Discovering healthcare provider behavior patterns through the lens of Medicare excess charge

The phenomenon of excess charge, where a healthcare service provider bills Medicare beyond the limit allowed for a medical procedure, is quite common in the United States public healthcare system. For example, in 2014, healthcare providers charged an average of 3.27 times (and up to 528 times) the allowable limit for cataract surgery. Previous research contends that such excess charges may be indicative of the actual amount that providers bill to non-Medicare patients and subsequent cost-shifting behavior, where a healthcare provider tries to recoup underpayment by Medicare from privately insured, self-pay, out-of-network, and uninsured patients.

The objective of this study is to examine the drivers of a provider's excess charge patterns, especially the extent to which the degree of excess charges may be associated with physician characteristics, Medicare reimbursement policy, or socioeconomic status and demographics of a provider's patient base.

Using data from the 2014 Medicare Provider Utilization files, we identify three procedures with the highest variation in Medicare reimbursements to study the excess charge phenomenon. We then employ a two-step cluster analysis within each procedure to identify distinct provider groups.

Each procedure code yielded distinct healthcare provider segments with specific patient demographics and related behavior patterns. Cluster silhouette coefficients indicate that these segments are unique. Three random subsamples from each procedure establish the stability of the clusters.

For each of the three procedures investigated in this study, a sizeable number of healthcare providers serving poorer, riskier patients are often paid significantly lower than their peers, and subsequently have the highest excess charges. For some providers, excess charges reveal possible cost-shifting to private insurance. Patterns of excess charges also indicate an imbalance of market power, especially in areas with lower provider competition and access to health care, thus leading to urban-rural healthcare disparities. Our results reinforce the call for price transparency and an upper limit to overbilling.