Hemiarthroplasty vs Primary Total Hip Arthroplasty For Displaced Fractures of the Femoral Neck in the Elderly

The Journal of Arthroplasty Vol. 27 No. 4 2012

Hemiarthroplasty vs Primary Total Hip Arthroplasty For Displaced Fractures of the Femoral Neck in the Elderly A Meta-Analysis Ai Zi-Sheng, PhD,* Gao You-Shui, PhD, MD,y Jing Zhi-Zhen, PhD, MD,z Yuan Ting, MD,y and Zhang Chang-Qing, PhD, MDy

Abstract: Current updated meta-analysis was designed to compare clinical effects of hemiarthroplasty (HA) vs primary total hip arthroplasty (THA) for displaced femoral neck fractures in elderly patients. Five randomized and 4 quasi-randomized controlled trials with a total 1208 patients were included for final analysis. It showed that mortality and postoperative infection between HA and THA had no statistical differences, that long-term reoperation rate of HA was higher than that of THA, that medium-term dislocation rate of HA was lower than that of THA, and that pain rates of HA in short-term and long-term were both higher than THA. Summarily, treatment of THA for elderly displaced femoral neck fracture could provide better results of reduced reoperation rate and pain relief; however, HA yielded a lower incidence of postoperative dislocation. Keywords: femoral neck fractures, total hip arthroplasty, hemiarthroplasty, the elderly, meta-analysis. © 2012 Elsevier Inc. All rights reserved.

Because of the growing aging population and high prevalence of osteoporosis, the incidence of femoral neck fractures shows a rising trend. In North America, for example, there are 280 000 hip fractures that require treatment every year [1], and by 2050, this number is expected to rise to 700 000 cases annually. It has been estimated that the annual medical costs owing to hip fractures will be more than 15 billion US dollars at this time [2]. Hemiarthroplasty (HA) and total hip arthroplasty (THA) are common methods for hip joint restoration [3]. Hemiarthroplasty and THA treatments for displaced femoral neck fractures in elderly patients both present advantages and disadvantages [4-6]. For displaced intra-

capsular fractures in elderly patients, if they have lower functional demands, there seems to be a consensus that HA is the preferred treatment [1]; however, for active, mentally alert, and relatively healthy elderly patients with displaced femoral neck fractures, the decision regarding whether to select HA or THA is controversial [1,7,8]. To provide a useful quantitative assessment regarding outcomes of HA or primary THA for displaced femoral neck fractures in patients older than 60 years, we performed a meta-analysis based on randomized and quasi-randomized controlled trials studies published up to December 2010.

Methods From the *Department of Preventive Medicine, College of Medicine, Tongji University, Shanghai, China; yDepartment of Orthopaedics, Shanghai Sixth People's Hospital, Shanghai Jiao Tong University, Shanghai, China; and zDepartment of Orthopaedics, Shanxi Provincial People's Hospital, Taiyuan, Shanxi Province, China. Submitted March 30, 2011; accepted July 28, 2011. The Conflict of Interest statement associated with this article can be found at doi:10.1016/j.arth.2011.07.009. Reprint requests: Zhang Chang-Qing, PhD, MD, Department of Orthopaedics, Shanghai Sixth People's Hospital, Shanghai Jiao Tong University, Shanghai 200233, China. © 2012 Elsevier Inc. All rights reserved. 0883-5403/2704-0015$36.00/0 doi:10.1016/j.arth.2011.07.009

Inclusion Criteria Only studies meeting the following criteria were included in this meta-analysis: (1) randomized controlled trials or quasi-randomized controlled trials comparing HA with THA, (2) patients with displaced fracture of the femoral neck (Garden stage III or IV), and (3) reported clinical outcomes. All patients in these studies were having their first HA or arthroplasty surgery. Exclusion Criteria Patients with the following conditions were excluded from the study: (1) undisplaced femoral neck fracture,

583

584 The Journal of Arthroplasty Vol. 27 No. 4 April 2012 (2) pathological fracture secondary to malignant disease, and (3) osteoarthritis or rheumatoid arthritis of the hip. Search Strategy ZSA and YSG independently performed a systemic search of the English medical literature of HA and primary THA for treatment of displaced femoral neck fractures in the elderly published between January 1966 and December 2010. Sources include the Cochrane database of randomized trials and PubMed. The search strategy was based on combinations of “femoral neck fracture,” “hemiarthroplasty,” and “total hip arthroplasty or total hip replacement.” Citations that included the key terms in the title, abstract, article, or medical subjects heading terms were retained. In addition, these searches were supplemented with manual searches of references within the published articles, major orthopedic textbooks, and principal journals. When necessary, the study authors were contacted for further information. Data Abstraction and Assessment of Methodological Quality Two of the authors (ZSA and YSG) independently extracted relevant data in duplicate, including fracture classification, mean age, sex distribution, prefracture comorbidity, cognitive function and mobility, average duration of the follow-up, and type of prosthesis. Two reviewers (ZSA and YSG) independently assessed the methodology of the selected articles according to method by Juni et al [9]. Quality criteria included treatment assignment, concealment of allocation, description of entry criteria, adherence to the intention-of-treatment principle, blinding, and handling of withdrawals. The reviewers resolved disagreements by discussion. Outcome Measures The outcome measures of this study included mortality rate at different follow-up times, pain, and main complications, which included infection (superficial and deep), dislocation, and reoperation. Statistical Analysis Meta-analysis was done with Review manager 4.2 (version 4.2 for Windows XP; Copenhagen: the Nordic Cochrane Centre, the Cochrane Collaboration, 2003). For categorical variable data, relative ratio (RR) and 95% confidence intervals (CIs) were calculated. Relative ratio is the risk of an event (or developing a disease) relative to exposure, which is the ratio of probability of the event development in the exposed vs a nonexposed group. For indicators of heterogeneity of pooled effect sizes, we calculated I 2, which indicates the heterogeneity in percentages, and we tested whether the level of heterogeneity was significant using the Q statistic. If the hypothesis of homogeneity was not rejected, the fixedeffected model (Mantel-Haenszel test) [10] was used to calculate the summary RR and the 95% CI. In all other

cases, a random-effects model (DerSimonian-Laird method) was used [11].

Results Study Characteristics After the inclusion and exclusion criteria outlined above, 2 reviewers (ZSA and YSG) independently read the title and/or abstract combined with manual search and references of retrieved articles and included 9 articles in this meta-analysis. All literatures involved THA vs HA surgeries for displaced femoral neck fractures in elderly patients, and the main outcome measures were postoperative complications and function recovery. The sample size of the data ranged from 40 to 252 cases with an average age of 69 to 81years and a follow-up time of 12 to 156 months. Both the THA and HA groups used were either cemented or uncemented. Detailed characteristics of the 9 studies are listed in Table 1. Based on the 9 studies, 1208 patients were available for analysis [12-20]. Of these, 561 patients received THA and 647 patients received HA. All 9 studies were randomized or quasi-randomized, and 5 studies had adequate randomization procedures [15-18,20]; the other 4 studies [12-14,19] were randomized based on odd and even hospital number (1), the day of the week (2), or fixed alternating sequence (1). Of the 9 studies, 5 studies [15-18,20] had concealment of allocation, whereas the other 4 studies [12-14,19] did not; 5 studies [15-18,20] had intent-to-treat analysis, whereas the remaining 4 studies [12-14,19] did not; 4 studies [12-14,19] had a loss of follow-up, whereas the other 5 studies [15-18,20] had complete follow-up; 2 studies [15,19] had a blinded outcome assessor, whereas in the remaining 7 studies [12-14,16-18,20], the outcome assessor was not blinded. Each study had specified entry criteria and defined outcome measures, and the detailed methodological qualities are listed in Table 2. As the follow-up time in each trial was different, the 9 trials were divided according to length of follow-up: within 1 year, within 5 years, and within 13 years (3 subgroups). Meta-analysis was carried out on each subgroup, and a comprehensive meta-analysis was also performed. Outcome Measures Mortality Mortality data were provided in 8 studies (1168 patients, 443 events): the number of events was 209 for HA (n = 624) and 194 for THA (n = 544). Effects sizes were homogeneous (χ 2 = 12.66, I 2 = 44.7%, P = .08). There were reductions in mortality rates within 5 years in patients who were treated with HA (RR, 0.88; 95% CI, 0.71-1.08; P = .23); however, these were not statistically significant. There was less mortality 1-year and 13-year postsurgery in patients with THA (RR, 1.14; 95% CI, 0.71-1.85; P = .58; RR, 1.06; 95% CI, 0.93-1.21; P = .39, respectively), but these results were not

HA vs THA For Elderly Femoral Neck Fractures  Zi-Sheng et al

585

Table 1. Characteristics of the Included Studies Study

Average Age (y)

FollowUp (mo)

Dorr et al [12] Skinner et al [13] Ravikumar and Marsh [14] Keating et al [15] Baker et al [16] Blomfeldt et al [17] Macaulay et al [18] Mouzopoulos et al [19] van der Bekerom et al [20]

69 81 81 75 75 81 79 74 81

48 12 156 24 36 12 24 48 60

Intervention

No. of Patients

THA

HA

THA

HA

Cemented Cemented Cemented Cemented Cemented Cemented Cemented or uncemented Cemented Cemented

Cemented or uncemented Uncemented Uncemented Cemented Cemented Cemented Cemented or uncemented Not specified Cemented

39 89 89 69 40 60 17 43 115

50 91 91 111 41 60 23 43 137

patients occurred in 27 (4.5%) of 604 hips compared with 89 (17.2%) of 518 who were treated with THA. Effect sizes were homogeneous (χ 2 = 7.3, I 2 = 17.9%, P = .29). There was a higher dislocation rate in the THA group in all 3 of the subgroups. These were not significant at 1 year (RR, 0.66; 95% CI, 0.32-1.37; P = .26) or at 13 years (RR, 0.65; 95% CI, 0.33-1.27; P = .21). However, within 5 years, 2 studies showed significance; van der Bekerom et al [20] showed greater dislocation rates in THA, and this difference was significant within 5 years (RR, 0.22; 95% CI, 0080.57; P = .02). The pooled analysis (Fig. 3) showed significant difference in the risk for dislocation between HA and THA; the pooled relative risk was 0.49 (95% CI, 0.32-0.75; P = .001). Overall, THA was associated with a greater risk of dislocation compared with primary HA.

statistically significant. Overall, mortality did not differ between the HA and THA groups (RR, 0.97; 95% CI, 0.85-1.11; P = .70) (Fig. 1). Reoperation Rates Reoperation rates were reported in 7 studies (1048 patients, 83 events). Pooled data presented 564 HAs and 484 THAs (the event of reoperation were 61 and 22 cases, respectively). Only Skinner et al [13] reported a greater likelihood of reoperation rates at 1 year occurring in the HA group (RR, 2.93; 95% CI, 0.98-8.75; P = .05). Five studies reported the reoperation rates within 5 years. There were 4 reports of higher reoperation rates in the HA group within 5 years [12,15,16,19], and 1 study showed a reduction in reoperation rates in the HA group, but this difference was not significant within 5 years (RR, 1.77; 95% CI, 0.93-3.36; P = .08). After 13 years of follow-up, there were significantly higher reoperation rates in the HA group (RR, 2.43; 95% CI, 1.53-3.84; P = .003) compared with the THA group. The pooled relative risk of reoperation rates after HA compared with THA was 2.43 (95% CI, 1.53-3.84; P = .0002). Fig. 2 shows that HA was associated with a greater risk of subsequent reoperation compared with THA.

Pain Rates Pain rates were reported in 2 studies (360 patients, 67 events). The pain rate was higher in patients treated with HA (62/182 [34.1%]) than those treated with THA (5/178 [2.8%]) (P b .001). Effect sizes were homogeneous (χ 2 = 1.40, I 2 = 28.7%, P = .24). The pain rates in short-term and long-term follow-up in the HA group were higher than that in the THA group (RR, 42.07; 95% CI, 2.59-684.01; P = .009; RR, 8.02; 95% CI, 3.32-

Dislocation Rates The 8 studies, totaling 1122 patients and 78 events, provided data on dislocation rates. Dislocation in HA

Table 2. Methodological Quality of the Included Studies

Study Dorr et al [12] Skinner et al [13] Ravikumar and Marsh [14] Keating et al [15] Baker et al [16] Blomfeldt et al [17] Macaulay et al [18] Mouzopoulos et al [19] van der Bekerom et al [20]

Random Method Odd and even hospital no. The day of week The day of week Randomized, computerized telephone service Randomized, sealed envelops Randomized, sealed envelops Randomized, opaque sealed envelops Fixed alternating sequence Computer randomization program

ITT indicates intention to treat.

Concealment of Allocation

ITT Analysis

Loss of Follow-up

Entry Criteria Specified

Outcome Measures Defined

Outcome Assessor Blinded

No No No Yes

No No No Yes

Yes No No Yes

Yes Yes Yes Yes

Yes Yes Yes Yes

No No No Yes

Yes Yes Yes No Yes

No Yes Yes Yes Yes

Yes No No Yes No

Yes Yes Yes Yes Yes

Yes Yes Yes Yes Yes

No No No Yes No

586 The Journal of Arthroplasty Vol. 27 No. 4 April 2012

Fig. 1. Relative risk and 95% CI for mortality.

19.36; P b .00001, respectively). The pooled analysis (Fig. 4) revealed that HA had significantly higher pain rates compared with THA (RR, 11.11; 95% CI, 4.8025.76; P b .00001). Infection Rates Six studies (690 patients, 27 events) provided infection rate data; infection occurred in 15 (4%) of 376 hips in

HA groups compared with 12 (3.8%) of 314 hips in THA groups. There was no effect size heterogeneity (χ 2 = 2.98, I 2 = 0%, P = .56). Fig. 5 shows that patients who underwent THA experienced higher infection rates within 1 year (RR, 0.92; 95% CI, 0.21-4.04; P = .91) and within 5 years (RR, 0.60; 95% CI, 0.18-1.99; P = .41); however, at 13 years, HA patients had a greater likelihood of infection compared with the THA group

Fig. 2. Relative risk and 95% CI for reoperation rates.

HA vs THA For Elderly Femoral Neck Fractures  Zi-Sheng et al

587

Fig. 3. Relative risk and 95% CI for dislocation rates.

(RR, 2.28; 95% CI, 0.61-8.55; P = .22). There were no statistical differences between THA and HA patients for any of the different time intervals. By interpreting all the data, it is apparent that infection rates did not differ between patients who underwent HA or THA (RR, 1.07; 95% CI, 0.51-2.22; P = .86) (Fig. 5).

Discussion The number and incidence of hip fractures increase exponentially with age in both sexes [8]. This rise will escalate, in part, because of the increase in the world's elderly population. Considering the enormous burden of this injury [8], orthopedists must improve clinical results for femoral neck fracture. There are 2

types of femoral neck fractures: displaced and undisplaced. The recommended treatment of undisplaced femoral neck fractures is closed/open reduction and internal fixation, irrespective of age [1]. Displaced intracapsular femoral neck fractures are very common injuries in elderly patients [21], and the treatments for displaced femoral neck fractures in this population commonly consist of reduction and internal fixation, HA, and THA [7,17,22]. There is no consensus on which treatment (HA or THA) has a better surgical outcome for displaced femoral neck fractures in the elderly [23]. Both THA and HA report advantages and disadvantages. When compared with HA, THA surgery is more complicated, has a longer operative time,

Fig. 4. Relative risk and 95% CI for pain rates.

588 The Journal of Arthroplasty Vol. 27 No. 4 April 2012

Fig. 5. Relative risk and 95% CI for infection rates.

carries a greater risk of dislocation, and has a higher initial cost [4], but it leads to fewer reoperations, better function, and less pain during rehabilitation [24]. Coates and Armour [25] advocate that THA for displaced femoral neck fractures in elderly patients should be used in cases of acetabular erosion with HA. The objective of our meta-analysis is to collect as much and as inclusive a data set as possible from prospective randomized and quasi-randomized controlled trials to determine whether THA is superior to HA for the treatment of displaced femoral neck fractures in elderly patients. To this end, we collected 9 studies that collectively included 1208 patients. Considering different follow-up durations in these studies, the data could not be directly pooled, and thus, we divided the data into 3 different subgroups (within 1 year, within 5 years, and at 13 years). Therefore, we obtained a result for each of the 3 subgroups as well as an overall result. Compared with HA, our meta-analysis has provided evidence that THA for the treatment of displaced femoral neck fracture has lower long-term risks of reoperation rates at 13 years postoperatively, which is in accordance with meta-analyses performed by Hopley et al [26]. With respect to mortality, there was an increased postoperative risk for THA patients compared with HA patients, but there was no statistical significance between the 2 groups at different times in our study. This result was not in accordance with results in Goh et al [27]. With respect to the dislocation rate, there were 4 studies within the first subgroups (within 1 year) and, except for the study by Skinner et al [13], 3 of these

studies [16-18] reported no dislocations in the HA groups, whereas the number of dislocations in the THA group was 3 and 1, respectively [16,18]. In the study by Blomfeldt et al [17], there were no dislocations in the THA group. In the second subgroups (within 5 years), there were 3 studies. Dorr et al [12] and Keating et al [15] reported the dislocation rate being higher in the THA group compared with the HA group. A significant difference between the 2 groups, in which there were no dislocations in the HA group and 8 dislocations in the THA group, was obtained. Accordingly, within 5 years, the result of our metaanalysis showed that THA had a greater risk of dislocation. At 13 years, Ravikumar and Marsh [14] reported that there was a higher dislocation rate in the THA group compared with the HA group. So far, 2 recent meta-analyses [26,27] reported a trend of a higher rate of dislocation in the THA group compared with the HA group; however, there were no significant differences in their studies. To our knowledge, our study is the first to conclude that there is a significant difference in dislocation rate between the 2 groups. The incidence of dislocation after artificial hip arthroplasty varies in reported series. Factors associated with dislocation include excessive anterversion or retroversion of the prosthesis, posterior capsulectomy, and excessive postoperative flexion or rotation with the hip adducted. It is noted that infection is also a common cause of postoperative dislocation. Fortunately, the dislocation rate has been reduced for the development of surgical techniques and newly designed implants. There are few articles to discuss the change in the dislocation rates over the years after femoral

HA vs THA For Elderly Femoral Neck Fractures  Zi-Sheng et al

neck fractures; therefore, a long-term and welldesigned study is desired in the near future. Our meta-analysis indicated that THA had a lesser likelihood of infection compared with HA, although the difference did not reach statistical significance. This result is in accordance with the study by Hopley et al [26]. Pain rates at 1 and 13 years were significantly higher in HA patients than that in THA patients, and these differences were significant.

Conclusion In summary, the treatment of THA for displaced femoral neck fractures in elderly patients may provide better results in terms of reoperation rate, pain relief, and functional improvement; however, displaced femoral neck fractures treated with HA yield a lower incidence of dislocation, shortened operative time, and decreased blood loss. In choosing surgical treatment of displaced femoral neck fractures, orthopedists should comprehensively consider physiological age, life expectancy, preexisting disease, quality-of-life demands, anticipated functional demands, psychological-mental status, as well as bone and joint quality. Larger, randomized controlled trials are needed to confirm the efficacy of THA vs HA for displaced femoral neck fractures in the elderly population. Clinical trials are also needed to include current techniques, modern devices, and a unified clinical outcome index that will be beneficial to developing a better guide for clinical practice via updated meta-analysis.

Acknowledgments This study was supported by a project from Shanghai Municipal Health Bureau (no. 2010212).

References 1. Bhandari M, Devereaux PJ, Tornetta III P, et al. Operative management of displaced femoral neck fractures in elderly patients: an international survey. J Bone Joint Surg Am 2005;87:2122. 2. Macaulay W, Pagnotto MR, Iorio R, et al. Displaced femoral neck fractures in the elderly: hemiarthroplasty versus total hip arthroplasty. J Am Acad Orthop Surg 2006;14:287. 3. Iorio R, Healy WL, Lemos DW, et al. Displaced femoral neck fractures in the elderly: outcomes and cost effectiveness. Clin Orthop Relat Res 2001;383:229. 4. Schmidt AH, Leighton R, Parvizi J, et al. Optimal arthroplasty for femoral neck fractures: is total hip arthroplasty the answer? J Orthop Trauma 2009;23:428. 5. Papandrea RF, Froimson MI. Total hip arthroplasty after acute displaced femoral neck fractures. Am J Orthop 1996; 25:85. 6. Cartlidge IJ. Primary total hip replacement for displaced subcapital femoral fractures. Injury 1981;13:249.

589

7. Healy WL, Iorio R. Total hip arthroplasty: optimal treatment for displaced femoral neck fractures in elderly patients. Clin Orthop Relat Res 2004;429:43. 8. Kyle RF. Fractures of the femoral neck. Instr Course Lect 2009;58:61. 9. Juni P, Altman DG, Egger M. Assessing the quality of controlled clinical trials. BMJ 2001;323:42. 10. Robins J, Greenland S, Breslow NE. A general estimator for the variance of the Mantel-Haenszel odds ratio. Am J Epidemiol 1986;124:719. 11. Dersimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials 1986;7:177. 12. Dorr LD, Glousman R, Hoy AL, et al. Treatment of femoral neck fractures with total hip replacement versus cemented and noncemented hemiarthroplasty. J Arthroplasty 1986; 1:21. 13. Skinner P, Riley D, Ellery J, et al. Displaced subcapital fractures of the femur: a prospective randomized comparison of internal fixation, hemiarthroplasty and total hip replacement. Injury 1989;20:291. 14. Ravikumar KJ, Marsh G. Internal fixation versus hemiarthroplasty versus total hiparthroplasty for displaced subcapital fractures of femur—13 year results of a prospective randomised study. Injury 2000;31:793. 15. Keating JF, Grant A, Masson M, et al. Displaced intracapsular hip fractures in fit, older people: a randomised comparison of reduction and fixation, bipolar hemiarthroplasty and total hip arthroplasty. Health Technol Assess 2005; 9(41):iii-iv, ix-x, 1. 16. Baker RP, Squires B, Gargan MF, et al. Total hip arthroplasty and hemiarthroplasty in mobile, independent patients with a displaced intracapsular fracture of the femoral neck. A randomized, controlled trial. J Bone Joint Surg Am 2006;88:2583. 17. Blomfeldt R, Törnkvist H, Eriksson K, et al. A randomised controlled trial comparing bipolar hemiarthroplasty with total hip replacement for displaced intracapsular fractures of the femoral neck in elderly patients. J Bone Joint Surg Br 2007;89:160. 18. Macaulay W, Nellans KW, Garvin KL, et al. Prospective randomized clinical trial comparing hemiarthroplasty to total hip arthroplasty in the treatment of displaced femoral neck fractures: winner of the Dorr Award. J Arthroplasty 2008;23:2. 19. Mouzopoulos G, Stamatakos M, Arabatzi H, et al. The four-year functional result after a displaced subcapital hip fracture treated with three different surgical options. Int Orthop 2008;32:367. 20. van der Bekerom MP, Hilverdink EF, Sierevelt IN, et al. A comparison of hemiarthroplasty with total hip replacement for displaced intracapsular fracture of the femoral neck: a randomised controlled multicentre trial in patients aged 70 years and over. J Bone Joint Surg Br 2010;92:1422. 21. Hedlund R, Lindgren U, Ahlbom A. Age- and sex-specific incidence of femoral neck and trochanteric fractures. An analysis based on 20,538 fractures in Stockholm County, Sweden, 1972-1981. Clin Orthop Relat Res 1987; 222:132. 22. York JD, Allen PG, Smith BP, et al. Prosthetic treatment of hip fractures in the elderly patient. J Surg Orthop Adv 2010;19:98.

590 The Journal of Arthroplasty Vol. 27 No. 4 April 2012 23. Raaymakers EL. Fractures of the femoral neck: a review and personal statement. Acta Chir Orthop Traumatol Cech 2006;73:45. 24. Rogmark C, Johnell O. Orthopaedic treatment of displaced femoral neck fractures in elderly patients. Disabil Rehabil 2005;27:1143. 25. Coates RL, Armour P. Treatment of subcapital fractures by primary total hip replacement. Injury 1979;11:132.

26. Hopley C, Stengel D, Ekkernkamp A, et al. Primary total hip arthroplasty versus hemiarthroplasty for displaced intracapsular hip fractures in older patients: systematic review. BMJ 2010;340:c2332. 27. Goh SK, Samual M, Su DH, et al. Meta-analysis comparing total hiparthroplasty with hemiarthroplasty in the treatment of displaced neck of femur fracture. J Arthroplasty 2009;24:400.