Long-Term Survival of the Native Hip After a Minimally Displaced, Nonoperatively Treated Acetabular Fracture

Clarke-Jenssen, John MD; Wikerøy, Annette K.B. MD; Røise, Olav MD, PhD; Øvre, Stein Arne MD, PhD; Madsen, Jan Erik MD, PhD

Journal of Bone & Joint Surgery - American Volume:
doi: 10.2106/JBJS.15.01154
Scientific Articles

Background: Few studies have evaluated the long-term results for nonoperatively treated acetabular fractures. The purpose of this study was to describe the long-term survival of the native acetabulum as well as the clinical and radiographic outcome for patients with nonoperatively treated acetabular fractures.

Methods: All patients with acetabular fractures are prospectively registered in our acetabular fracture database and followed up at regular intervals for up to 20 years. We identified 236 patients (237 fractures) who had been treated nonoperatively between 1994 and 2004; 51 patients with incomplete data were excluded. For the survival analysis, 186 fractures with an average follow-up of 9 years (range, 1 to 20 years) were included. For the long-term clinical outcome, 104 patients with an average follow-up of 12.1 years (range, 9 to 20 years) were included.

Results: The 10-year survival of the native hips was 94% (111 hips were at risk). Eighty-nine percent of the patients had a good or excellent Harris hip score, and 88% had a good or excellent Merle d’Aubigné and Postel score. The most important negative predictor for clinical outcome and survival of the hip was a fracture step-off of ≥2 mm measured in the obturator oblique radiograph.

Conclusions: Nonoperative treatment of minimally displaced acetabular fractures yields good to excellent long-term results. For patients with a questionable indication for fracture surgery, oblique radiographs (Judet views) are a helpful tool in the decision-making process, as a fracture step-off of ≥2 mm is a strong predictor for a poor clinical and radiographic result at 10 years.

Level of Evidence: Prognostic Level II. See Instructions for Authors for a complete description of levels of evidence.

Author Information

1Orthopaedic Department, Division of Surgery and Clinical Neuroscience, Oslo University Hospital, Oslo, Norway

2Orthopaedic Department, Akershus University Hospital, Lørenskog, Norway

3Institute of Clinical Medicine, Faculty of Medicine, University of Oslo, Oslo, Norway

E-mail address for J. Clarke-Jenssen: John.Clarke-Jenssen@ous-hf.no

Article Outline

Acetabular fractures occur infrequently, with an incidence of approximately 3 per 100,000 people per year1. In the trauma community, there is a clear consensus that unstable or substantially displaced fractures with joint incongruity are best treated surgically with anatomic reduction and stable fracture fixation2,3. Historically, displaced acetabular fractures were treated conservatively with bed rest and traction4, and some recent reports have indicated that this approach is still used5,6. However, the risk for complications is high7,8, and the rate of success is questionable4,9. Some reports in the literature have described survival rates of >90% from 8 to 10 years after nonoperative treatment10,11. The clinical results, however, were closely associated with the initial displacement of the fracture.

Consequently, most pelvic fracture centers presently reserve nonoperative treatment for patients with minimally displaced (<2 or ≤3 mm) fractures and a congruent and stable hip. An intact weight-bearing dome with a roof arc angle of >45° may also indicate a nonoperative approach, as would complete both-column fractures exhibiting joint congruency2,3,12.

Few studies have described the long-term results for patients treated nonoperatively with early mobilization. The purpose of the present work was therefore to investigate the long-term survival of the native hip and the clinical results of nonoperatively treated acetabular fractures in patients at our institution.

Back to Top | Article Outline

Materials and Methods

Data were extracted from our acetabular fracture database, established in 1994 by one of the senior authors (O.R.). All acetabular fractures treated by the department are registered prospectively. Data on the mechanism of injury, radiographic parameters, fracture classification, treatment, and follow-up data, including the Harris hip score and the modified Merle d’Aubigné and Postel score, have been prospectively registered at the time of injury and at follow-up evaluations at 3 and 6 months and at 1, 2, 5, 10, 15, and 20 years after treatment.

We identified all patients treated nonoperatively between 1994 and 2004, to allow for a minimum of 10 years of follow-up. During this time period, nonoperative treatment was advocated when the intra-articular step-off was <2 mm or with secondary joint congruency, as seen in some both-column fractures. Fractures involving the acetabular wall(s) but with a stable joint were included. If serious medical comorbidities were contraindications to operative treatment, a step-off of ≥2 mm was accepted. For octogenarian patients with poor bone stock, there was a slightly higher acceptance for fracture displacement. The nonoperative treatment regimen included early mobilization with partial weight-bearing for the first 8 weeks. Seventeen patients early in the series were treated with skeletal traction for a limited time period (2 to 12 weeks). A total of 236 patients (237 fractures) who received conservative treatment during the time period were identified and constituted the study group. Fifty-one of the patients were excluded; they included 30 who had died before 1 year of follow-up had been completed, 11 who had not been followed because of severe medical comorbidities, and 10 who had been lost to follow-up before 1 year (Fig. 1). In the same period, 257 patients were treated operatively.

In the hip survival analysis, all patients who had not died or been lost to follow-up within the first year after injury were included. A total of 185 patients (186 fractures) with an average follow-up of 9 years (range, 1 to 20 years) were available for analysis.

For the long-term clinical follow-up, all patients with complete data on the Harris hip score and/or modified Merle d’Aubigné score at a minimum of 9 years were included (Fig. 1). One hundred and four patients had been followed for an average of 12.1 years (range, 9 to 20 years). The Harris hip score was considered excellent (100 to 90 points), good (89 to 80 points), fair (79 to 70 points), or poor (<70 points)13. The Merle d’Aubigné score, as modified by Matta et al.14, was categorized as excellent (18 points), good (17, 16, or 15 points), fair (14 or 13 points), or poor (<13 points).

Patients were evaluated with 3 standard radiographs including an anteroposterior and 2 oblique (Judet) pelvic views, made as described by Letournel and Judet15. Ninety-three percent of the patients were evaluated with a computed tomography (CT) scan. All radiographs were assessed prospectively by the treating orthopaedic pelvic trauma surgeon, and there were 4 surgeons involved in treating these patients. In all 3 radiographs, measurements of the intra-articular diastasis and step-off, as well as the roof arc as described by Matta et al.14, were made (Fig. 2). The CT scans were analyzed for evidence of fracture lines in the upper 10 mm of the weight-bearing dome. The fractures were classified according to the system described by Letournel16 (Table I). They were also categorized by complexity into elementary or those with associated fractures for the purpose of statistical analysis.

Data were analyzed for hip survival, with total hip arthroplasty as the end point. Survival data were analyzed for differences with respect to fracture step-off, diastasis, roof arc, CT-verified fracture lines in the upper 10 mm of the weight-bearing dome, sex, and age of ≥60 years.

Back to Top | Article Outline

For the survival analysis, STATA software (version 13.1; StataCorp) was used. Data were analyzed with the Kaplan-Meier method, and differences in group survivorship were analyzed with the log-rank test. Cox regression was used to identify negative predictors and estimate hazard ratios. Competing risk regression was also applied to adjust for the number of patients who had died during the long follow-up period. For the clinical data, SPSS software (version 22; IBM) was used. The Mann-Whitney U test was applied because of non-normal distribution of the Harris hip and modified Merle d’Aubigné scores and also when analyzing the development of osteoarthritis. The level of significance was set at p = 0.05.

Back to Top | Article Outline


Survival of the Native Acetabulum

Twelve of 185 patients (186 fractures) had a total hip arthroplasty at an average of 6.9 years (range, 1.4 to 15.9 years) after the injury. The average 10-year survival for 111 hips at risk was 94% (95% confidence interval [CI], 89% to 97%). The data do not support estimation of survival beyond 10 years because of the relatively low number of observations beyond 10 years (Fig. 3). The average age at the time of injury was 49 years (range, 12 to 92 years). The most significant negative predictor was an intra-articular step-off of ≥2 mm, measured in the obturator oblique projection (Fig. 4). There was also significantly worse survival with an intra-articular step-off of ≥2 mm in the other projections (Table II). An intra-articular diastasis of ≥2 mm also predicted a significantly lower survival rate for the hip when detected in the obturator oblique radiograph; however, this was not the case for the anteroposterior and iliac oblique radiographs. When CT scans were analyzed for the presence or absence of fracture lines in the upper 10 mm of the weight-bearing dome, there was a nearly significant (p = 0.086) trend toward decreased joint survival when the posterior part of the dome was involved (Table II).

The roof arc14 was used to measure fracture involvement in the weight-bearing dome. We stratified the data into a roof arc of ≥45° or <45°. There was a significantly worse survival rate when the posterior roof arc was <45° (p = 0.046) measured on the iliac oblique radiographs. The anteroposterior radiographs indicated a significantly reduced survival for the incongruent joints compared with congruent ones (p = 0.040), with congruency defined as parallelism of the femoral and acetabular subchondral plates. This could not be shown for the oblique radiographs.

Individual fracture type or complexity (elementary or associated) had no significant impact on hip joint survival. In addition, there was no difference in joint survival between sexes or between a patient age of ≥60 or <60 years.

Cox regression showed similar results, with a significant increase in hazard ratios only for a step-off of ≥2 mm in all radiographic projections. When adjusting for competing risk in the regression model, the only significantly increased hazard ratio was observed for a step-off of ≥2 mm as seen on the obturator oblique radiograph (Table III).

Survival analysis was initially developed to study mortality. When it is used with other end points, such as survival of the native hip, there is a possibility of introducing bias because of competing risks17,18. In our study, there was 21% mortality during the follow-up period. When we adjusted for this factor by performing a competing risk regression analysis, it showed a decrease in the hazard ratios (Table III), which may be because the Kaplan-Meier estimator tends to overestimate the probability of failure in the face of competing risk (Fig. 5).

Back to Top | Article Outline
Clinical Results

The average Harris hip score at the time of follow-up was 93.4 points (range, 28 to 100 points), with 89% good or excellent scores. The average modified Merle d’Aubigné score was 16.6 points (range, 8 to 18 points), with 88% good or excellent scores (Table IV). The most important negative predictor for clinical outcome was a fracture step-off of ≥2 mm measured in the obturator oblique radiograph, with significant differences for both the Harris hip score and modified Merle d’Aubigné scores in comparison with those of patients with less displaced fractures. The mean Harris hip score was 89 points for patients with a step-off of ≥2 mm and 94 points for patients with a step-off of <2 mm. This difference was also seen in the modified Merle d’Aubigné score for patients with a fracture step-off of ≥2 mm in the anteroposterior radiographs. The measured fracture displacement in the iliac oblique radiographs did not predict the long-term outcome. Also, there were no differences in the clinical results with respect to the presence or absence of fracture lines in the weight-bearing dome, the sex or age of the patient, or the complexity of the fracture (Table V). The baseline data on the patients providing clinical results compared with those who were included only in the survival analysis showed no difference with respect to the sex of the patient or the Letournel classification. The patients providing clinical results were, however, an average of 20 years younger and were more likely to have a high-energy injury, but less likely to have a dislocation, than those in the survival analysis.

Back to Top | Article Outline
Radiographic Outcome

Ninety-seven patients (98 acetabular fractures) were available for complete radiographic examination, with an average follow-up of 12.5 years (range, 9 to 20 years). The radiographs were evaluated for osteoarthritis as described by Letournel and Judet2. The initial fracture radiographs showed a mean step-off of 1.3 mm (range, 0 to 15 mm) in the anteroposterior radiograph, 0.8 mm (range, 0 to 12 mm) in the obturator oblique radiograph, and 1.0 mm (range, 0 to 15 mm) in the iliac oblique radiograph. The mean diastasis was 2.4 mm (range, 0 to 25 mm) in the anteroposterior radiograph, 1.9 mm (range, 0 to 25 mm) in the obturator oblique radiograph, and 2 mm (range, 0 to 16 mm) in the iliac oblique radiograph.

At the time of the latest follow-up evaluation, 53% of the patients had no osteoarthritis of the affected hip, 25% had Letournel grade 1, 7% had grade 2, 7% had grade 3, and 8% had grade 5. The development of osteoarthritis was analyzed with respect to intra-articular step-off and a diastasis of ≥2 mm. We found that an intra-articular step-off of ≥2 mm significantly increased the risk for developing osteoarthritis when measured in the obturator oblique (p = 0.001) and iliac oblique radiographs (p = 0.011). Furthermore, a fracture diastasis of ≥2 mm in the iliac oblique radiograph (p = 0.005) predicted secondary osteoarthritis development. No such correlation was found in the other projections. The presence of fracture lines in the upper 10 mm of the weight-bearing dome was not significantly associated with the development of osteoarthritis (p = 0.061).

Back to Top | Article Outline


In this prospective cohort study, our main aims were to determine the long-term survival of the native hip in patients who received conservative treatment for an acetabular fracture and to analyze influencing factors. The overall 10-year survival of the native hip was excellent, with a survival rate of 94% (95% CI, 89% to 97%).

Few studies have described long-term survival of the native hip after conservative acetabular fracture treatment. In a slightly younger group of 35 patients (mean age, 42 years), with substantially greater fracture displacement of a mean of 16 mm (range, 5 to 27 mm), Sen and Veerappa reported survival of 33 of 35 hips after a mean follow-up of 4.1 years (range, 2 to 12 years)6. Although the survival rate for the hips was good, the clinical results showed that only 18 patients (51%) had a good or excellent Merle d’Aubigné score.

In a cohort described by Rowe and Lowell, the survival rate was 85% (53 of 62 hips) at a mean follow-up of 6 years19. The full spectrum of comminuted, severely displaced, and nondisplaced fractures was included, and a close correlation was observed between displacement and later survival of the hip, with 100% survival in the group of 21 nondisplaced fractures. That series was published before CT scans were available.

In the present study, 50% of the failures occurred during the first 4 years, with an average time to total hip arthroplasty of 6.9 years. This is in contrast to the findings of Hesp and Goris, who reported that 51 of 55 native hips in their series survived, with all failures occurring in the first 2 years10.

The most significant factor in predicting a poor survival of hips in our series was a step-off of ≥2 mm in the obturator oblique radiograph. If one was present, the long-term survival decreased toward 60%. In a report on 72 patients who had an acetabular fracture, Dunet et al. noted that 42 patients were treated nonoperatively and had an average follow-up of 6.8 years20. Nine of the patients received a secondary total hip arthroplasty. Those authors also noted that a lack of joint congruency in the posterior part of the joint appeared to be a significant predictor of poor survival.

The development of osteoarthritis was strongly associated with a step-off of ≥2 mm as measured on the obturator oblique radiograph (p = 0.001). This association was also confirmed in the iliac oblique projection (p = 0.011), whereas no significant association was found for fracture step-offs in the anteroposterior radiograph. The clinical results were also significantly reduced for this group of patients with a greater fracture displacement. The mean group differences, however, were within what has generally been accepted as a minimal clinically important difference. This does not change the fact that there are large individual differences in the clinical scores, probably of considerable interest to the individual patient.

Sen and Veerappa reported results similar to ours after a medium-term follow-up interval in their study from 20096. They found that approximately 80% of the 18 patients with a fracture reduction of ≤3 mm had a good or excellent Merle d’Aubigné score at an average follow-up of 4.1 years. Only 14% of the 14 patients with a fracture displacement of >3 mm had good or excellent scores6. With even greater displacements, the clinical results seem to deteriorate further. In 21 nonoperatively treated patients with a minimum initial displacement of 5 mm, Matta et al. reported that only 24% had a good or excellent Merle d’Aubigné score after a mean follow-up of 3.7 years14. That study also confirmed a good correlation between clinical and radiographic outcome.

Tornetta, in his series of 41 posterior wall fractures, showed that good to excellent results could be achieved in >90% of the patients, providing that the hip and the fracture pattern were stable when tested with the patient under general anesthesia21. Similar results were obtained by McNamara et al.22. The series described by Tornetta had a mean follow-up period of only 2.7 years. The average time to failure in our study was 6.9 years, with only 50% of the failures occurring within the first 4 years. This fact should be remembered when interpreting reported short-term results for acetabular fractures.

Fracture involvement of the weight-bearing dome, described as the roof arc, has been used as a predictor for outcome and an indication for surgery by several authors14,21,23. In our series, we could only show a significantly reduced survival of the hip when the posterior roof arc was <45°. For fracture lines in the upper 10 mm of the weight-bearing dome, there was a nonsignificant trend (p = 0.086) toward reduced hip survival when the fracture lines were situated in the posterior part of the joint.

Long-term register studies have inherent weaknesses. When we started our database, the use of radiographs was still the gold standard in the evaluation of acetabular fractures, and to ensure comparable data, we continued their routine use. Furthermore, we found it easier to standardize the measurements on radiographs, even though the use of CT scans gradually has become the examination of choice and has since proven to be an accurate method24,25. Also, the intraobserver and interobserver reliability of our measurements have not been validated

The use of the Harris hip score and the modified Merle d’Aubigné score to evaluate clinical outcomes is not optimal26,27; however, on the basis of the fact that most of the investigators reporting on this patient group used them6,16,28,29, we chose them as the primary outcome measures when we started the register in 1994. With the development of newer and better validated scores for both hip function and general patient satisfaction, we now include the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) and the Hip disability and Osteoarthritis Outcome Score (HOOS) as well as the EuroQol-5D (EQ-5D) in the follow-up of patients with an acetabular fracture.

The present study describes a cohort of patients with an acetabular fracture treated nonoperatively for various reasons, such as minimal joint displacement, secondary joint congruency, or medical contraindications to surgery. This creates a heterogeneous group of patients and fractures, and the results must be interpreted with this in mind. Even so, to our knowledge, the present study includes one of the largest cohorts of patients with nonoperatively treated acetabular fractures with long-term follow-up and provides useful insight regarding the outcome for this patient group over time.

Nonoperative treatment of acetabular fractures yields good to excellent long-term clinical and radiographic results for minimally displaced fractures. A loss of joint congruency with an intra-articular step-off of ≥2 mm leads to an increased rate of secondary osteoarthritis, a worsened functional outcome, and reduced survival of the native hip. The obturator oblique radiograph proved to be the most sensitive radiographic predictor.

For the acetabular fracture with a borderline surgical indication, the addition of oblique radiographic projections (Judet views) can be a useful tool in the decision-making process, as a fracture step-off of ≥2 mm has proven to be a strong predictor for an inferior long-term clinical and radiographic result. Acetabular fractures with a step-off of <2 mm can safely be treated nonoperatively. A greater step-off in the obturator oblique radiograph is prognostic of a poor clinical result for conservatively treated acetabular fractures.

Investigation performed at the Oslo University Hospital, Oslo, Norway

Disclosure: There was no external source of funding. The Disclosure of Potential Conflicts of Interest forms are provided with the online version of the article.

Back to Top | Article Outline


1. Laird A, Keating JF. Acetabular fractures: a 16-year prospective epidemiological study. J Bone Joint Surg Br. 2005 ;87(7):969–73.
2. Letournel E, Judet R. General principles of management of acetabular fractures. In: Elson RA, editor. Fractures of the acetabulum. 2nd ed. Berlin: Springer; 1993. p 347–61.
3. Olson SA, Zlowodzki M. Decision making: nonoperative and operative indications for treatment of acetabular fractures. In: Tile M, Helfet D, Kellam J, Vrahas M, editors. Fractures of the pelvis and acetabulum: principles and methods of management. 4th ed., vol. 2. Davos Platz, Switzerland: AOTrauma; 2015. p 481–96.
4. Lansinger O. Fractures of the acetabulum. A clinical, radiological and experimental study. Acta Orthop Scand Suppl. 1977;165:1–125.
5. Nandi SN, Manda P, Bhakta A. Result of conservative management in central fracture dislocation of hip. J Indian Med Assoc. 2012 ;110(7):481–4.
6. Sen RK, Veerappa LA. Long-term outcome of conservatively managed displaced acetabular fractures. J Trauma. 2009 ;67(1):155–9.
7. Rossvoll I, Finsen V. Mortality after pelvic fractures in the elderly. J Orthop Trauma. 1989;3(2):115–7.
8. Harper CM, Lyles YM. Physiology and complications of bed rest. J Am Geriatr Soc. 1988 ;36(11):1047–54.
9. Pecorelli F, Della Torre P. Fractures of the acetabulum: conservative treatment and results. Ital J Orthop Traumatol. 1987 ;13(3):307–18.
10. Hesp WL, Goris RJ. Conservative treatment of fractures of the acetabulum. Results after longtime follow-up. Acta Chir Belg. 1988 ;88(1):27–32.
11. Heeg M, Oostvogel HJ, Klasen HJ. Conservative treatment of acetabular fractures: the role of the weight-bearing dome and anatomic reduction in the ultimate results. J Trauma. 1987 ;27(5):555–9.
12. Gänsslen A, Hildebrand F, Krettek C. Conservative treatment of acetabular both column fractures: does the concept of secondary congruence work? Acta Chir Orthop Traumatol Cech. 2012;79(5):411–5.
13. Harris WH. Traumatic arthritis of the hip after dislocation and acetabular fractures: treatment by mold arthroplasty. An end-result study using a new method of result evaluation. J Bone Joint Surg Am. 1969 ;51(4):737–55.
14. Matta JM, Anderson LM, Epstein HC, Hendricks P. Fractures of the acetabulum. A retrospective analysis. Clin Orthop Relat Res. 1986 ;205:230–40.
15. Letournel E, Judet R. Radiology of the normal acetabulum. In: Elson RA, editor. Fractures of the acetabulum. 2nd ed. Berlin: Springer; 1993. p 29–61.
16. Letournel E. Acetabulum fractures: classification and management. Clin Orthop Relat Res. 1980 ;151:81–106.
17. Biau DJ, Latouche A, Porcher R. Competing events influence estimated survival probability: when is Kaplan-Meier analysis appropriate? Clin Orthop Relat Res. 2007 ;462:229–33.
18. Wongworawat MD, Dobbs MB, Gebhardt MC, Gioe TJ, Leopold SS, Manner PA, Rimnac CM, Porcher R. Editorial: estimating survivorship in the face of competing risks. Clin Orthop Relat Res. 2015 ;473(4):1173–6. Epub 2015 Feb 11.
19. Rowe CR, Lowell JD. Prognosis of fractures of the acetabulum. J Bone Joint Surg Am. 1961;43:30–92.
20. Dunet B, Tournier C, Billaud A, Lavoinne N, Fabre T, Durandeau A. Acetabular fracture: long-term follow-up and factors associated with secondary implantation of total hip arthroplasty. Orthop Traumatol Surg Res. 2013 ;99(3):281–90. Epub 2013 Apr 4.
21. Tornetta P 3rd. Non-operative management of acetabular fractures. The use of dynamic stress views. J Bone Joint Surg Br. 1999 ;81(1):67–70.
22. McNamara AR, Boudreau JA, Moed BR. Nonoperative treatment of posterior wall acetabular fractures after dynamic stress examination under anesthesia: revisited. J Orthop Trauma. 2015 ;29(8):359–64.
23. Øvre S, Madsen JE, Røise O. Acetabular fracture displacement, roof arc angles and 2 years outcome. Injury. 2008 ;39(8):922–31. Epub 2008 Jul 1.
24. Garrett J, Halvorson J, Carroll E, Webb LX. Value of 3-D CT in classifying acetabular fractures during orthopedic residency training. Orthopedics. 2012 ;35(5):e615–20.
25. Borrelli J Jr, Ricci WM, Steger-May K, Totty WG, Goldfarb C. Postoperative radiographic assessment of acetabular fractures: a comparison of plain radiographs and CT scans. J Orthop Trauma. 2005 ;19(5):299–304.
26. Ovre S, Sandvik L, Madsen JE, Roise O. Modification of the Harris hip score in acetabular fracture treatment. Injury. 2007 ;38(3):344–9. Epub 2006 Sep 1.
27. Nilsdotter A, Bremander A. Measures of hip function and symptoms: Harris Hip Score (HHS), Hip Disability and Osteoarthritis Outcome Score (HOOS), Oxford Hip Score (OHS), Lequesne Index of Severity for Osteoarthritis of the Hip (LISOH), and American Academy of Orthopedic Surgeons (AAOS) Hip and Knee Questionnaire. Arthritis Care Res (Hoboken). 2011 ;63(Suppl 11):S200–7.
28. Grimshaw CS, Moed BR. Outcomes of posterior wall fractures of the acetabulum treated nonoperatively after diagnostic screening with dynamic stress examination under anesthesia. J Bone Joint Surg Am. 2010 ;92(17):2792–800.
29. Matta JM. Fractures of the acetabulum: accuracy of reduction and clinical results in patients managed operatively within three weeks after the injury. J Bone Joint Surg Am. 1996 ;78(11):1632–45.
Copyright 2016 by The Journal of Bone and Joint Surgery, Incorporated