Bisphosphonates reduce the risk of hip and vertebral fragility fractures associated with osteoporosis [3, 22, 23], but several studies have shown that they increase the risk of “atypical” fractures [37, 38]. Atypical femur fractures are transverse or short oblique, noncomminuted subtrochanteric or diaphyseal femur fractures that result from low-energy trauma in the setting of long-term antiresorptive therapy . Continuous bisphosphonate use for > 3 years has been implicated as a risk factor for atypical femur fractures and is thought to be related to osteoclast inhibition and suppression of bone turnover, resulting in hypermineralized, brittle bone [30, 36, 43]. Although the relative risks of atypical femur fracture can be high in patients with chronic antiresorptive use (range, 2.1-128) compared with patients without bisphosphonate exposure, their absolute risk is low (three to 50 cases per 100,000 person-years) . However, when these fractures do occur, they are associated with delayed healing compared with typical geriatric hip fractures .
In addition, patients sustaining a unilateral atypical femur fracture are at increased risk for subsequent contralateral fracture [20, 37]. Prophylactic fixation of an incomplete atypical femur fracture appears effective, whereas nonoperative treatment risks persistent pain and fracture completion (Fig. 1) [2, 10, 29]. However, there is little evidence beyond case series and expert opinion to guide the use of prophylactic fixation, although prophylaxis of the contralateral hip after unilateral slipped capital femoral epiphysis has been explored . We are unaware of any studies focusing on the cost-effectiveness of this intervention; the question of when to perform prophylactic fixation in patients with atypical femur fracture lends itself well to decision and cost-effectiveness analyses because treatment options are well defined (prophylactic surgery or no surgery) and outcomes are distinct .
Materials and Methods
We used Markov modeling to determine whether contralateral prophylactic femur fracture fixation is cost-effective after a bisphosphonate-associated atypical femur fracture and, if so, what patient-related factors may influence that determination. This method evaluates the relative costs and benefits of alternative clinical strategies over the course of a patient’s lifetime. To enter the model, patients had to present with an atypical femur fracture (Fig. 2). The default age for simulated patients was 70 years, and patients aged 60 to 90 years were assessed in the analysis because the majority of atypical femur fractures occur in this demographic [9, 37].
A Markov model has three core components: event probabilities, costs, and utilities. Our cycle length was 1 year, and costs are from the health payer perspective. The following describes the process for constructing each component of the Markov model.
Event probability is the likelihood that a clinical scenario will occur such as contralateral fracture risk and risk of perioperative complications from prophylactic or fracture fixation.
The age-adjusted likelihood of sustaining an atypical femur fracture was derived from the available evidence, and patients were categorized as standard versus high risk for fracture based on radiographic and demographic factors that predispose to fracture. Studies that followed the natural history of the contralateral side after unilateral fracture were utilized to determine baseline fracture risk, because these reflect patients entering our model who sustained an index atypical femur fracture [2, 14, 20, 24, 37, 40]. Our standard risk patient characteristics reflect the cohort examined in a prospective study by Dell et al. , which followed 142 patients with atypical femur fractures with a mean age of 69 years and body mass index of 27 kg/m2 of whom 90% had taken bisphosphonates for a mean duration of 6 years. Among these patients, 23% sustained a contralateral subtrochanteric fracture. In this study, prodromal thigh pain was noted in 69% of patients and many patients had diffuse cortical thickening of the lateral cortex on radiographs before fracture, factors known to be associated with the pathogenesis of the disease [9, 37]. Additionally, we defined patients with standard risk as those without symptoms or radiographic findings in the contralateral femur at presentation [9, 41].
Our high-risk scenario was based on the study by Lo et al. , who reported that among 38 women with atypical femur fractures, 15 developed stress or complete fractures in the contralateral femur. Compared with typical femur fractures in patients older than 60 years old, a contralateral fracture was more likely among younger patients and those of Asian descent . Further factors that may place patients at high risk for contralateral fracture include varus proximal femoral geometry, femoral bowing, prodromal thigh pain, and radiographic changes such as periosteal beaking, but the effect size has not been well defined [2, 12, 16, 19, 24, 25, 27, 28, 30, 35, 37]. Therefore, we defined our high-risk patients as those with more than one of the following risk factors: Asian descent, varus proximal femoral geometry, femoral bowing, prodromal pain, or radiographic changes such as periosteal beaking or a transverse radiolucent line in the contralateral femur .
The probability of contralateral fracture decreases with time from the index fracture with nearly all fractures occurring within 5 years. We therefore assumed that a negligible risk of fracture existed after 5 years [9, 14]. This is consistent with recommendations that patients discontinue bisphosphonates or initiate anabolic therapy such as teriparatide to mitigate fracture risk. We thus created a dynamic model whereby fracture risk varies with patient risk profile and time from the index fracture to establish the probabilities of fracture (Table 1).
Complication rates were primarily adapted from studies about fragility hip fractures, because studies on atypical femur fracture-specific complication rates are of limited size and generalizability [4, 13, 21, 29, 31-33]. Complications were defined according to the National Surgical Quality Improvement Program categorical groups as used in the hip fracture risk calculator developed by Pugely et al. . Prophylactic surgery was assumed to have a lower risk of complications than fracture fixation, and all complications were modeled to occur within 1 year of the initial fracture [2, 4, 13, 29]. Evidence input for complication rates is described (Table 1). We assumed patients would not undergo simultaneous fracture fixation and contralateral prophylaxis.
Costs were defined from the healthcare insurer (payer) perspective applied from the Medicare 2015 fee schedule and include the average national surgeon reimbursement for open treatment of per-/subtrochanteric femur fracture with an intramedullary nail (Current Procedural Terminology code [CPT] 27245), prophylactic treatment of the femur (CPT 27495), initial inpatient hospital care (CPT 99222), followup office visits (CPT 99203, 99213), and hospital payment for Diagnosis-Related Group codes 480, 481, or 482 (hip and femur procedures [nonjoint] with or without a minor or major complication) (Table 2) [7, 8]. Facility prices were used for CPT 27245, 27495, and 99222, whereas nonfacility fees were used for CPT 99203 and 99213. Cost ranges were determined using the ratio of single-facility maximum and minimum payments to the average national payment. Skilled nursing facility (SNF) stays and the costs of surgical complications were also modeled. The average SNF stay for a hip fracture was determined from the literature with the relative length of stay for prophylactic fixation and complications assessed by surgeon experience [6, 15, 17]. It was assumed that there were no new costs 1 year after fracture or prophylaxis. Costs were discounted at 3% per year .
Utility attributes a value to the state of health or disease. We measured utility as quality-adjusted life-years (QALYs) over the patient’s remaining lifetime discounted at 3% per year . We assessed the literature for patient surveys, which attributed utilities and disutilities to the following health states after a patient sustained a contralateral femur fracture: year of fracture, postprophylaxis with or without surgical complications, and death. In states except for death, QALYs and costs were determined from the literature (Table 1). We attributed a larger QALY loss to postfracture complications because these complications such as nonunion are generally more morbid than postprophylaxis complications [2, 40].
QALY reductions for health states were rooted in studies that assess the impact of hip fractures on quality of life [1, 12, 42]. These publications used validated tools to assess patient mobility, self-care, activities of daily living, pain, and emotional well-being. We applied these findings with clinical judgment to reach baseline QALY changes, which were then subjected to sensitivity analysis.
We used all-cause annual US mortality tables to model expected death rates . During the year after either prophylaxis or contralateral atypical femur fracture, we assumed an increased mortality rate if there were complications. Expected mortality was increased by a factor of 2.5 in the year after prophylaxis contralateral fracture and then returned to baseline. Increased mortality rates after surgical complications were extracted from a prospective observational cohort study of > 2000 elderly patients presenting with hip fractures .
Given the inherent uncertainty in probability, cost, and utility estimates, we performed univariate, deterministic sensitivity analyses across all model inputs to evaluate how results changed with variation in parameter inputs with low and high ranges determined from published studies (Table 1). The incremental cost-effectiveness ratio (ICER) was calculated by dividing the difference in costs from the payer perspective using Medicare reimbursements by the difference in QALYs between the two treatment arms. We interpreted changes in policy that lowered cost while increasing QALYs as cost-saving, ICER of < USD 50,000 per QALY as definitely cost-effective, ICER between USD 50,000 and USD 100,000 per QALY as likely cost-effective, and ICER greater than USD 100,000 or loss in QALYs as not cost-effective . Additionally, two-way sensitivity analysis was performed with the three most sensitive model parameters over the ranges reported elsewhere for high-risk patients [2, 4, 9, 13, 29, 37]. These analyses demonstrate how simultaneous varying of two model inputs will affect the ICER.
For patients who sustain an atypical femur fracture with standard risk for contralateral fracture at age 70 years, prophylactic fixation of the contralateral femur is not cost-effective at a willingness to pay (WTP) of USD 100,000/QALY and is definitely cost-effective for patients aged 70 years with a high fracture risk. Performing prophylaxis for standard-risk patients resulted in slightly greater lifetime QALYs (11.3 versus 11.2 for no prophylaxis) at a higher lifetime cost (USD 11,500 versus USD 3600 for no prophylaxis), resulting in an ICER of USD 131,300/QALY. For patients deemed at high risk of contralateral atypical femur fracture, prophylaxis resulted in slightly greater lifetime QALYs (11.3 versus 11.1 for no prophylaxis) at higher lifetime cost (USD 11,500 versus USD 6400 for no prophylaxis), resulting in an ICER of USD 22,400/QALY. In both standard-risk and high-risk patients, the prophylactic strategy is associated with higher lifetime costs resulting from the cost of performing prophylactic surgery, but also higher lifetime QALYs driven by a reduction in disutility from fewer long-term complications and poor outcomes from surgeries for atypical femur fracture as well as prevention of disutility caused by the contralateral atypical femur fracture in the year of the injury.
Prophylactic fixation for patients with standard risk is likely cost-effective only when patients are between 60 and 66 years of age, whereas prophylaxis is definitely cost-effective for high-risk patients between 60 and 84 years of age and likely cost-effective between 84 and 89 years of age (Fig. 3). We define patients at high fracture risk as those with more than one risk factor including Asian ethnicity, prodromal pain, femoral bowing, varus proximal geometry, or radiographic changes in the contralateral femur such as periosteal beaking or a transverse radiolucent line because these patients are more likely to have a contralateral fracture if no intervention is performed [20, 37, 41]. Prophylactic surgery is more cost-effective among patients on the younger end of our studied population (60-90 years old) because older patients have a lower life expectancy.
Sensitivity analysis was performed across all inputs in the model based on the ranges described (Table 1).
ICER values are most affected by changes to cost of prophylaxis treatment, patient age, and probability of prophylaxis complication, whereas changes to probability of fracture within 5 years affect ICER values to a lesser extent (Fig. 4). Prophylaxis for a 70-year-old patient with high risk is likely cost-effective at a WTP threshold of USD 100,000/QALY when the cost of prophylaxis is less than USD 29,400, the probability of prophylaxis complication is < 21%, or if the patient is younger than 89 years old and within ranges tested in all other model parameters.
Two-way sensitivity analysis revealed that with zero probability of a prophylaxis complication for high-risk patients, the cost of prophylaxis could be as high as USD 26,300 and prophylactic surgery would still be definitely cost-effective, whereas at a 22% probability of a prophylaxis complication, the cost of prophylaxis surgery must be less than USD 7420 to be definitely cost-effective (Fig. 5A). Additionally, as the probability of prophylaxis complications increases, the cost of prophylaxis must decrease to maintain cost-effectiveness. At zero probability of a prophylaxis complication for high-risk patients, prophylaxis would be considered definitely cost-effective for patients at any age between 60 and 90 years, but at a 22% probability of prophylaxis complications, prophylaxis only for patients between 60 and 66 years would be considered likely cost-effective (Fig. 5B). These findings further suggest that a low probability of prophylaxis complications can offset the effects of increased patient age and cost of prophylactic complications on cost-effectiveness.
Patients who have one atypical femur fracture have been found to be at increased risk of contralateral atypical femur fracture [20, 37]. Orthopaedic surgeons who are presented with these patients have to mitigate the risk of a contralateral fracture. No clear, validated guidelines exist to support decision-making . We used Markov modeling to assess whether prophylactically fixing the contralateral femur in these patients would be cost-effective and, if so, in which patient demographics. We chose Markov modeling to assess simulated patients with atypical femur fracture because these patients have distinct health states with defined costs, utilities, and transition probabilities and because Markov modeling is a well-accepted approach for cost-effectiveness analysis . At a WTP threshold of USD 100,000/QALY, this study indicates prophylactic fixation of the contralateral side after unilateral atypical femur fracture is not cost-effective among patients older than 66 years old with a standard risk of fracture and is cost-effective for patients between 60 and 89 years of age with a high risk of fracture. Sensitivity analysis revealed that our model was quite robust, meaning our results did not change with broad variation in the assumed risks, costs, and utilities. Our findings are therefore generalizable to patients, providers, and payers faced with the decision of whether to prophylactically treat the contralateral hip after atypical femur fracture.
The study has several limitations. First, there are no randomized clinical trials assessing prophylactic management of the contralateral femur in patients with atypical femur fracture, and the relatively low absolute risk of an atypical femur fracture means that there is a limited body of literature from which model inputs could be extracted . Atypical femur fracture-specific fracture risks, complication rates, and costs were used where available, but were based on small prospective and retrospective studies of limited generalizability. Some values for both prophylactic and traumatic femur fixation complications and health state utilities were estimated from studies on fragility hip fractures and are not specific to the atypical femur fracture population, because these values were not found in the literature [4, 12, 13, 21, 33]. The values for the model inputs were not determined through a systematic review, although we attempted to use data from the largest studies that most closely matched the patient demographics being assessed. Costs were determined from national average Medicare reimbursement rates, which may not be representative of individual institutional costs. Additionally, the length of SNF stays was partially determined by expert opinion, and average daily cost of a SNF stay was based on the national average Medicare reimbursement rate . We selected a WTP threshold of USD 100,000/QALY, which is an arbitrary but commonly reported value in the literature . Another weakness is the creation of a single high-risk fracture group that includes multiple known risk factors. We did not model the specific contribution of every known risk factor, because the effect sizes are unclear. Studies have indicated that prodromal hip pain, radiographic changes such as periosteal beaking, varus proximal femoral geometry, Asian ethnicity, and curvature of the femoral diaphysis may increase fracture risk [16, 19, 25, 35, 37, 38]. Finally, no pharmacologic or nonoperative modalities for decreasing fracture risk were tested. Fracture risk may be decreased by stopping antiresorptive medications, restricted weightbearing, and medical management with calcium, vitamin D, or teriparatide, and surgical treatment is currently advised if pain persists after 2 to 3 months of nonoperative treatment [5, 16, 37].
In our model, prophylactic fixation of the contralateral femur was not a cost-effective strategy for patients with standard fracture risk, but was cost-effective for patients at high risk for atypical femur fractures associated with bisphosphonate use. A systematic review performed by Toro et al. to create a decision-making algorithm for the management of patients who present with a complete atypical femur fracture and an incomplete contralateral fracture suggested that prophylactic fixation of the contralateral fracture should be the gold standard to allow for early weightbearing and prevent progression of the fracture . These patients share many characteristics with patients we define as high risk such as prodromal pain, varus geometry, and radiographic changes, and many of these findings are highlighted as minor features for diagnosing an atypical femur fracture in the Second ASBMR Report on Atypical Femur Fractures . Our novel findings corroborate these conclusions with cost-effectiveness data supporting prophylactic fixation in high-risk patients.
In our analysis, we identified several variables that influenced the cost-effectiveness of prophylactic fixation. Changes to the cost of prophylaxis treatment, probability of prophylaxis complications, and cost of fracture treatment were among the variables that generated the largest variation in the ICER, as may be expected (Fig. 4). Additionally, patient age of presentation significantly affected the ICER with patients closer to 60 years of age benefiting more from prophylaxis compared with older patients, which is consistent with a previously published cost-effectiveness analysis that found that prophylactic fixation of the contralateral hip after unilateral fragility hip fracture could be cost-effective in women younger than 70 years or between 71 and 75 years with a 30% greater relative risk for contralateral fracture . We found the probability of fracture within 5 years, the QALY reduction associated for each year after fracture complication, and the probability of fracture complication affected the ICER as well but to a lesser extent.
After sustaining a bisphosphonate-associated atypical femur fracture, the patient and clinician must decide between observation and prophylactic fixation of the contralateral side. This decision also has important financial implications for the healthcare system. The principal advantage of observation is avoiding the morbidity and cost of additional surgery; the principal advantage of prophylaxis is avoiding the unpredictability, acute disability, and potential complications of a subsequent contralateral atypical femur fracture, which is associated with greater pain, prolonged hospitalization, urgent surgical intervention, and delayed healing. Our findings suggest that prophylactic surgical intervention should be considered a cost-effective intervention for patients at high risk of sustaining a contralateral fracture after presenting with unilateral atypical femur fracture, specifically those who have more than one risk factor such as Asian ethnicity, prodromal pain, femoral geometry changes, or radiographic findings on the contralateral femur. A better understanding of the degree of risk contributed by known radiographic and demographic parameters would be beneficial to guide management, which could be accomplished with a large cohort study that prospectively documents known risk factors for atypical femur fractures.
1. Arias E, Heron M, Xu J. United States Life Tables, 2012.
Hyattsville, MD, USA: National Vital Statistics Reports, Centers for Disease Control and Prevention; 2016;65.
2. Banffy MB, Vrahas MS, Ready JE, Abraham JA. Nonoperative versus prophylactic treatment of bisphosphonate-associated femoral stress fractures. Clin Orthop Relat Res. 2011;469:2028-2034.
3. Beaupre LA, Morrish DW, Hanley DA, Maksymowych WP, Bell NR, Juby AG, Majumdar SR. Oral bisphosphonates are associated with reduced mortality after hip fracture. Osteoporos Int. 2011;22:983-991.
4. Bellabarba C, Herscovici D Jr, Ricci WM. Percutaneous treatment of peritrochanteric fractures using the Gamma nail. Clin Orthop Relat Res. 2000;375:30-42.
5. Black DM, Bauer DC, Schwartz AV, Cummings SR, Rosen CJ. Continuing bisphosphonate treatment for osteoporosis--for whom and for how long? N Engl J Med. 2012;366:2051-2053.
6. Bozic KJ, Ward L, Vail TP, Maze M. Bundled payments in total joint arthroplasty: targeting opportunities for quality improvement and cost reduction. Clin Orthop Relat Res.
7. CMS. National Physician Fee Schedule Relative Value File2015B. Available at: https://www.cms.gov/apps/physician-fee-schedule/search/search-criteria.aspx
. Accessed July 26, 2018.
9. Dell RM, Adams AL, Greene DF, Funahashi TT, Silverman SL, Eisemon EO, Zhou H, Burchette RJ, Ott SM. Incidence of atypical nontraumatic diaphyseal fractures of the femur. J Bone Miner Res. 2012;27:2544-2550.
10. Egol KA, Park JH, Prensky C, Rosenberg ZS, Peck V, Tejwani NC. Surgical treatment improves clinical and functional outcomes for patients who sustain incomplete bisphosphonate-related femur fractures. J Orthop Trauma. 2013;27:331-335.
11. Egol KA, Park JH, Rosenberg ZS, Peck V, Tejwani NC. Healing delayed but generally reliable after bisphosphonate-associated complete femur fractures treated with IM nails. Clin Orthop Relat Res. 2014;472:2728-2734.
12. Ekstrom W, Nemeth G, Samnegard E, Dalen N, Tidermark J. Quality of life after a subtrochanteric fracture: a prospective cohort study on 87 elderly patients. Injury. 2009;40:371-376.
13. Faucett SC, Genuario JW, Tosteson AN, Koval KJ. Is prophylactic fixation a cost-effective method to prevent a future contralateral fragility hip fracture? J Orthop Trauma. 2010;24:65-74.
14. Giusti A, Hamdy NA, Papapoulos SE. Atypical fractures of the femur and bisphosphonate therapy: a systematic review of case/case series studies. Bone. 2010;47:169-180.
15. Grabowski DC. Post-acute and Long-term Care: A Primer on Services, Expenditures and Payment Methods.
Washington, DC, USA: US Department of Health and Human Services, Office of Disability, Aging and Long-Term Care Policy; 2010.
16. Hagen JE, Miller AN, Ott SM, Gardner M, Morshed S, Jeray K, Alton TB, Ren D, Abblitt WP, Krieg JC. Association of atypical femoral fractures with bisphosphonate use by patients with varus hip geometry. J Bone Joint Surg Am. 2014;96:1905-1909.
17. Harada ND, Chun A, Chiu V, Pakalniskis A. Patterns of rehabilitation utilization after hip fracture in acute hospitals and skilled nursing facilities. Med Care. 2000;38:1119-1130.
18. Kocher MS, Bishop JA, Hresko MT, Millis MB, Kim YJ, Kasser JR. Prophylactic pinning of the contralateral hip after unilateral slipped capital femoral epiphysis. J Bone Joint Surg Am. 2004;86:2658-2665.
19. Kwek EB, Goh SK, Koh JS, Png MA, Howe TS. An emerging pattern of subtrochanteric stress fractures: a long-term complication of alendronate therapy? Injury. 2008;39:224-231.
20. Lo JC, Huang SY, Lee GA, Khandelwal S, Provus J, Ettinger B, Gonzalez JR, Hui RL, Grimsrud CD. Clinical correlates of atypical femoral fracture. Bone. 2012;51:181-184.
21. Lovy AJ, Kim JS, Di Capua J, Somani S, Shim S, Keswani A, Hasija R, Wu Y, Joseph D, Ghillani R. Intramedullary nail fixation of atypical femur fractures with bone marrow aspirate concentrate leads to faster union: a case-control study. J Orthop Trauma. 2017;31:358-362.
22. Lozano-Calderon SA, Colman MW, Raskin KA, Hornicek FJ, Gebhardt M. Use of bisphosphonates in orthopedic surgery: pearls and pitfalls. Orthop Clin North Am. 2014;45:403-416.
23. Lyles KW, Colon-Emeric CS, Magaziner JS, Adachi JD, Pieper CF, Mautalen C, Hyldstrup L, Recknor C, Nordsletten L, Moore KA, Lavecchia C, Zhang J, Mesenbrink P, Hodgson PK, Abrams K, Orloff JJ, Horowitz Z, Eriksen EF, Boonen S. Zoledronic acid and clinical fractures and mortality after hip fracture. N Engl J Med. 2007;357:1799-1809.
24. Meier RP, Perneger TV, Stern R, Rizzoli R, Peter RE. Increasing occurrence of atypical femoral fractures associated with bisphosphonate use. Arch Intern Med. 2012;172:930-936.
25. Murphy CG, O'Flanagan S, Keogh P, Kenny P. Subtrochanteric stress fractures in patients on oral bisphosphonate therapy: an emerging problem. Acta Orthop Belg. 2011;77:632-637.
26. Neumann PJ, Cohen JT, Weinstein MC. Updating cost-effectiveness--the curious resilience of the USD 50,000-per-QALY threshold. N Engl J Med. 2014;371:796-797.
27. Neviaser AS, Lane JM, Lenart BA, Edobor-Osula F, Lorich DG. Low-energy femoral shaft fractures associated with alendronate use. J Orthop Trauma. 2008;22:346-350.
28. Nieves JW, Cosman F. Atypical subtrochanteric and femoral shaft fractures and possible association with bisphosphonates. Curr Osteoporos Rep. 2010;8:34-39.
29. Oh CW, Oh JK, Park KC, Kim JW, Yoon YC. Prophylactic nailing of incomplete atypical femoral fractures. ScientificWorldJournal. 2013;2013:450148.
30. Park-Wyllie LY, Mamdani MM, Juurlink DN, Hawker GA, Gunraj N, Austin PC, Whelan DB, Weiler PJ, Laupacis A. Bisphosphonate use and the risk of subtrochanteric or femoral shaft fractures in older women. JAMA. 2011;305:783-789.
31. Parker MJ, Myles JW, Anand JK, Drewett R. Cost-benefit analysis of hip fracture treatment. J Bone Joint Surg Br. 1992;74:261-264.
32. Pearse EO, Redfern DJ, Sinha M, Edge AJ. Outcome following a second hip fracture. Injury. 2003;34:518-521.
33. Pugely AJ, Martin CT, Gao Y, Klocke NF, Callaghan JJ, Marsh JL. A risk calculator for short-term morbidity and mortality after hip fracture surgery. J Orthop Trauma. 2014;28:63-69.
34. Roche JJ, Wenn RT, Sahota O, Moran CG. Effect of comorbidities and postoperative complications on mortality after hip fracture in elderly people: prospective observational cohort study. BMJ. 2005;331:1374.
35. Sasaki S, Miyakoshi N, Hongo M, Kasukawa Y, Shimada Y. Low-energy diaphyseal femoral fractures associated with bisphosphonate use and severe curved femur: a case series. J Bone Miner Metab. 2012;30:561-567.
36. Schilcher J, Michaelsson K, Aspenberg P. Bisphosphonate use and atypical fractures of the femoral shaft. N Engl J Med. 2011;364:1728-1737.
37. Shane E, Burr D, Abrahamsen B, Adler RA, Brown TD, Cheung AM, Cosman F, Curtis JR, Dell R, Dempster DW, Ebeling PR, Einhorn TA, Genant HK, Geusens P, Klaushofer K, Lane JM, McKiernan F, McKinney R, Ng A, Nieves J, O'Keefe R, Papapoulos S, Howe TS, van der Meulen MC, Weinstein RS, Whyte MP. Atypical subtrochanteric and diaphyseal femoral fractures: second report of a task force of the American Society for Bone and Mineral Research. J Bone Miner Res. 2014;29:1-23.
38. Shane E, Burr D, Ebeling PR, Abrahamsen B, Adler RA, Brown TD, Cheung AM, Cosman F, Curtis JR, Dell R, Dempster D, Einhorn TA, Genant HK, Geusens P, Klaushofer K, Koval K, Lane JM, McKiernan F, McKinney R, Ng A, Nieves J, O'Keefe R, Papapoulos S, Sen HT, van der Meulen MC, Weinstein RS, Whyte M. Atypical subtrochanteric and diaphyseal femoral fractures: report of a task force of the American Society for Bone and Mineral Research. J Bone Miner Res. 2010;25:2267-2294.
39. Sonnenberg FA, Beck JR. Markov models in medical decision making: a practical guide. Med Decis Making. 1993;13:322-338.
40. Swart E, Makhni EC, Macaulay W, Rosenwasser MP, Bozic KJ. Cost-effectiveness analysis of fixation options for intertrochanteric hip fractures. J Bone Joint Surg Am. 2014;96:1612-1620.
41. Toro G, Ojeda-Thies C, Calabrò G, Moretti A, Guerra GMD, Caba-Doussoux P, Iolascon G. Management of atypical femoral fracture: a scoping review and comprehensive algorithm. BMC Musculoskelet Disord. 2016;17:227. 2016 May 23. [Epub ahead of print]
42. Tosteson AN, Gabriel SE, Grove MR, Moncur MM, Kneeland TS, Melton LJ 3rd. Impact of hip and vertebral fractures on quality-adjusted life years. Osteoporos Int. 2001;12:1042-1049.
43. Wysowski DK, Greene P. Trends in osteoporosis treatment with oral and intravenous bisphosphonates in the United States, 2002-2012. Bone. 2013;57:423-428.
44. Zilberberg MD, Shorr AF. Understanding cost-effectiveness. Clin Microbiol Infect. 2010;16:1707-1712.