Early Orthogeriatric Treatment of Trauma in the Elderly
A Systematic Review and Metaanalysis
; ; ; ; ; ; ; ;
Background: More than 125 000 hip fractures occur in Germany every year, with a one-year mortality of about 25%. To improve treatment outcomes, models of cooperation between trauma surgery and geriatrics have been developed. Their benefit has not yet been unequivocally demonstrated.
Methods: We systematically searched the Medline database and the Cochrane Library for prospective randomized controlled trials in which the treatment of elderly patients with fractures by the trauma surgery service alone was compared with preoperatively initiated collaborative treatment by the trauma surgery and geriatric services (“orthogeriatric” treatment). We investigated three treatment outcome variables—length of hospital stay, in-hospital mortality, and one-year mortality—in a metaanalysis.
Results: The five trials of hip fracture treatment that met the selection criteria all had relatively small study populations and a high risk of bias. The outcomes with respect to hospital stay differed greatly among trials (I2: 88.5%), and geriatric intervention was not found to have any statistically significant effect (0.06 days, 95% confidence interval [CI]: –3.74 to 3.62 days). The relative risk of dying in the hospital was 0.66 for orthogeriatric treatment (95% CI: 0.28–1.55, p = 0.34), and the hazard ratio for one-year mortality was 0.79 in favor of orthogeriatric treatment (95% CI: 0.57 to 1.10, p = 0.17). A metaanalysis of functional outcomes was not possible.
Conclusion: Only a few randomized controlled trials of early orthogeriatric treatment have been performed, and these trials are of limited quality. Due to low case numbers, a benefit from interdisciplinary orthogeriatric treatment could not clearly be demonstrated. Further trials are needed.
As a result of demographic change, with the number of elderly persons in industrialized countries rising, the number of fractures in geriatric patients is also increasing.
Typical geriatric fractures are fractures of the proximal femur, the proximal humerus, the distal radius, vertebral bodies, and the pelvis, and increasingly also periprosthetic fractures (1). These fractures, which are associated with osteoporosis and falls, pose great challenges to treating physicians. On the one hand, altered bone structure complicates surgical care, and on the other comorbidities often lead to complications (2).
The most significant socio-economic factor in this setting is proximal femoral fractures. These are fractures of the neck of the femur and pertrochanteric and subtrochanteric fractures (ICD-10 S72.0 to S72.2 ). In 2009 more than 125 000 patients over the age of 70 years received inpatient treatment for the principal diagnosis proximal femoral fracture in Germany alone (4). Despite great advances in surgical care, treatment outcomes remain disappointing (5). One-year mortality for proximal femoral fractures is approximately 25% (6), and around one-third of patients lose their independence within the same period (7). The direct annual costs of illness are estimated at €2.5 billion in Germany alone (8). Clearly, then, optimum patient care is important. In order to be able to deal better with these patients’ multimorbidity, various models for collaborative orthogeriatric care of patients with proximal femoral fractures have been developed worldwide in recent years (9). To date it has only been shown that patients with proximal femoral fractures benefit from orthogeriatric care during rehabilitation (10). It has not yet been possible to provide unambiguous evidence of an advantage for interdisciplinary orthogeriatric care begun perioperatively, although individual studies have been published, some of the results of which are promising (9).
This systematic review and metaanalysis is intended to represent the current state of scientific knowledge on the possible benefit for patients with typical geriatric fractures of orthogeriatric care begun perioperatively.
The first stage in the compilation of this article was a systematic review of Medline and the Cochrane Library performed by two reviewers. Next, a metaanalysis of the various outcome parameters was performed, to the extent that this was possible. Both these steps were performed in line with the PRISMA Statement (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) (11). The search terms “fragility fracture,” “geriatric fracture,” “osteoporotic fracture,” “hip fracture,” “femur fracture,” “femoral fracture,” “humerus fracture,” “humeral fracture, “periprosthetic fracture,” and “vertebral fracture” combined with “multidisciplinary,” “comanagement,” “co-management,” “interdisciplinary,” “comprehensive care,” and the terms “ortho-geriatric” and “orthogeriatric” were used to identify studies which investigated orthogeriatric care begun perioperatively for geriatric trauma patients (Figure 1). Study inclusion criteria were selected on the basis of the PICOS (Patient Group, Intervention, Control Intervention, Outcome, Study Design) question (Table 1), and risk of bias was evaluated using the criteria shown in Table 2. In a second stage, metaanalysis was performed on the studies identified during the review of the literature using the random effects model.
The above-mentioned search terms identified 558 citations, 49 of which were literature citations (39 original articles and 10 reviews) that met the inclusion criteria and of which the full text was available (Figure 1). Manual searches of the bibliographies of the selected original articles did not identify any further publications. Patients with proximal femoral fractures were investigated in all the selected literature citations.
Of the 39 identified original articles, 10 were randomized controlled trials (eTable 1). However, closer examination revealed that the articles by Shyu et al. (16–19, 21) and Stenvall et al. (14, 15) were based on one study or one study cohort each. These were included only once each. The metaanalysis therefore included a final total of five studies, involving 970 patients altogether (482 intervention, 488 control).
Risk of bias
All the studies included in the metaanalysis were shown to have a high risk of bias (Table 2).
Some of the parameters examined varied between studies, so metaanalysis was performed only on length of hospital stay, hospital mortality, and long-term mortality (eTable 2).
Length of hospital stay
Metaanalysis of length of hospital stay revealed very high heterogeneity (I2: 88.5%) and a difference of –0.06 days (95% confidence interval [CI]: –3.74 to 3.62), meaning that the length of hospital stay in the intervention group was not significantly shorter (p = 0.97; Figure 2).
This parameter, too, yielded heterogeneous results (I2: 31.6%). Only Vidán et al. found a significant decrease in hospital mortality as a result of geriatric intervention (12). Overall, analysis showed a relative risk of 0.66 in favor of the intervention, which was statistically insignificant (95% CI: 0.28 to 1.55; p = 0.34; Figure 3).
These results were more homogeneous than those for the other parameters (I2: 0%). Only the small study by Uy et al. showed a higher mortality rate in the intervention group (13). In the other four studies, more patients died in the control groups. With a hazard ratio (HR) of 0.79 (95% CI: 0.57 to 1.10), metaanalysis did not show significantly lower mortality as a result of geriatric intervention (p = 0.17; Figure 4).
Other study results
For the other outcome parameters, Stenvall et al. showed a significant decrease in the rate of complications during hospital stay; this was also true of patients who fell during the same period (14). At the same time, in this group of patients preoperative length of hospital stay was lower than in the other studies; however, there were no major differences between the two patient groups within studies. The rate of patients who recovered their preinjury walking ability or preinjury activity level was higher in the intervention group (15) (eTable 2). Shyu et al. found that at various points during the research the functional outcome in the intervention group was better than in the control group (16–19). Nevertheless, not all of the results were significant, as shown by those of Naglie et al. (20), Vidán et al. (12), and Uy et al. (13) (eTable 2). Because methods of assessment varied and the parameters chosen were heterogenous, we did not perform a metaanalysis on functional outcome. Only Shyu et al. investigated health-related quality of life (HRQoL). They found higher HRQoL in the intervention group (21).
This systematic review and metaanalysis are intended to investigate whether interdisciplinary orthogeriatric care begun perioperatively is more beneficial than trauma surgery alone. Only five studies could be included. All five studies investigated patients with proximal femoral fractures. Interdisciplinary care led to a decrease in hospital mortality and one-year mortality, but these decreases were not statistically significant.
The selected search terms identified a number of potentially eligible studies. All the studies investigated patients with proximal femoral fractures. This highlights the worldwide significance of these fractures due to their frequency, their high treatment costs, and their lasting effect on patients’ quality of life, but also as a tracer diagnosis for suitable treatment of fragility fractures. However, in this setting it is doubtful whether findings on treatment of patients with proximal femoral fractures can be extrapolated to to patients with other fragility fractures without further information.
Most of the identified studies on proximal femoral fractures were nonrandomized studies (Figure 1). Although most of the results of these studies were promising (22–29) they were not included in our metaanalysis because they were nonrandomized.
The five included studies, which were conducted in various countries, investigated various models of collaboration described in Pioli (30). Inclusion and exclusion criteria also differed between the included studies (eTable 1). This explains the observed heterogeneity of the studies’ results—however, the results cannot be simply extrapolated to the German health-care system. The control groups received trauma surgery only. Only a few studies compared individual models, so the models cannot be evaluated. Mazzola et al. only found earlier mobilization with interdisciplinary treatment directly after hospital admission versus interdisciplinary treatment begun postoperatively. However, this was an observational study in which patient characteristics differed between the two patient groups (33). In our opinion, local circumstances must be taken into account when developing an interdisciplinary treatment approach.
Other studies showed that patients with proximal femoral fractures, some without concomitant treatment by a geriatrician, benefited from structured, multidisciplinary, multiprofessional treatment according to treatment pathways (27, 34–37). Among other things, these approaches include early surgery; early mobilization with full weight bearing postoperatively; internal guidelines covering common problems in orthogeriatric patients such as delirium, malnutrition, coagulation management, and pain; and finally structured discharge management.
This metaanalysis was therefore only able to investigate whether early interdisciplinary orthogeriatric treatment of various types was superior to conventional trauma surgery.
Length of hospital stay
Length of hospital stay was approximately the same in both treatment groups (Figure 2). It was therefore impossible to determine whether length of hospital stay was affected by the different treatment models. The study by Naglie et al. was noteworthy for its significantly longer length of stay, 29.2 days, in the intervention group (control: 20.9 days) (20). One explanation for this might be that the study included only one geriatric internist visit, which may have prolonged inpatient stay. Length of inpatient hospital stay in this study was significantly above the mean length of acute medical hospitalization in Germany, which is currently 15.5 days (median: 13 days) (31). Other studies not included in the metaanalysis also yielded heterogeneous results regarding length of hospital stay (9). Differences in the length of hospital stay may be the result of differences in different countries' health care systems and the features of local structures as a quality criterion in the treatment of orthogeriatric patients. Length of hospital stay in the same framework conditions would itself be problematic as a quality criterion for patient care, as it does not provide any information on quality of treatment.
The relative risk was 0.66 (95% CI: 0.28 to 1.55) in favor of interdisciplinary treatment versus trauma surgery alone, and there was no significant decrease in hospital mortality (Figure 3). Vidán et al. explain their significant reduction in mortality (0.6% versus 5.5%) by prevention or better management of systemic complications in the intervention group. In the study by Vidán et al. the rate of major systemic complications was significantly lower in the intervention group (45.2% versus 61.7%) (12). As Vidán et al. investigated Pioli’s fourth model, joint interdisciplinary orthogeriatric care from admission to discharge, the figures might provide evidence that intensive interdisciplinary care offers patients the most benefits. Astonishingly, only one of 162 patients (0.6%) died in the study by Shyu et al. (18). In contrast, the most recent mortality rate in Germany was 5.2% (31). This can be explained by the study’s exclusion criteria of terminal diseases and major cognitive impairment, which are associated with higher mortality.
As described in other studies, proximal femoral fractures are associated with considerable long-term mortality. According to a current review of the literature, one-year mortality is approximately 25% (6). Mortality in the studies included in our metaanalysis was significantly lower: an average of 15%. However, in one study the follow-up period was only six months, and in another only four months. Although the data from both studies were included in the metaanalysis using Perneger’s method (32), the varying follow-up times are a limitation. No significant reduction in one-year mortality was shown for the intervention group versus the control group (Figure 4). However, it should be remembered that the studies in this metaanalysis included only a small number of cases. In our opinion, the one-year mortality figures are evidence of a lasting effect of early interdisciplinary care that should be investigated in more detail in further studies. Bachmann et al. also included studies that concentrated on rehabilitation in a metaanalysis and were able to corroborate the positive effect of collaborative orthogeriatric care on one-year mortality (10).
Further outcome parameters
Functional improvement in patients in the intervention group was achieved in most studies. These included, among others, a reduced rate of falls (14) and more frequent recovery of preinjury walking ability (15, 18), preinjury mobility (12, 20), or preinjury ADL (activities of daily living) (12, 15) (eTable 2). Because of differing measurement tools and outcome parameters, these could not be included in metaanalysis, although the recovery of independence associated with functional status is itself of great importance to patients. Only Shyu et al. investigated patients’ health-related quality of life. They found that it was significantly positively affected by interdisciplinary treatment (21). The studies analyzed here examined only length of hospital stay, in view of the costs incurred. Treatment problems are already emerging in orthogeriatric care. For example, in a survey for 2009, 37% of hospitals reported problems in follow-up care of patients with proximal femoral fractures (38). The various models of interdisciplinary collaboration may help to reduce these treatment problems. If these models improve quality of care and therefore also functional outcome, there may also be financial advantages for all those providing funding, as a result of a reduction in the demand for care. This means that health economic evaluation of interdisciplinary care versus conventional treatment approaches is essential to overall assessment.
Risk of bias
The power of this metaanalysis is limited by the high risk of bias of all the studies it included (Table 2). In addition, there is a risk of bias across studies as a result of publication bias or selective reporting in individual studies. Although we did not perform analysis of publication bias because of the small number of studies and the high risk of bias within studies, we do not believe that there are good-quality and therefore expensive studies on this important subject that have not been published. Selective reporting, on the other hand, is more likely in our opinion. A further limitation is the fact that the literature search involved only two databases.
In the literature overall there is a variety of evidence that early orthogeriatric collaboration leads to improved outcomes for orthogeriatric patients. The small number of randomized controlled trials, which had small case numbers and were of limited quality, did not allow this benefit to be demonstrated with certainty in our metaanalysis. Randomized controlled trials with sufficient patient numbers should therefore be conducted, investigating functional parameters, quality of life, and financial issues.
Conflict of interest statement
Dr. Riem has received publication fees from Thieme and reimbursement of conference participation costs from the German Society for Orthopedics and Trauma Surgery, the German Society for Trauma Surgery, and the European Society for Trauma and Emergency Surgery. She has received payment for a research project she herself initiated from the German Society for Trauma Surgery.
Prof. Liener, Dr. Buecking, Dr. Bliemel, Dr. Timmesfeld, Prof. Ruchholtz, Prof. Hartwig, Dr. Eschbach, and Dr. Friess declare that no conflict of interest exists.
Manuscript received on 15 October 2012, revised version accepted on
11 January 2013.
Translated from the original German by Caroline Devitt, M.A.
Dr. med. Benjamin Bücking
Hospital for Trauma, Hand, and Reconstructive Surgery
Universitätsklinikum Gießen und Marburg GmbH
Baldingerstr., 35043 Marburg, Germany
@For eReferences please refer to:
eMethods + eTables:
Dr. med. Eschbach
Institute of Medical Biometrics and Epidemiology, Philipps Universität Marburg: Dr. rer. nat. Timmesfeld
Department of Orthopedics and Trauma Surgery, Diakonissenkrankenhaus Karlsruhe:
Dr. med. Riem, Prof. Dr. med. Hartwig
Department of Trauma, Hand- and Reconstructive Surgery, St. Clemens Hospital Sterkrade: Dr. med. Friess
Clinic for Orthopedics and Trauma Surgery, Marienhospital Stuttgart: Prof. Dr. med. Liener
|1.||Gesundheitsberichterstattung des Bundes: Diagnosedaten der Krankenhäuser ab 2000; available at: www.gbe-bund.de/oowa921-install/servlet/oowa/aw92/dboowasys921.xwdevkit/xwd_init?gbe.isgbetol/xs_start_neu/&p_aid=i&p_aid=36857213&nummer=702&p_sprache=D&p_indsp=522&p_aid=9442669; Last accessed on 14 June 2012.|
|2.||Roche JJ, Wenn RT, Sahota O, et al.: Effect of comorbidities and postoperative complications on mortality after hip fracture in elderly people: prospective observational cohort study. BMJ 2005; 331: 1374. CrossRef MEDLINE PubMed Central|
|3.||ICD-10: Version: 2010. International Statistical Classification of Diseases and Related Health Problems 10th Revision; available at: http://apps.who.int/classifications/apps/icd/icd10online/. Last accessed on 2 May 2012.|
|4.||Statistisches Bundesamt Deutschland. Gesundheit. Grunddaten der Krankenhäuser 2009. Statistisches Bundesamt, Wiesbaden 2011; available at: www.destatis.de/jetspeed/portal/cms/Sites/destatis/Internet/DE/Content/Publikationen/Fachveroeffentlichungen/Gesundheit/Krankenhaeuser/GrunddatenKrankenhaeuser2120611097004,property=file.pdf Last accessed on 14 June 2011.|
|5.||Roth T, Kammerlander C, Gosch M, Luger TJ, Blauth M: Outcome in geriatric fracture patients and how it can be improved. Osteoporos Int 2010; 21(Suppl 4): 615–9. CrossRef MEDLINE|
|6.||Hu F, Jiang C, Shen J, et al.: Preoperative predictors for mortality following hip fracture surgery: A systematic review and meta-analysis. Injury 2012; 43: 676–85. CrossRef MEDLINE|
|7.||Becker C, Gebhard F, Fleischer S, et al.: Prediction of mortality, mobility and admission to long-term care after hip fractures. Unfallchirurg 2003; 106: 32–8. CrossRef MEDLINE|
|8.||Konnopka A, Jerusel N, König HH: The health and economic consequences of osteopenia- and osteoporosis-attributable hip fractures in Germany: estimation for 2002 and projection until 2050. Osteoporos Int 2009; 20: 1117–29. CrossRef MEDLINE|
|9.||Kammerlander C, Roth T, Friedman SM, et al.: Ortho-geriatric service—a literature review comparing different models. Osteoporos Int 2010; 21(Suppl 4): 637–46. CrossRef MEDLINE|
|10.||Bachmann S, Finger C, Huss A, et al.: Inpatient rehabilitation specifically designed for geriatric patients: systematic review and meta-analysis of randomised controlled trials. BMJ 2010; 340: 1718. CrossRef MEDLINE|
|11.||Moher D, Liberati A, Tetzlaff J, et al.: Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. BMJ 2009; 339: 2535. CrossRef MEDLINE|
|12.||Vidán M, Serra JA, Moreno C, et al.: Efficacy of a comprehensive geriatric intervention in older patients hospitalized for hip fracture: a randomized, controlled trial. J Am Geriatr Soc 2005; 53: 1476–82. CrossRef MEDLINE|
|13.||Uy C, Kurrle SE, Cameron ID: Inpatient multidisciplinary rehabilitation after hip fracture for residents of nursing homes: a randomised trial. Australas J Ageing 2008; 27: 43–4. CrossRef MEDLINE|
|14.||Stenvall M, Olofsson B, Lundström M, et al.: A multidisciplinary, multifactorial intervention program reduces postoperative falls and injuries after femoral neck fracture. Osteoporos Int 2007; 18: 167–75. CrossRef MEDLINE PubMed Central|
|15.||Stenvall M, Olofsson B, Nyberg L, et al.: Improved performance in activities of daily living and mobility after a multidisciplinary postoperative rehabilitation in older people with femoral neck fracture: a randomized controlled trial with 1-year follow-up. J Rehabil Med 2007; 39: 232–8. CrossRef MEDLINE|
|16.||Shyu YI, Liang J, Wu CC, et al.: A pilot investigation of the short-term effects of an interdisciplinary intervention program on elderly patients with hip fracture in Taiwan. J Am Geriatr Soc 2005; 53: 811–8. CrossRef MEDLINE|
|17.||Shyu YI, Liang J, Wu CC, et al.: Interdisciplinary intervention for hip fracture in older Taiwanese: benefits last for 1 year. J Gerontol A Biol Sci Med Sci 2008; 63: 92–7. CrossRef MEDLINE|
|18.||Shyu YI, Liang J, Wu CC, et al.: Two-year effects of interdisciplinary intervention for hip fracture in older Taiwanese. J Am Geriatr Soc 2010; 58: 1081–9. CrossRef MEDLINE|
|19.||Shyu YI, Tsai WC, Chen MC, et al.: Two-year effects of an interdisciplinary intervention on recovery following hip fracture in older Taiwanese with cognitive impairment. Int J Geriatr Psychiatry 2012; 27: 529–3. CrossRef MEDLINE|
|20.||Naglie G, Tansey C, Kirkland JL, et al.: Interdisciplinary inpatient care for elderly people with hip fracture: a randomized controlled trial. CMAJ 2002; 167: 25–32. MEDLINE PubMed Central|
|21.||Shyu YI, Liang J, Wu CC, et al.: An interdisciplinary intervention for older Taiwanese patients after surgery for hip fracture improves health-related quality of life. BMC Musculoskelet Disord 2010; 11: 225. CrossRef MEDLINE PubMed Central|
|22.||Khasraghi FA, Christmas C, Lee EJ, et al.: Effectiveness of a multidisciplinary team approach to hip fracture management. J Surg Orthop Adv 2005; 14: 27–31. MEDLINE|
|23.||Friedman SM, Mendelson DA, Kates SL, et al.: Geriatric co-management of proximal femur fractures: total quality management and protocol-driven care result in better outcomes for a frail patient population. J Am Geriatr Soc 2008; 56: 1349–56. CrossRef MEDLINE|
|24.||Friedman SM, Mendelson DA, Bingham KW, et al.: Impact of a comanaged Geriatric Fracture Center on short-term hip fracture outcomes. Arch Intern Med 2009; 169: 1712–7. CrossRef MEDLINE|
|25.||Zuckerman JD, Sakales SR, Fabian DR, et al.: Hip fractures in geriatric patients. Results of an interdisciplinary hospital care program. Clin Orthop Relat Res 1992; 274: 213–25. MEDLINE|
|26.||Koval KJ, Chen AL, Aharonoff GB, et al.: Clinical pathway for hip fractures in the elderly: the Hospital for Joint Diseases experience. Clin Orthop Relat Res 2004; 425: 72–81. CrossRef MEDLINE|
|27.||Roberts HC, Pickering RM, Onslow E, et al.: The effectiveness of implementing a care pathway for femoral neck fracture in older people: a prospective controlled before and after study. Age Ageing 2004; 33: 178–84. CrossRef MEDLINE|
|28.||Khan R, Fernandez C, Kashifl F, et al.: Combined orthogeriatric care in the management of hip fractures: a prospective study. Ann R Coll Surg Engl 2002; 84: 122–4. MEDLINE PubMed Central|
|29.||Fisher AA, Davis MW, Rubenach SE, et al.: Outcomes for older patients with hip fractures: the impact of orthopedic and geriatric medicine cocare. J Orthop Trauma 2006; 20: 172–8. CrossRef MEDLINE|
|30.||Pioli G, Giusti A, Barone A: Orthogeriatric care for the elderly with hip fractures: where are we? Aging Clin Exp Res 2008; 20: 113–22. MEDLINE|
|31.||AQUA – Institut für angewandte Qualitätsförderung und Forschung im Gesundheitswesen: Bundesauswertung zum Erfassungsjahr 2011. 17/1 – Hüftgelenksnahe Femurfraktur. Qualitätsindikatoren. Available at: www.aqua-institut.de. Last accessed on 31 May 2012.|
|32.||Perneger TV: Estimating the relative hazard by the ratio of logarithms of event-free proportions. Contemp Clin Trials 2008; 29: 762–6. CrossRef MEDLINE|
|33.||Mazzola P, De Filippi F, Castoldi G, et al.: A comparison between two co-managed geriatric programmes for hip fractured elderly patients. Aging Clin Exp Res 2011; 23: 431–6. MEDLINE|
|34.||Gholve PA, Kosygan KP, Sturdee SW, et al.: Multidisciplinary integrated care pathway for fractured neck of femur. A prospective trial with improved outcome. Injury 2005; 36: 93–8. MEDLINE|
|35.||Choong PF, Langford AK, Dowsey MM, et al.: Clinical pathway for fractured neck of femur: a prospective, controlled study. Med J Aust 2000; 172: 423–6. MEDLINE|
|36.||Olsson LE, Karlsson J, Ekman I: The integrated care pathway reduced the number of hospital days by half: a prospective comparative study of patients with acute hip fracture. J Orthop Surg Res 2006; 1: 3. CrossRef MEDLINE PubMed Central|
|37.||Swanson CE, Day GA, Yelland CE, et al.: The management of elderly patients with femoral fractures. A randomised controlled trial of early intervention versus standard care. Med J Aust 1998; 169: 515–8. MEDLINE|
|38.||Bliemel C, Bieneck F, Riem S, et al.: Subsequent treatment following proximal femoral fracture – who, when, where? Assessment of the current situation in Germany. Z Orthop Unfall 2012; 150: 210–7. CrossRef MEDLINE|
|e1.||DerSimonian R, Laird N: Meta-analysis in clinical trials. Control Clin Trials 1986; 7: 177–88. CrossRef MEDLINE|
|e2.||Higgins JP, Thompson SG: Quantifying heterogeneity in a meta-analysis. Stat Med 2002 ; 21: 1539–58. CrossRef MEDLINE|
|e3.||Perneger TV: Estimating the relative hazard by the ratio of logarithms of event-free proportions. Contemp Clin Trials 2008; 29: 762–6. CrossRef MEDLINE|