ACA) may have reduced geographic variation in spending, especially postacute care spending. Objective To investigate how geographic variation in Medicare fee-for-service per-beneficiary spending changed from 2007 to 2018 before and after passage of the ACA. Design, Setting, and Participants This cross-sectional study included all fee-for-service Medicare enrollees 65 years or older from January 1, 2007, to December 31, 2018. The fee-for-service Medicare Geographic Variation Public Use File was used to group hospital referral regions (HRRs) in each year into deciles (10 equal groups) based on per-beneficiary total spending. The difference between the per-beneficiary monthly spending in each decile and the national mean, as well as the ratio of per-beneficiary total spending in the top deciles to that of the bottom decile, were reported. Data analysis occurred from July 22, 2019, to October 21, 2021. Main Outcomes and Measures Per-beneficiary spending on hospital inpatient, hospital outpatient, physician, and postacute care (and type of postacute care). Results There were 27.2 million fee-for-service beneficiaries in 2007 (58.0% women) and 28.3 million beneficiaries in 2018 (55.9% women). Per-beneficiary Medicare spending was $9691 in 2007 and $9847 in 2018 (using inflation-adjusted 2018 dollars). Geographic variation in Medicare spending was stable from 2007 to 2011 and then declined steadily from 2012 to 2018. The ratio of per-beneficiary total Medicare spending in the HRRs in the top decile to the bottom decile was 1.68 in 2007 ($415 monthly difference in spending) but only 1.56 ($361 monthly difference in spending) in 2018 (estimated change, -0.12 [95% CI, -0.21 to -0.02]; P = .01). Focusing on specific spending categories, the only statistically significant reductions in geographic variation were found for home health; the ratio of home health spending among HRRs in the top to bottom deciles of total Medicare spending fell from 5.14 in 2007 to 3.45 in 2018 (change, -1.69 [95% CI, -3.30 to -0.09]; P = .04). Conclusions and Relevance Geographic variation in total per-beneficiary Medicare spending fell from 2007 to 2018, with home health spending being a key factor associated with geographic variation. The ACA's value-based payment programs and enhanced integrity efforts in home health provide a possible explanation for the decrease. This cross-sectional study explores which components of Medicare spending were associated with reductions in geographic variation and potential policy mechanisms after passage of the Affordable Care Act among enrollees 65 years or older.