Surgery for hip fracture yields societal benefits that exceed the direct medical costs

Abstract

Background: A hip fracture is a debilitating condition that consumes significant resources in the United States. Surgical treatment of hip fractures can achieve better survival and functional outcomes than nonoperative treatment, but less is known about its economic benefits.

Questions/purposes: We asked: (1) Are the societal benefits of hip fracture surgery enough to offset the direct medical costs? (2) Nationally, what are the total lifetime benefits of hip fracture surgery for a cohort of patients and to whom do these benefits accrue?

Methods: We estimated the effects of surgical treatment for displaced hip fractures through a Markov cohort analysis of patients 65 years and older. Assumptions were obtained from a systematic literature review, analysis of Medicare claims data, and clinical experts. We conducted a series sensitivity analyses to assess the effect of uncertainty in model parameters on our estimates. We compared costs for medical care, home modification, and long-term nursing home use for surgical and nonoperative treatment of hip fractures to estimate total societal savings.

Results: Estimated average lifetime societal benefits per patient exceeded the direct medical costs of hip fracture surgery by USD 65,000 to USD 68,000 for displaced hip fractures. With the exception of the assumption of nursing home use, the sensitivity analyses show that surgery produces positive net societal savings with significant deviations of 50% from the base model assumptions. For an 80-year-old patient, the breakeven point for the assumption on the percent of patients with hip fractures who would require long-term nursing home use with nonoperative treatment is 37% to 39%, compared with 24% for surgical patients. Nationally, we estimate that hip fracture surgery for the cohort of patients in 2009 yields lifetime societal savings of USD 16 billion in our base model, with benefits and direct costs of USD 21 billion and USD 5 billion, respectively. For an 80-year-old, societal benefits ranged from USD 2 billion to USD 32 billion, using our range of estimates for nursing home use among nonoperatively treated patients who are immobile after the fracture.

Conclusions: Surgical treatment of hip fractures produces societal savings. Although the magnitude of these savings depends on model assumptions, the finding of societal savings is robust to a range of parameter values.

Level of evidence: Level III, economic and decision analyses. See the Instructions for Authors for a complete description of levels of evidence.

Fig. 1A–B

(A) A decision tree shows the treatment pathway and health states in the Markov model of hip fractures. The surgical branch of intracapsular fractures consists of four health states: dead, well, infection revision, and aseptic revision (infection and aseptic revisions are represented by one oval in the figure). In the first year after surgery, living patients enter the well state. The well state includes good and fair outcomes. For patients in the well state, they can die, stay in that health state, or have a revision surgery in the subsequent year. The nonoperative branch consists of three states: dead, survive - immobile, and survive - mobile. Once patients enter either survive - immobile or survive - mobile, they stay there until they die. (B) The surgical branch of extracapsular fractures consists of five health states: dead, well, conversion to arthroplasty, infection revision arthroplasty, and aseptic revision arthroplasty (infection and aseptic revisions are represented by one oval in the figure). The well state includes good and fair outcomes, because distribution and utility data for these separate health states were unavailable for extracapsular fracture. Patients can die, do well, or undergo conversion surgery to arthroplasty during the first year. For patients who had a conversion to arthroplasty, they can die, stay in that state, or have a revision arthroplasty in the subsequent year. The nonoperative branch consists of three states: dead, survive - immobile, and survive - mobile. Once patients enter either survive - immobile or survive - mobile, they stay there until they die.