1612 C OPYRIGHT Ó 2014

BY

T HE J OURNAL

OF

B ONE

AND J OINT

S URGERY, I NCORPORATED

Cost-Effectiveness Analysis of Fixation Options for Intertrochanteric Hip Fractures Eric Swart, MD, Eric C. Makhni, MD, MBA, William Macaulay, MD, Melvin P. Rosenwasser, MD, and Kevin J. Bozic, MD, MBA Investigation performed at Columbia University Medical Center, New York, NY, and the University of California, San Francisco, California

Background: Intertrochanteric hip fractures are a major source of morbidity and financial burden, accounting for 7% of osteoporotic fractures and costing nearly $6 billion annually in the United States. Traditionally, ‘‘stable’’ fracture patterns have been treated with an extramedullary sliding hip screw whereas ‘‘unstable’’ patterns have been treated with the more expensive intramedullary nail. The purpose of this study was to identify parameters to guide cost-effective implant choices with use of decision-analysis techniques to model these common clinical scenarios. Methods: An expected-value decision-analysis model was constructed to estimate the total costs and health utility based on the choice of a sliding hip screw or an intramedullary nail for fixation of an intertrochanteric hip fracture. Values for critical parameters, such as fixation failure rate, were derived from the literature. Three scenarios were evaluated: (1) a clearly stable fracture (AO type 31-A1), (2) a clearly unstable fracture (A3), or (3) a fracture with questionable stability (A2). Sensitivity analysis was performed to test the validity of the model. Results: The fixation failure rate and implant cost were the most important factors in determining implant choice. When the incremental cost for the intramedullary nail was set at the median value ($1200), intramedullary nailing had an incremental cost-effectiveness ratio of $50,000/quality-adjusted life year when the incremental failure rate of sliding hip screws was 1.9%. When the incremental failure rate of sliding hip screws was >5.0%, intramedullary nails dominated with lower cost and better health outcomes. The sliding hip screw was always more cost-effective for A1 fractures, and the intramedullary nail always dominated for A3 fractures. As for A2 fractures, the sliding hip screw was cost-effective in 70% of the cases, although this was highly sensitive to the failure rate. Conclusions: Sliding hip screw fixation is likely more cost-effective for stable intertrochanteric fractures (A1) or those with questionable stability (A2), whereas intramedullary nail fixation is more cost-effective for reverse obliquity fractures (A3). These conclusions are highly sensitive to the fixation failure rate, which was the major influence on the model results.

Peer Review: This article was reviewed by the Editor-in-Chief and one Deputy Editor, and it underwent blinded review by two or more outside experts. It was also reviewed by an expert in methodology and statistics. The Deputy Editor reviewed each revision of the article, and it underwent a final review by the Editor-in-Chief prior to publication. Final corrections and clarifications occurred during one or more exchanges between the author(s) and copyeditors.

A

pproximately 150,000 intertrochanteric hip fractures occur per year in the United States, accounting for 7% of all osteoporotic fractures1,2. They represent a major cost burden to the United States health-care system, with total treatment costs estimated at approximately $6 billion annually1,2. Therefore, increasing scrutiny is being directed at all aspects of hip fracture treatment, including fixation type, and several studies have recently evaluated factors associated with the choice of fixation type3-5.

Implant choice is frequently determined by whether the intertrochanteric fracture is ‘‘stable’’ or ‘‘unstable’’6-8. Stable fractures include those that course from the greater trochanter inferomedially toward the lesser trochanter with preservation of the posteromedial cortex, the lesser trochanter, and the lateral wall. Those with a compromised medial cortex, loss of lateral wall integrity, extension into the subtrochanteric region, or a reverse obliquity pattern have been traditionally considered unstable6,8,9.

Disclosure: None of the authors received payments or services, either directly or indirectly (i.e., via his or her institution), from a third party in support of any aspect of this work. One or more of the authors, or his or her institution, has had a financial relationship, in the thirty-six months prior to submission of this work, with an entity in the biomedical arena that could be perceived to influence or have the potential to influence what is written in this work. No author has had any other relationships, or has engaged in any other activities, that could be perceived to influence or have the potential to influence what is written in this work. The complete Disclosures of Potential Conflicts of Interest submitted by authors are always provided with the online version of the article.

J Bone Joint Surg Am. 2014;96:1612-20

d

http://dx.doi.org/10.2106/JBJS.M.00603

1613 TH E JO U R NA L O F B O N E & JO I N T SU RG E RY J B J S . O RG V O LU M E 96 -A N U M B E R 19 O C T O B E R 1, 2 014 d

d

d

C O S T -E F F E C T I V E N E S S A N A LY S I S O F F I X AT I O N O P T I O N S FO R INTERTROC HANTERIC HIP FRACTU RES

Fig. 1

AO classification of intertrochanteric (type 31-A) femoral fractures. Fig. 1-A Type A1 is simple with minimal comminution. Fig. 1-B Type A2 is more comminuted, with medial calcar or lateral buttress comminution and/or subtrochanteric extension. Fig. 1-C Type A3 is unstable, with reverse obliquity or transverse orientation. (Reproduced, with permission, from: M¨uller AO Classification of Fractures—Long Bones. https://www.aofoundation.org/ Documents/mueller_ao_class.pdf. Copyright by AO Foundation, Switzerland.)

A variety of implants are available for surgical treatment of intertrochanteric fractures, but the most common designs can be categorized as extramedullary sliding hip screws or intramedullary nails. Fixation of stable fractures with sliding hip screw constructs have generally had favorable outcomes6,9,10. Intramedullary nailing has recently become the treatment of choice for unstable fractures6, although this transition is controversial6,8,10. According to the AO fracture classification system (Fig. 1), type A1 is universally considered stable and type A3 is generally considered unstable. A2 fractures are less clear, but a common opinion is that A2-1 fractures are stable whereas A2-2 and A2-3 fractures are unstable, based on the degree of increasing medial comminution and orientation of the fracture line8,10,11. Despite the routine use of intramedullary nailing for unstable fractures, the biomechanical advantages of this fixation method have yet to be clearly correlated with clinical superiority. This lack of evidence has led to ambiguity in the decisionmaking process, specifically with regard to the appropriate use of intramedullary nails, which are generally more expensive. The reason for this confusion is multifold: first, many studies of fixation failure rates for intertrochanteric fractures have included multiple fracture types or the authors did not segregate by precise fracture classification9. The inclusion of both stable and unstable patterns in the same cohort likely results in an underestimation of the failure risk of using sliding hip screws for unstable fractures. Second, there is a well-documented lack of interobserver reliability in classifying intertrochanteric fractures, especially within AO subtypes12-15, further compromising the validity of correlating outcomes with specific fracture characteristics. Finally, many earlier studies (prior to 2000) showed relatively high intraoperative and postoperative rates of complications, particularly femoral fractures, with the use of intramedullary nailing16, an issue that appears to have been largely resolved with the most recent generation of intramedullary

nails17; this has added an important confounding variable to the observed failure/reoperation rates in earlier trials. As a result of this uncertainty, surgeons treating a questionably unstable intertrochanteric fracture are frequently faced with a difficult decision—namely, whether to use a more expensive implant (an intramedullary nail) to potentially minimize the risk of failure or be ‘‘cost conscious’’ and use a less expensive implant (a sliding hip screw) whenever possible. The authors of several recent studies have noted increased use of intramedullary nails3-5, and concern has been raised about the increased cost associated with this trend. However, to our knowledge, no one has evaluated this complex decision process from a formal costeffectiveness perspective. The purpose of this study was to use expected-value decision-analysis techniques to model this common clinical scenario. The question we addressed was: in the face of ambiguous clinical data, what is the most cost-effective strategy for fixation of an intertrochanteric fracture when there is concern that the fracture pattern might be unstable? Materials and Methods Study Design

T

his economic decision analysis was performed according to the guidelines in 18 the Panel on Cost-Effectiveness Analysis in Health and Medicine . Our reference case was a theoretical cohort of elderly patients with an intertrochanteric hip fracture. All costs were considered from a societal perspective and based on those reported in the primary literature. The time horizon was eight years, which, 19 based on actuarial tables, was the median survival for our theoretical cohort .

Model Design An expected-value decision-analysis model was constructed to estimate the total costs and total gains in quality-adjusted life years (QALYs) with two competing strategies: fracture fixation with (1) a sliding hip screw or (2) an intramedullary nail. We assumed that, other than fixation choice and method of implantation, all other medical treatment was identical.

1614 TH E JO U R NA L O F B O N E & JO I N T SU RG E RY J B J S . O RG V O LU M E 96 -A N U M B E R 19 O C T O B E R 1, 2 014 d

d

d

C O S T -E F F E C T I V E N E S S A N A LY S I S O F F I X AT I O N O P T I O N S FO R INTERTROC HANTERIC HIP FRACTU RES

Fig. 2

Decision tree model used to predict cost and outcome as a function of fixation type. After the initial decision nodes, probability nodes (circles) and outcome nodes (triangles) are shown. IT = intertrochanteric. Our decision tree model, outlined in Figure 2, describes four major health states after surgery: (1) healed fracture, (2) perioperative femoral fracture requiring revision, (3) implant failure or nonunion requiring revision, or (4) death in the perioperative period. Revision surgical procedures for femoral fracture and implant failure were also assigned mortality rates. Final health state was modeled as (1) survival after successful fixation, (2) survival after successful revision, or (3) death. After the model was established and tested with sensitivity analysis, a probabilistic sensitivity analysis was conducted to evaluate outcomes according to AO fracture type (A1, A2, or A3). As definitive failure rates have not been reported, the failure rate was allowed to vary within the ranges reported in 7,16 meta-analyses of those fracture types , and a Monte Carlo simulation was utilized to evaluate the effect that these different parameters had on costs and health utility outcomes.

Model Inputs (Table I) Costs A summary of the values used in the reference case for the model and the ranges used in the probabilistic sensitivity analysis are given in Table I. The

major costs included in the model were associated with implant use and revision surgery. Implant costs for the base case were based on median 20 values in the published literature , which describe an increased cost with intramedullary nailing of approximately $1200 per case. However, a larger range of incremental costs ($900 to $1500) was used in the sensitivity analysis because of the large regional and institutional variations in device costs. Costs of revision surgery for femoral fracture or fixation failure were based on published reports. Prior data have shown that revision surgery due to failure of fixation would cost approximately $30,000 in today’s value of cur21,22 rency, or about twice the cost of the primary fracture surgery . This is consistent with published costs of revision hip arthroplasty, which have ranged 23-25 from $20,000 to $40,000 . Other costs associated with acute hospitalization and subsequent rehabilitation were assumed to be equal between the treatment groups. Although these costs are large, the literature has not shown a significant difference in length of stay, postoperative function, transfusion requirements, or operating room utilization between implant types, provided that the fracture healed 16,26,27 successfully .

1615 TH E JO U R NA L O F B O N E & JO I N T SU RG E RY J B J S . O RG V O LU M E 96 -A N U M B E R 19 O C T O B E R 1, 2 014 d

d

d

C O S T -E F F E C T I V E N E S S A N A LY S I S O F F I X AT I O N O P T I O N S FO R INTERTROC HANTERIC HIP FRACTU RES

TABLE I Input Values Used to Calculate the Cost-Effectiveness of Fixation Method for Intertrochanteric Hip Fractures Description Cost ($) Sliding hip screw Intramedullary nail Revision surgery for failure of fixation Perioperative femoral fracture rate (%) Sliding hip screw Intramedullary nail

Value

Range

Reference

2000 3200 30,000

1900-2100 2800-3600 20,000-40,000

20 20 21-24

0.0-0.2 0.0-0.5

17,37,40-45 17,36,37,40-45

0.1 0.25

Fixation failure rate (%) Sliding hip screw AO type A1 AO type A2 AO type A3

0 0 40.0

21.0-1.0 23.0-3.0 30.0-50.0

26,40,46,47 49,50,54-58 59-62

Intramedullary nail AO type A1 AO type A2 AO type A3

0 0 20.0

21.0-1.0 23.0-3.0 10.0-30.0

25,40,46 48,54 59,60

6

5.5-12

63,64

20.0 6.0

15.0-30.0 2.0-8.0

30,65-69 19,66

During initial fracture surgery/hospitalization

6.0

3.0-12.0

29,66

Due to revision surgery for failure of sliding hip screw or intramedullary nail

6.0

0.0-9.0

22,38,71

80 0.7 20.1

74-85 0.63-0.8 20.2-0

Expected time to loss of fixation (mo) Probability of death (%) All causes £1 yr after fracture >1 yr after fracture

Quality of life/utility Age at time of fracture After successful initial fixation of intertrochanteric fracture Utility loss after revision of failed fixation

Health-Related Quality of Life (Health Utility) Health-related quality of life was estimated with use of published data on 27-34 . It quality of life after successful fixation of intertrochanteric fractures ranges from 0.6 to 0.8, and we chose a health utility of 0.7 for our reference case. It was assumed that there was no difference in health utility based on fixation 6,16,35-37 type, provided the fracture healed successfully . For patients requiring revision surgery, there was an estimated disutility of 10%. This is based on literature reports that functional outcomes are roughly 38,39 22 equivalent or slightly worse in the revision setting .

Clinical Outcome Probabilities Historically, the perioperative femoral fracture rate was higher when an in16 tramedullary nail, as opposed to a sliding hip screw, was used . However, more recent investigations have shown that this complication rate has decreased 17 significantly with the most recent generation of intramedullary nail devices . In the most recent studies, the increased risk of femoral fracture from intra17,36,37,40-45 medullary nail use has ranged from 0% to 0.5% . For the purposes of our analysis, 0.25% was chosen to represent the femoral fracture rate with intramedullary nail use. Rates of secondary fixation failure were estimated on the basis of data from published trials and divided by AO fracture type. A1 fractures are usually considered stable, and multiple studies and meta-analyses have shown that both sliding hip screws and intramedullary nails have an absolute failure risk of

66-69 19,24,27-29,32-34 22,38,39 26,40,46,47

approximately 1% when used to fix these fractures . A 95% confidence interval of ±1.0% was used in our probabilistic analysis. The failure rates of A2 fractures, considered ‘‘possibly unstable,’’ are controversial. There have been few high-quality trials with dedicated analysis of 48-50 unstable fractures , as many studies have also included fractures with stable 42,51-53 patterns . The failure rate of A2 fractures after fixation with a sliding hip 54,55 screws has been cited to range from 3% (comparable with the rates seen 56-58 . In 2006 a meta-analysis, Jones et al., after intramedullary nailing) to 19% specifically looked at unstable patterns and found an overall cutout rate of 47 3.3% . They found no significant difference in incremental failure rate between intramedullary nails and sliding hip screws, with a total margin for error of about 6.0%. Thus, for our analysis, the fixation failure rate was set at 3.0% for both intramedullary nails and sliding hip screws, with a 95% confidence interval ranging from 0% to 6%. Finally, there are few randomized controlled trials that specifically assessed A3 fracture types, and in many studies supplemental fixation was used 50 along with sliding hip screws , probably because of the commonly held belief that A3 fractures are too unstable for isolated sliding hip screw fixation. The most recent studies that specifically evaluated failure rates of A3 fractures showed very high failure rates for sliding hip screws, ranging from 34% to 59,60 56% , and even extramedullary fixed-angle devices such as dynamic con61 dylar screws can have failure rates of up to 26% . Intramedullary nails used in 60-62 . These ranges similar studies had failure rates ranging from 5.0%), there is both an improvement in health utility and cost savings (i.e., intramedullary nailing dominates). When the ICER is positive, there is an improvement in health utility that comes with an associated increased cost. We used that value in our analysis, although it was allowed to vary from 10% to 30% in the probabilistic sensitivity analysis. Fixation failure was modeled to occur at an average of six months postoperatively, as this was the most com63,64 monly reported time to loss of fixation . Mortality rates during the initial hospitalization and the first postoperative year were derived from multiple reports. They ranged from 15% to 30,65-69 30% , and 20% was used for our reference case. This was assumed to be identical for both methods of fixation, consistent with the published litera16 ture . After the initial year postsurgery, mortality was assumed to return to 70 19 population-based norms and estimated with use of actuarial tables , resulting in an average life expectancy of eight years after the fracture. The mortality rate associated with revision surgery has generally been similar to that 22,38,71 associated with primary fracture surgery and was assumed to be identical in the reference case.

Analysis The model was created and analyzed with use of decision analysis software (TreeAge Pro 2012 v12.2; Williamstown, Massachusetts). Foldback analysis was then conducted to determine the expected costs and health state utility values for each branch of the decision tree. Sensitivity analysis was performed to determine the effect of variability in critical model parameters on the overall strategy and to confirm the validity of the model. Fixation failure rates, implant costs, revision surgery costs, mortality rates, and health utility estimates were all varied within the ranges reported in the literature in order to test the sensitivity of the results. The fixation failure rate for sliding hip screws was one of the critical model parameters and had a wide range of reported values; thus, a probabilistic statistical analysis was conducted on model outcome as a function of failure rate. A Monte Carlo sensitivity analysis was used to simulate a variety of different failure rates (with ranges determined by the fracture classifications specified above). Acceptability curves were generated to summarize results given the uncertainty of some of the parameter estimates used in the analy72,73 sis . These curves serve to visually demonstrate the probability of a given strategy being cost-effective, given an arbitrary cost-effectiveness ratio (expressed in U.S. dollars per QALY). Incremental net monetary benefit curves were also used to estimate outcomes in monetary terms as a function of a set

74

cost-effectiveness ratio . They were used to determine at what cost-effectiveness ratio each treatment strategy becomes beneficial on the whole and to visually quantify the magnitude of the financial benefit.

Source of Funding There was no external source of funding for this study

Results Reference Case ased on the reference case parameters depicted in Table I, intramedullary nail fixation was the dominant strategy—i.e., it provided both a direct cost savings as well as an improvement in health state utility—when the incremental failure rate of sliding hip screws rose above 5.0% (Fig. 3). When the incremental failure rate was 1.9%, the incremental cost-effectiveness ratio was still below $50,000/QALY, although that number quickly rose to $100,000/QALY when the failure rate for sliding hip screws was

Cost-effectiveness analysis of fixation options for intertrochanteric hip fractures.

Intertrochanteric hip fractures are a major source of morbidity and financial burden, accounting for 7% of osteoporotic fractures and costing nearly $...
998KB Sizes 0 Downloads 8 Views