Variation in the use of differentially-reimbursed hip fracture procedures.

Abstract

Background: This series of three papers examine the patient, surgeon and hospital factors that were associated with intramedullary nail (IMN) use, the geographic variation in IMN use, and the association between provider volume and short-term mortality following the surgical treatment of Medicare intertrochanteric hip fracture patients during 2000 through 2002. Methods: Medicare claims and enrollment data 2000-2002 identified Medicare beneficiaries age 65 or older who underwent an inpatient surgery to treat an intertrochanteric femur fracture using either an IMN or plate/screws. Surgeon and hospital characteristics from the MPIER physician enrollment and Provider of Services (hospital) files were merged with the claims. Logistic regression was used to examine the geographic variation in IMN use by state and year, with and without adjustment for patient factors. Generalized linear mixed models with fixed and random effects were used to model the association between surgeon and hospital factors, and device choice (IMN or plate/screws), while controlling for patient factors. Fixed effects regression analyses using generalized estimating equations (GEE) were used to examine the association of provider volume and inpatient through 90-day mortality following intertrochanteric hip fracture surgery with internal fixation. Results: Geographic variation in IMN use from 2000-2002 was extensive, was largely not explained by patient factors, and the observed surgeon practice patterns persisted over time. Surgeon factors, resident case involvement and teaching hospital status were strong predictors of IMN use. Surgeons under age 45, those operating at more than one hospital and doctors of osteopathy were significantly more likely to use IMN. Surgeon more than hospital factors accounted for high IMN use. For the short-term mortality outcomes, the magnitude of the positive hospital volume effect was smaller among hip fracture patients than has been noted among elective arthroplasty patients. There were no significant mortality differences among patients of the lowest compared with the highest volume hip fracture surgeons. Conclusions: Device choice for Medicare intertrochanteric hip fracture patients was not based primarily on patient factors in 2000-2002. Device choice was and continues to be driven by factors other than substantial clinical outcomes evidence, particularly within certain subsets of providers. Both orthopaedic faculty involved in the training of new surgeons and early-career surgeons who are taught evidence-based medicine principles were the providers observed to have the highest IMN use. In addition to preferential training, higher Medicare reimbursement to surgeons for IMN is likely contributing to substantial increases in IMN use when a less-expensive procedure would give similar outcomes in the majority of cases. Subsequent to an RVU payment incentive, IMN use has increased dramatically, yet this increased IMN use has resulted in no better quality for most patients and at a higher cost to both hospitals and Medicare.