Treatment of Intertrochanteric Femur Fractures with Long versus Short Cephalomedullary Nails



 

Cameron Sadeghi, MD1; Heather A. Prentice, PhD2; Kanu M. Okike, MD3; Elizabeth W. Paxton, PhD2

Perm J 2020;24:19.229 [Full Citation]

https://doi.org/10.7812/TPP/19.229
E-pub: 06/10/2020

ABSTRACT

Context: Prior studies regarding indications for long vs short cephalomedullary nails in the treatment of intertrochanteric fractures had limited sample sizes and follow-up, suggesting a need for further investigation.

Objective: To evaluate the association between cephalomedullary nail length and outcomes for the treatment of intertrochanteric femur fractures.

Design: Cohort study using Kaiser Permanente’s Hip Fracture Registry. A total of 5526 patients who underwent surgical treatment with cephalomedullary nails for an intertrochanteric femur fracture (2009-2014) were identified: 3108 (56.2%) with long nails and 2418 (43.8%) with short nails. Cox proportional hazards model regression was used to evaluate risks of all-cause revision and revision for periprosthetic fracture. Linear regression was used to evaluate operative time, estimated blood loss, and length of stay. Propensity score weights were used in all models to balance nail groups on patient and device characteristics.

Main Outcome Measures: All-cause revision surgery.

Results: No association was found in risk of all-cause revision (hazard ratio = 0.75, 95% confidence interval [CI] = 0.48-1.15) or revision for periprosthetic fracture (hazard ratio = 0.59, 95% CI = 0.23-1.48) for long nails compared with short nails. Use of longer nails resulted in 18.80 more minutes of operative time (95% CI = 17.33-20.27 minutes), 41.10 mL more of estimated blood loss (95% CI = 31.71-50.48 mL), and a longer hospitalization (8.4 hours; β = 0.35, 95% CI = 0.12-0.58 hours).

Conclusion: These findings suggest that routine use of short cephalomedullary nails is safe and effective in the treatment of intertrochanteric fractures.

INTRODUCTION

Cephalomedullary nails are the most common fixation device used in the treatment of intertrochanteric femur fractures.1 Cephalomedullary nails are commonly divided into long and short devices. Little consensus exists regarding indications for long vs short nails in the treatment of intertrochanteric fractures. Early designs of short cephalomedullary nails were associated with a higher risk of periprosthetic fracture, but with the newer generation of nails, this risk was reduced.2

Experts have recommended long nails for unstable and highly displaced fracture patterns because of the perceived protective effect of long nails in preventing refracture in this population of patients, who commonly have osteoporotic bones and a higher risk of falls.3 Historically, long nails were recommended to reduce the risk of future periprosthetic femur fractures. Results of prior studies have failed to identify differences in failure or periprosthetic fracture rates between long and short nails, but they demonstrated a reduced surgical time and blood loss for short nails.4-9 However, these studies have been limited by small sample sizes and short postoperative follow-up, suggesting a need for further investigation.

Using data from a US hip fracture registry (HFR), we sought to investigate the following questions: 1) Is there a difference in revision risk, either all-cause or for periprosthetic fracture, when using long vs short cephalomedullary nails in the treatment of intertrochanteric femur fractures? 2) Does operative time, estimated blood loss, or inhospital length of stay (LOS) differ for patients who receive a long vs short cephalomedullary nail in the treatment of intertrochanteric femur fractures?

METHODS
Study Design, Setting, and Sample

We conducted a retrospective cohort study using data from an integrated health care system’s HFR, the Kaiser Permanente Hip Fracture Registry. Kaiser Permanente covers more than 12.2 million members throughout 8 US geographical regions (ie, Colorado, Georgia, Hawaii, Mid-Atlantic, Northern California, Northwest, Southern California, and Washington). Members of the integrated health care system have been previously shown to be representative of the regional population served,10,11 increasing generalizability of study results.

The study sample was selected using the HFR and consisted of patients with a closed femur-based fracture of the intertrochanteric section who underwent primary fixation procedures from 2009 through 2014. Intertrochanteric fractures were identified using International Classification of Diseases, Ninth Revision (ICD-9) code 820.21 and adjudicated through implant information. Only fractures fixed with the 2 highest volume implants in the HFR, the Gamma3 Nailing System (Stryker Orthopaedics, Kalamazoo, MI) and the trochanteric femoral nail (TFN, DePuy Synthes, West Chester, PA), were included. The study consisted of 5526 hip fracture procedures: 3108 (56.2%) with long nails and 2418 (43.8%) with short nails.

Data Source

A detailed summary of data collection procedures for our HFR has been previously described.12 Started in 2009, this surveillance tool for all surgically treated hip fractures performed in 4 regions of our health care system (Hawaii, Northern California, Northwest, and Southern California) captures patient, procedure, implant, surgeon, and hospital information using the electronic health record (EHR), administrative claims data, health care membership data, other institutional databases, and mortality records. From 2009 to 2014, the registry included 26,873 procedures. Coverage is 100% for all surgically treated hip fractures performed in Kaiser Permanente. Outcomes are longitudinally monitored after the index procedure using electronic screening algorithms and are validated by trained clinical content experts using the EHR.

Treatment

The treatment of interest was length of the cephalomedullary nail. For this study, long nails were defined as those extending into the distal metadiaphysis and short nails as 170 to 180 mm in length. Nails that were not classified as long or short according to these definitions were excluded from the analysis. Data for all implants (including nails) were entered into the EHR at the time of implantation via a barcode scan. This detailed implant information is extracted from the EHR to the registry and is reviewed by clinical content experts, who classify each implant into its respective category.

Outcome

The primary outcome of interest was all-cause revision surgery. Revision was defined as any reoperation performed after the index procedure where an implant was exchanged. Secondary endpoints included revision for periprosthetic fracture, operative time (in minutes), estimated blood loss (in milliliters), and in hospital LOS (in days). Revision outcomes were time-to-event with follow-up time defined as the time from the index procedure date until the date of revision surgery, health care membership termination, death, or study end date (December 31, 2014), whichever came first. Date of membership termination and death for survival endpoints were treated as censored cases, with survival time based on the time those cases exited the study sample.

Statistical Analysis

Several potential confounders of treatment were considered, including age, body mass index (BMI), American Society of Anesthesiologists (ASA) classification, sex, race/ethnicity, and use of interlocking screws. Average treatment effects were estimated by incorporating inverse probability of treatment weights in a Cox proportional hazards regression for survival endpoints (revision surgery) and a linear regression model for continuous endpoints (operative time, estimated blood loss, LOS). The use of weights can induce dependence in the data, and an effective option for variance estimation is a nonparametric bootstrap.13 Given the presence of missing data on some covariates, for each bootstrap sample a single imputation (imputation model included all potential confounders, outcomes,14 and treatment) was performed15 under a fully conditional specification.16 Subsequently, inverse probability of treatment weights were calculated (ie, using a logistic regression model with the listed confounders as predictors of treatment assignment and a caliper restriction of 0.20 standard deviation [SD] of the logit propensity score17,18), and a treatment effect was estimated.

We used 500 bootstrap samples to calculate the treatment effect estimate (mean of the samples) and the variance of the estimate. Hazard ratios (HRs) for survival endpoints, b estimates for continuous endpoints, 95% normal-theory confidence intervals (CIs), and 2-sided p-values were reported. Analyses were performed using R version 3.3.0 software, and a = 0.05 was the statistical significance threshold used for this study.

RESULTS

Among the 5526 hip fracture procedures, there were 96 all-cause revisions (1.7%). Of the 96 revisions, 50 (1.6%) were among the 3108 cases using long nails and 46 (1.9%) were among the 2418 with short nails. Twenty-seven (0.5%) revisions were for periprosthetic fracture specifically: 13 (0.4%) among long nails and 14 (0.6%) among short nails. Operative times, estimated blood loss, and LOS data were available from 5493 (99.4%), 4696 (85.0%), and 5504 (99.6%) of the 5526 hip fractures, respectively. Mean (SD) operative time, estimated blood loss, and LOS for long vs short nail groups were as follows: 62.7 minutes (SD = 33.1) vs 47.4 minutes (SD = 22.8 minutes), 135.7 mL (SD = 151.7 mL) vs 99.8 mL (SD = 105.5 mL), and 5.57 days (SD = 4.43 days) vs 5.34 days (SD = 4.24 days).

Member terminations during the follow-up period included 201 patients (3.6%), and there were 2027 deaths (36.7%).

Cephalomedullary Nail Length and Revision Risk

Unadjusted cumulative incidence curves for time to all-cause revision stratified by nail length are displayed in Figure 1. Propensity score weights significantly reduced the imbalance of the devices on the covariates (all standardized differences < 0.01; Table 1). After the application of inverse probability of treatment weights, we failed to observe evidence of a difference in risk of all-cause revision when long nails were compared with short nails (HR = 0.75, 95% CI = 0.48-1.15, p = 0.186). A subgroup analysis based on conditional (regression-adjusted) proportional hazards models indicated a lack of evidence supporting a difference between Gamma3 and TFN (reference group) among long nails (HR = 1.13, 95% CI = 0.63-2.04, p = 0.675) and short nails (HR = 0.79, 95% CI = 0.39-1.61, p = 0.521), with respect to all-cause revision. The use of interlocking screws was not associated with a higher risk among long nails (HR = 0.90, 95% CI = 0.51-1.58, p = 0.703) or short nails (HR = 0.78, 95% CI = 0.35-1.76, p = 0.554; reference group = no interlocking screw). The primary reason for all-cause revision was fixation failure/symptomatic implant for both long and short nails, 56.0% and 52.2%, respectively (Table 2). A similar conclusion was reached when we looked specifically at risk of revision for periprosthetic fracture for long vs short nails (HR = 0.59, 95% CI = 0.23-1.48, p = 0.258).

19 229 figure1

19 229 table1

19 229 table2and3

Secondary Outcomes

In adjusted models, long nails were associated with 18.80 more minutes of operative time (95% CI = 17.33-20.27 minutes, p < 0.001; Table 3). There was 41.1 mL more of estimated blood loss for long nails compared with short nails (95% CI = 31.71-50.48 mL, p < 0.001), and patients who received a long nail had a longer inhospital LOS (8.4 hours; b = 0.35, 95% CI = 0.12-0.58 hours, p = 0.003).

DISCUSSION

Cephalomedullary nails are the most common method for stabilizing intertrochanteric femur fractures, but controversy exists regarding the indications for long vs short nails. Several studies have compared the use of long and short nails, but many of these have been limited by sample size and length of follow-up. Therefore, we evaluated the risk of all-cause revision for a large cohort of patients treated with long vs short cephalomedullary nails using the Kaiser Permanente Hip Fracture Registry. Even in our cohort of more than 5500 patients, we failed to observe a difference between nail length and risk of revision, but long nails did have a longer operative time, a greater estimated blood loss, and a longer LOS.

In general, we found high union rates and a low incidence of implant revision for intertrochanteric femur fractures with cephalomedullary nails: 1.9% for all-cause revision and 0.5% for revision due to periprosthetic fracture specifically. This is in line with results of recent studies that have reported the risk of ipsilateral femur refracture in the range of 0.5% to 10%.4-6,8,19,20 In a larger study sample, after adjusting for a number of potential confounders, our revision findings are consistent with results of prior studies with smaller sample sizes investigating long vs short nails.5,8,9,21-23 Most of these prior studies had sample sizes of around 200 patients, with the exception of the study by Liu et al,9 which included 899 patients. There has been concern reported regarding heterogeneity across implant devices.5 However, our results held even when stratifying the data by patients treated with Gamma3 vs TFN nails. The most common reason for revision surgery, regardless of nail length, was the presence of symptomatic implants. The registry lacked detail on whether this was due to a cut-out or removal of a painful implant, and this is a study limitation.

Our findings on operative time and estimated blood loss are similar to those of a prior report.5 Our LOS findings are similar to those of a prior study21 but contrast with another study that reported no difference in LOS.5 One reason for this discrepancy might be a longer procedure time for localizing and placing a distal interlocking screw in long nails. Furthermore, our analysis accounted for a number of patient and surgical factors through propensity score weighting before the evaluation of outcomes.

Although our findings for operative time, estimated blood loss, and LOS are statistically significant, we acknowledge that the differences observed in these outcomes by nail length may be of limited clinical significance. An increase in LOS by 8 hours for patients treated with long nails may be related to an inherent selection bias in which patients with more severe fracture patterns are more likely to be treated with a long nail. Higher rates of transfusion related to more blood loss have also been previously reported in prior studies, which could also prolong LOS.5,21 Our registry lacks detail on transfusions; therefore, this could not be investigated in the present study.

Our study is not without limitations. This study is observational in nature, and causation cannot be inferred. To mitigate confounding due to differences across implant designs outside of nail length, we restricted the study sample to the 2 highest volume designs in the HFR. Furthermore, although we attempted to account for potential confounders in our analysis, residual confounding due to unmeasured variables is a possibility. For example, we were unable to account for fracture type and surgeon decision making. A surgeon may judge a fracture, such as those with large posteromedial fragments or subtrochanteric extension, to be more unstable, and preferentially select a long cephalomedullary nail in the treatment of these fractures. We also were unable to perform radiographic review for the entire study sample but relied instead on diagnostic coding to identify patients with intertrochanteric fractures. However, we do not expect diagnosis coding of intertrochanteric fractures to depend on receipt of a long or short nail.

Study strengths include use of our institution’s hip fracture registry as the data source, which prospectively collects information on a predefined set of variables. Outcomes are longitudinally tracked using algorithms and are validated through manual chart review, which increases the internal validity of our study.

CONCLUSION

In a cohort of more than 5500 patients with intertrochanteric fractures who underwent hip fracture surgery using a cephalomedullary nail, we failed to observe a difference in risk of all-cause revision and revision for periprosthetic fracture regardless of use of long or short nails. Short cephalomedullary nails resulted in shorter operative times, with a lower estimated blood loss, and a shorter LOS. These findings suggest that a surgeon’s routine use of short cephalomedullary nails may be appropriate in the treatment of intertrochanteric fractures.

Disclosure Statement

The author(s) have no conflicts of interest to disclose.

Acknowledgments

We acknowledge the Kaiser Permanente orthopedic surgeons who contribute to the Kaiser Permanente Hip Fracture Registry as well as the Surgical Outcomes and Analysis Department staff, which coordinates registry operations. The authors also acknowledge Brian H. Fasig, PhD, for his ongoing support of the Kaiser Permanente Hip Fracture Registry and quality control management.

Kathleen Louden, ELS, of Louden Health Communications performed a primary copyedit.

Author Affiliations

1 Department of Orthopaedic Surgery, Southern California Permanente Medical Group, San Diego, CA, USA

2 Surgical Outcomes and Analysis, Kaiser Permanente, San Diego, California, USA

3 Department of Orthopaedic Surgery, Kaiser Moanalua Medical Center, Honolulu, HI, USA

Corresponding Author

Cameron Sadeghi, MD ()

References
1.    Lugovskaya N, Vinson DR. Paroxysmal atrial fibrillation and brain freeze: A case of Anglen JO, Weinstein JN. Nail or plate fixation of intertrochanteric hip fractures: changing pattern of practice. A review of the American Board of Orthopaedic Surgery Database. J Bone Joint Surg Am 2008 Apr;90(4):700-7.
    2.    Bhandari M, Schemitsch E, Jonsson A, Zlowodzki M, Haidukewych GJ. Gamma nails revisited: gamma nails versus compression hip screws in the management of intertrochanteric fractures of the hip: a meta-analysis. J Orthop Trauma. 2009 Jul;23(6):460-4.
    3.    Haidukewych GJ. Intertrochanteric fractures: ten tips to improve results. J Bone Joint Surg Am. 2009 Mar 1;91(3):712-9.
    4.    Hou Z, Bowen TR, Irgit KS, Matzko ME, Andreychik CM, Horwitz DS, et al. Treatment of pertrochanteric fractures (OTA 31-A1 and A2): long versus short cephalomedullary nailing. J Orthop Trauma. 2013 Jun;27(6):318-24.
    5.    Boone C, Carlberg KN, Koueiter DM, Baker KC, Sadowski J, Wiater PJ, et al. Short versus long intramedullary nails for treatment of intertrochanteric femur fractures (OTA 31-A1 and A2). J Orthop Trauma. 2014 May;28(5):e96-e100.
    6.    Kleweno C, Morgan J, Redshaw J, Harris M, Rodriguez E, Zurakowski D, et al. Short versus long cephalomedullary nails for the treatment of intertrochanteric hip fractures in patients older than 65 years. J Orthop Trauma. 2014 Jul;28(7):391-7.
    7.    Kanakaris NK, Tosounidis TH, Giannoudis PV. Nailing intertrochanteric hip fractures: short versus long; locked versus nonlocked. J Orthop Trauma. 2015 Apr;29 Suppl 4:S10-6.
    8.    Vaughn J, Cohen E, Vopat BG, Kane P, Abbood E, Born C. Complications of short versus long cephalomedullary nail for intertrochanteric femur fractures, minimum 1 year follow-up. Eur J Orthop Surg Traumatol. 2015 May;25(4):665-70.
    9.    Liu J, Frisch NB, Mehran N, Qatu M, Guthrie ST. Short-term Medical Complications Following Short Versus Long Cephalomedullary Nails. Orthopedics. 2018 Sep 1;41(5):e636-e42.
    10.    Karter AJ, Ferrara A, Liu JY, Moffet HH, Ackerson LM, Selby JV. Ethnic disparities in diabetic complications in an insured population. JAMA. 2002 May 15;287(19):2519-27.
    11.    Koebnick C, Langer-Gould AM, Gould MK, Chao CR, Iyer RL, Smith N, et al. Sociodemographic characteristics of members of a large, integrated health care system: comparison with US Census Bureau data. Perm J. 2012 Summer;16(3):37-41.
    12.    Inacio MC, Weiss JM, Miric A, Hunt JJ, Zohman GL, Paxton EW. A Community-Based Hip Fracture Registry: Population, Methods, and Outcomes. Perm J. 2015 Summer;19(3):29-36.
    13.    Austin PC. Variance estimation when using inverse probability of treatment weighting (IPTW) with survival analysis. Stat Med. 2016 Dec 30;35(30):5642-55.
    14.    White IR, Royston P. Imputing missing covariate values for the Cox model. Stat Med. 2009 Jul 10;28(15):1982-98.
    15.    Shao J, Sitter RR. Bootstrap for Imputed Survey Data. J Am Stat Assoc. 1996;91(435):1278-88.
    16.    van Buuren S. Multiple imputation of discrete and continuous data by fully conditional specification. Stat Methods Med Res. 2007 Jun;16(3):219-42.
    17.    Hong G. Marginal mean weighting through stratification: Adjustment for selection bias in multilevel data. J Educ Behav. 2010;35:499-531.
    18.    Hong G. Marginal mean weighting through stratification: A generalized method for evaluating multivalued and multiple treatments with nonexperimental data. Psychol Methods. 2012 Mar;17(1):44-60.
    19.    Okcu G, Ozkayin N, Okta C, Topcu I, Aktuglu K. Which implant is better for treating reverse obliquity fractures of the proximal femur: a standard or long nail? Clin Orthop Relat Res. 2013 Sep;471(9):2768-75.
    20.    Lindvall E, Ghaffar S, Martirosian A, Husak L. Short Versus Long Intramedullary Nails in the Treatment of Pertrochanteric Hip Fractures: Incidence of Ipsilateral Fractures and Costs Associated With Each Implant. J Orthop Trauma. 2016 Mar;30(3):119-24.
    21.    Krigbaum H, Takemoto S, Kim HT, Kuo AC. Costs and Complications of Short Versus Long Cephalomedullary Nailing of OTA 31-A2 Proximal Femur Fractures in U.S. Veterans. J Orthop Trauma. 2016 Mar;30(3):125-9.
    22.    Hulet DA, Whale CS, Beebe MJ, Rothberg DL, Gililland JM, Zhang C, et al. Short Versus Long Cephalomedullary Nails for Fixation of Stable Versus Unstable Intertrochanteric Femur Fractures at a Level 1 Trauma Center. Orthopedics. 2019 Mar 1;42(2):e202-e9.
    23.    Shannon SF, Yuan BJ, Cross WW 3rd, Barlow JD, Torchia ME, Holte PK, et al. Short Versus Long Cephalomedullary Nails for Pertrochanteric Hip Fractures: A Randomized Prospective Study. J Orthop Trauma. 2019 Oct;33(10):480-6.

Keywords: cephalomedullary nail, hip fracture, intertrochanteric fracture, revision surgery, registry

ETOC

Click here to join the eTOC list or text ETOC to 22828. You will receive an email notice with the Table of Contents of The Permanente Journal.

CIRCULATION

2 million page views of TPJ articles in PubMed from a broad international readership.

Indexing

Indexed in MEDLINE, PubMed Central, EMBASE, EBSCO Academic Search Complete, CrossRef, and SciVerse/Scopus.


                                             

 

 

ISSN 1552-5767 Copyright © 2020 thepermanentejournal.org.

All Rights Reserved.