Complications after internal screw fixation of nondisplaced femoral neck fractures in elderly patients: A systematic review

Abstract

Objective

Internal fixation with cannulated screws for nondisplaced femoral neck fractures in the elderly has substantial reoperation and mortality rates. The selection of either internal fixation or arthroplasty for nondisplaced femoral neck fracture is debatable.

Methods

We performed a systematic review of the literature regarding complications in the internal fixation of nondisplaced femoral neck fractures in elderly (>60 years old) patients. We searched in multiple databases (PubMed, Web of Science, Embase, and Cochrane Library) for articles in this area; there was no limitation over the publication year.

Results

A total of 1971 patients were identified from 16 published studies. All these patients were over 60 years old. The minimum follow-up after the surgical procedure was 11 months (range: 11–183 months). A total of 329 fractures (16.7%) with radiographic and clinical failures after fixation were identified with regard to stable femoral neck fractures. The single most common complication after surgery was nonunion (129/329), with a pooled percentage of 39.2%. Osteonecrosis was found to be the second most common cause of revision surgery (31.9%). The overall reoperation rate attributable to surgical complications was 15.2% (300/1971 patients). Conversion to hip arthroplasty was performed in 244 patients (12.4%) after primary fixation.

Conclusion

Our study elucidated further the complication rate of nondisplaced femoral neck fractures treated with internal screw fixation. Since the failure rate of screw fixation for stable femoral neck fractures in elderly patients is not low, we believe that hemiarthroplasty is a reasonable treatment option in select patients.

Level of Evidence

Level III, Therapeutic study

There are approximately 312,000 hip fractures occurring annually in the United States population, and it is estimated that there will be 500,000 cases by 2040 (1). The incidence of elderly femoral neck fractures has been rising every year due to an increase in the average life span along with recent developments in medical technology (2). Problems after treating displaced femoral neck fractures have been extensively investigated, but less attention has been paid to nondisplaced fractures, even though the incidence of nondisplaced femoral neck fractures is approximately half that of displaced fractures (3).

Although the accepted treatment for nondisplaced hip fractures in young patients is osteosynthesis, this issue still remains highly controversial in the elderly population. Internal fixation is a common surgical intervention for the treatment of nondisplaced femoral neck fractures in the elderly population (4). However, a higher-than-expected number of internal-fixation-related complications have occurred after surgery in elderly patients due to poor bone quality (5, 6). Han et al. reported that the major complication rate of nondisplaced femoral neck fractures in patients over the age of 70 years was as high as 34.6% after an internal fixation operation (7). Several studies have reported that the clinical results of nondisplaced fractures treated by osteosynthesis are not so satisfactory in elderly patients due to the presence of osteonecrosis, nonunion, implant-related complications, or residual pain (6, 8). Nevertheless, our current knowledge regarding complications of internal screw fixation in elderly patients is confined to a few reviews from single institutions or from multicenter national registries from a single country, even though the incidence rates may differ between the patients of different ethnic backgrounds and nationalities.

The present study was designed to evaluate the complication rate of screw fixation for nondisplaced femur neck fractures in elderly patients with a review of the literature and a pooled analysis. Therefore, we asked the following questions. (1) What are the common postoperative complications requiring surgical revisions? (2) What are the overall complications and reoperation rates after fixation?

Materials and Methods

Literature research

We performed a systematic review of the available literature using multiple separate search strategies. Four computer databases (PubMed, Web of Science, Embase, and Cochrane Library) were searched with the phrases “femur neck fracture,” “nondisplaced,” “elderly,” and “fixation” in different combinations. Two independent reviewers separately completed this search, and the results were duplicated two times by each reviewer. The initial search was performed on May 10, 2019, with an update on July 10, 2019, to ensure accuracy. No additional study was identified when the search was repeated.

Selection criteria

The inclusion criteria for the studies were (1) randomized controlled trials, nonrandomized or quasi-randomized controlled trials, prospective cohort trials, or retrospective comparative studies; (2) English-written articles dealing with human species; (3) electronic publications that reported cases of complication after screw fixation; (4) articles reporting cases with elderly (more than 60 years old) patients; (5) both retrospective and prospective series; and (6) only those articles that investigated the final outcomes, including complications.

The exclusion criteria were (1) biomechanical or animal studies, (2) conference presentations, (3) abstracts only, (4) articles not reporting the postoperative follow-up period and particular complications (3), (5) extracapsular hip fracture [OTA/AO 31-B2.1 (9)], (6) displaced femoral neck fractures, and (7) articles that did not deal with the treatment of postoperative complications (10). Limits for the number of patients in each study or the minimum duration of follow-up were not used, and no limitation was imposed on the publication year. Only published clinical studies were included. Due to the limited evidence available on this topic, case series were also included in our study. For the study period, patients older than 60 years were defined as elderly. Fractures were categorized as nondisplaced if they were AO/OTA 31B3.1-3 (Garden stages I and II) (11).

Study selection

Searching the aforementioned databases yielded a total of 576 articles. A simplified flowchart enumerating this search process is shown in Figure 1. First, PubMed database yielded 259 articles; second, the Web of Science database-using the same search strategy-yielded 81 articles. There were 201 articles that appeared in more than one of the 4 searches, yielding a total of 375 unique articles. Abstracts and full texts of the retrieved articles were independently read by 2 authors, and all the relevant articles were read in their entirety. In addition, the reference lists of the included studies were screened to minimize the risk of missing relevant articles. In cases of disagreement, a consensus was reached through a discussion between both the authors. Stringent exclusion criteria were applied, ultimately yielding 16 appropriate articles. After including prospective studies, most of the larger cohorts giving an answer or at least an insight into the clinical problems were selected for this review. Preferred reporting items for systematic reviews and meta-analyses (PRISMA) guidelines were followed (12).

Figure 1.

A PRISMA flowchart illustrating the selection of studies included in our systematic review

Data extraction

We used a standard data extraction form to extract the relevant data from the eligible articles. Two reviewers independently extracted the data. If necessary, we contacted the corresponding authors of the included studies to ensure that the information was integrated as well as to retrieve missing data, if any. The following data were extracted. Level of evidence of the included studies, demographics (age and gender), underlying disease, classification of fractures (Garden stages), time to surgery after admission, follow-up period, intervention (technique and treatment protocol), outcomes subsequent to the treatment such as clinical union and postoperative complications, and mortality after surgery. As the measurements of the patient characteristics and outcomes were consistent among the trials, we pooled these data to determine the mean values. The pooling of the functional assessment datasets was not attempted because of significant variability in the criteria.

Results

Search results and studies included

Our systematic literature review from the PubMed, Web of Science, Embase, and Cochrane Library literature searches revealed a total of 1971 patients from 16 selected articles that were reported from 1996 to 2019, including two prospective studies (5, 13). Although complete data were not available, data such as age, gender, underlying medical/surgical conditions, and postoperative complications were collected. Age, mean follow-up time, and treatment outcomes were clearly identified in all the reports.

Demography

The patients who were treated for femur neck fractures between 1983 and 2015 were included in the present study. The mean age of the patients was 78.7 years (range: 60–106 years). The minimum follow-up period after the surgical procedure was 11 months (range: 11–183 months). A vast majority of the patients were female (female/male: 977/515). Two papers did not indicate the gender distributions (6, 11). It was difficult to analyze the time to surgery after admission due to reporting inconsistencies. Demographic information is listed in Table 1.

Table 1.

Demographic information in the studies (NA: not available)

Author	Journal	Year	Study period	Country	Number of patients	Male/Female	Mean age
Dolatowski et al.	J Bone Joint Surg Am.	2019	2012–2015	Norway	111	27/84	83.2
Lu et al.	Arch Orthop Trauma Surg	2017	2008–2010	China	41	12/29	85.85
Han et al.	Eur J Trauma Emerg Surg	2016	2002–2008	Korea	52	14/38	77.6
Min et al.	Hip Pelvis	2016	2008–2014	Korea	25	5/20	72.3
Kang et al.	BMC Musculoskeletal Disorders	2016	1996–2013	Korea	81	27/54	73.1
Kain et al.	Clin Orthop Relat Res	2014	2005–2008	USA	121	NA	80
Kim et al.	Arch Orthop Trauma Surg	2014	1999–2011	Korea	58	18/40	77.5
Kim et al.	Hip Pelvis	2014	2007–2010	Korea	33	12/21	77
Manohara et al.	Journal of Orthopaedic Surgery	2014	2004–2008	Singapore	100	23/77	78
Murphy et al.	Clin Orthop Relat Res	2013	1998–2009	USA	358	NA	81
Parker et al.	Injury	2008	1989–2002	UK	346	102/244	80.8
Bjørgul et al.	Acta Orthopaedica	2007	1998–2003	Norway	225	64/161	80.4
Yih-Shiunn et al.	International Orthopaedics	2007	1997–2003	Taiwan	84	35/49	71.6
Shimizu et al.	Arch Orthop Trauma Surg	2007	1983–2002	Japan	49	8/41	75.4
Chen et al.	The Journal of TRAUMA	2005	1996–2001	China	37	15/22	83.76
Chiu et al.	Arch Orthop Trauma Surg	1996	1983–1992	Taiwan	250	153/97	73

NA: not available

Treatment outcomes

All the trials reported postoperative complications such as nonunion, subsequent osteonecrosis, implant failure, and peri-implant fractures. Overall, there were a total of 329 fractures (16.7%) with radiographic and clinical failures after fixation for a stable femoral neck fracture. Nonunion (129/329) was determined to be the most common reason for reoperation in patients with nondisplaced fractures treated with internal fixation, where the pooled percentage was 39.2%. Osteonecrosis was the second most common cause of revision surgery (Table 2). The overall reoperation rate attributable to surgical complications was 15.2% (300 patients). Conversion to hip arthroplasty was performed in 244 patients (12.4%) after primary fixation (Table 3). The 1-year follow-up mortality rates were reported by 6 studies: these values ranged from 1.2% to 23%. Clinical outcomes assessed by the patient-reported outcomes were inconsistently reported, making the interpretation of these data challenging.

Table 2.

Postoperative complications

Author	Nonunion	Osteonecrosis	Implant failure	Peri-implant fracture	Deep infection	Others (Osteoarthritis, local pain)
Dolatowski et al.	10	7	7	1	1	0
Lu et al.	3	4	3	0	0	0
Han et al.	8	6	1	2	1	0
Min et al.	1	1	0	1	0	0
Kang et al.	4	6	4	0	0	0
Kain et al.	6	3	4	2	0	3
Kim et al.	0	4	1	0	0	0
Kim et al.	0	1	6	2	0	0
Manohara et al.	3	5	0	0	0	0
Murphy et al.	20	9	19	6	0	0
Parker et al.	38	13	0	2	0	0
Bjørgul et al.	16	10	7	2	1	6
Yih-Shiunn et al.	2	8	3	0	0	0
Shimizu et al.	2	6	0	0	0	0
Chen et al.	1	4	1	0	0	0
Chiu et al.	15	18	9	0	0	0
Total (%)	129 (39.2)	105 (31.9)	65 (19.7)	18 (5.4)	3 (0.9)	9 (2.7)

Table 3.

Reoperations

Author	Number of patients	Reoperation	Convertion to arthroplasty	One-year mortaility (%)
Dolatowski et al.	111	22	12	25 (23)
Lu et al.	41	8	7	NA
Han et al.	52	16	15	2 (3.8)
Min et al.	25	3	2	NA
Kang et al.	81	5	5	1 (1.2)
Kain et al.	121	16	12	NA
Kim et al.	58	4	4	4 (6.9)
Kim et al.	33	8	6	NA
Manohara et al.	100	8	8	4 (4.0)
Murphy et al.	358	54	43	NA
Parker et al.	346	50	37	NA
Bjørgul et al.	225	42	36	22 (9.8)
Yih-Shiunn et al.	84	8	8	NA
Shimizu et al.	49	8	8	NA
Chen et al.	37	6	6	NA
Chiu et al.	250	42	35	NA
Total (%)	1971	300 (15.2)	244 (12.4)

NA: not available

Discussion

For Garden stage I or II femoral neck fractures in patients less than 65 years old, internal fixation is generally preferred. However, the treatment method for nondisplaced femoral neck fractures in the elderly remains debatable and is often surgeon-dependent. The aim of this study was to evaluate the failure rate of internal fixation for stable femoral neck fractures in elderly patients with a pooled analysis of the reported cases.

The most common early postoperative complications occurring after the fixation of stable femur neck fractures are nonunion and failure of fixation (14, 15). The known risk factors for nonunion are screw fixation in a triangular configuration, displaced fracture, borderline and unacceptable reduction, and increased screw-shaft subchondral purchase over the femoral neck (16). Parker et al. demonstrated that increasing age and female sex were the primary risk factors for healing complications after femoral neck fractures due to osteoporosis (17). A number of studies have noted osteoporosis and age as risk factors for nonunion (16). In our pooled analysis, we have confirmed that the single most common complication after multiple screw fixation in elderly patients was nonunion (pooled percentage: 39.2%). Osteonecrosis of the femoral head is also a well-known complication of femur neck fractures, which is caused by alterations in the blood supply; it often develops 2–3 years after treatment (14). Therefore, follow-up should be maintained for at least 2 years after fixation, allowing the detection and treatment of avascular necrosis (15). Notably, it seems that younger patients exhibit greater vulnerability toward osteonecrosis of the femoral head than older individuals because patients aged ≤70 years were at an increased risk of developing osteonecrosis after nondisplaced femoral neck fractures treated with percutaneous screws (18, 19).

Despite these complications, internal fixation remains the conventional treatment of choice for nondisplaced femoral neck fractures even in elderly patients (20). It has several advantages over endoprosthetic replacement because it is less invasive and cost-effective (21). Chen et al. demonstrated that the overall union rate was 94.59% (35 patients) at 6 months after primary internal fixation in patients older than 80 years (22). Min et al. reported that a satisfactory level of bone union (96.0%) was achieved by performing internal fixation in patients over 65 years of age with stable femoral neck fractures (23). Internal fixation is a less invasive technique than arthroplasty and may be expected to reduce the mortality rate.

On the contrary, although the prognosis of stable fractures is favorable, the complication rate in elderly patients is not low and the frequency of reoperation is higher than that in arthroplasty (3, 24, 25). The failure rate of osteosynthesis is reported to be as high as ~20–35% in femoral neck fractures due to osteonecrosis, nonunion, and fixation failure; further, postoperative ambulation might be delayed in this population due to difficulty in achieving firm fixation (7, 26). This risk is greater for superaged patients, for whom the mortality rate after such a fracture is also higher. According to our systematic review, there were a total of 329 fractures (16.7%) with radiographic and clinical failures after fixation for a stable femoral neck fracture, and the overall reoperation rate attributable to surgical complications was 15.2%. Our data confirmed the current literature data according to which a higher-than-expected number of internal-fixation-related complications occur after surgery in elderly patients. There are numerous different opinions on the risk factors that cause such complications in internal fixation. Displacement, insufficient reductions, and severe osteoporosis could be the common causes of complications (7, 27). In particular, elderly Asian women have low bone density as compared to westerners, which may be a risk factor that can lead to osteoporotic fractures and show higher possibilities of nonunion and fixation failure after femoral neck fractures (28).

Successful management requires the surgeon to balance the potential risks and healthcare costs associated with surgery in the cases of femur neck fractures. One of these risks is reoperation. The reoperation rate after nondisplaced fracture treatment in elderly patients is a matter of concern and can lead to multiple hazards. Higher reoperation rates have been associated with the use of fixation in femur neck fractures, resulting in costlier treatments than those in hemiarthroplasty (29, 30). In the present study, the overall reoperation rate attributable to surgical complications was 15.2%. Therefore, in contemporary studies, increasing evidence supports the use of hemiarthroplasty over internal fixation in the elderly population in cases of stable femur neck fractures (5, 11). Dolatowski et al. demonstrated that elderly patients with a nondisplaced femoral neck fracture may benefit from being treated with latest-generation hemiarthroplasty rather than screw fixation (13). Hemiarthroplasty with less postoperative complications, low reoperation rate, and better function recovery at an early stage provides a worthwhile choice of treatment for elderly patients with nondisplaced femoral neck fractures (5). According to a recent meta-analysis study, there were no significant difference with respect to mortality rates for mid- or long-term follow-ups between internal fixation and arthroplasty (31). Furthermore, fixation is known to be associated with a higher risk of contralateral hip fracture than that in arthroplasty (32). Further multicenter studies are needed to establish objective treatment guidelines for stable femur neck fractures in superaged patients.

Our study has several limitations. First, most of the included studies were retrospective in nature, resulting in unavoidable reporting bias. Only 2 prospective cohort studies assessing a total of 152 patients were included in our systematic review. In addition, a few patients with short-term follow-up (the minimum follow-up period was 11 months) were included in our study. Second, although we utilized a systematic methodology that we believe could be reproducible, it may be possible that different search terms and data sources would have provided additional studies that could have met the selection criteria. Because most of the studies comprised patients with both nondisplaced and displaced fractures, it was difficult to isolate the data regarding nondisplaced fractures. Therefore, we established strict selection and exclusion criteria to reduce selection bias as much as possible. A strictly designed and adequately powered randomized controlled trial in the future is essential.

Conclusion

Based on this systematic review, the overall failure rate of internal fixation for nondisplaced femoral neck fractures in elderly patients was not low. The reasons for this higher frequency of reoperation may be related to poor bone quality, patient age, and certain technical factors. Although multiple screw fixation has been recommended in the literature for nondisplaced femoral neck fractures, we believe primary arthroplasty may be a better option for the treatment of nondisplaced fractures in select elderly patients. Additional prospective studies with the involvement of a larger number of patients with stable femur neck fractures (so that a wide range of ethnic backgrounds can be included) will help improve our ability to avoid devastating outcomes.

HIGHLIGHTS.

• The failure rate of internal fixation for stable femoral neck fractures in elderly patients is not low.

• The reoperation rate attributable to surgical complications was 15.2%.

• Primary arthroplasty may be a better option for nondisplaced fractures in select elderly patients.