Utilization Trends, Patient-Demographics, and Comparison of Medical Complications of Sliding Hip Screw or Intramedullary Nail for Intertrochanteric Fractures: A Nationwide Analysis from 2005 to 2014 of the Medicare Population

Abstract

Background

Studies demonstrate an increase incidence of intertrochanteric fractures within the United States. Matched studies evaluating intertrochanteric fractures managed with either sliding hip screw (SHS) or intramedullary nail (IMN) within the Medicare population are limited. The purpose of this study was to investigate: 1) annual utilization trends; 2) patient demographics; and 3) complications including mortality.

Methods

A retrospective query using a nationwide database was performed. Patients undergoing SHS or IMN for intertrochanteric fractures were identified. The query yielded a total of 37,929 patients utilizing SHS (n = 11,665) or IMN (n = 26,264). Patients were matched 1:1 based on comorbidities. Primary outcomes included: utilization trends, patient demographics, 90-day complications, and 90-day readmission rates. Linear regression analyses were used to compare utilization trends. Pearson’s c2 analyses were used to compare patient-demographics, medical complications, and 90-day readmission rates. A p-value less than 0.05 was considered statistically significant.

Results

Linear regression analysis demonstrated a statistically significant decrease in utilization of SHS for IT fractures (p<0.0001); whereas utilization for IMN stayed consistent (p=0.36). IMN had significantly higher prevalence of comorbidities compared to SHS, notably, hyperlipidemia (70.6 vs. 62.6%; p<0.0001). Based on 1:1 match, IMN patients had significantly higher rates of 90-day medical complications, such as respiratory failure (11.0 vs. 8.1%; p<0.0001) and VTE (4.2 vs. 3.2%; p<0.001; however, there was not a statistical difference in postoperative infection (1.4 vs. 1.5%, p=0.06). There was no statistical difference in 90-day mortality between IMN and SHS cohorts (0.19 vs .13%, p = 0.249).

Conclusion

This analysis demonstrates a difference in utilization of SHS and IMN for patients with IT fractures. Patients with IMN had significantly higher prevalence of comorbid conditions and incidence of 90-day postoperative complications compared to SHS patients. The study can be utilized by orthopaedic surgeons to potentially anticipate healthcare utilization depending on implant selection.

Level of Evidence: III

Introduction

Intertrochanteric femur fractures are a common injury pattern in the United States as we observe a growing elderly population with subsequent age-related bone changes, drawing attention to surgical management, cost trends, patient specific factors, morbidity, and mortality. Annual incidence of 150,000 has been reported.^1,2 Furthermore, the incidence of elderly hip fractures and cost are modeled to increase from 2005 to 2025 by 51% and 49%, respectively.³ Conversely, some studies have reported a decrease in incidence of hip fractures and subsequent mortality with an increase in comorbidities between 1995 and 2005, which has coincided with increasing use of bisphosphonate therapy in this patient population.^1,4-9

The possibility of increasing incidence has raised concern over the burden of cost of management as hip fractures contribute disproportionately to annual expenditure compared to other osteoporotic fractures, $6 billion USD per year.^7,10-16 It has been reported that patients typically require $40,000 in medical resources in the first year after injury and around $5000 in the following years.^3,17,18 The preferred treatment for these fractures remains surgical with the choice of modality primarily based on fracture pattern and surgeon preference. Currently, the two mainstay implant options in treating intertrochanteric femur fractures are the sliding hip screw (SHS) with side plate, introduced in the 1950s, and the intramedullary nail (IMN) with an SHS component, also referred to as the cephallomedullary nail (CMN), which came into common practice in the 1990s.^10,19-21

A national trend exists for younger surgeons preferring IMN over SHS despite lacking a strong body of evidence supporting the benefit of this transition.^2,9,10,22 It has been recently reported that 92.4% of IT fractures have been estimated to be managed with IMN, representing a 49.1% increase in the study period.²² Advantages of the IMN are thought to provide an internal buttress effect against fracture collapse and medialization of the distal fracture component, especially in unstable fracture patterns. Additionally, IMN offers limited soft tissue insult by percutaneous insertion in an era of minimally invasive focus. However, use of IMN traditionally have posed the risk of intraoperative and postoperative femur fracture, which may have been improved by the development of more anatomically conforming implant designs and increased surgeon familiarity with technique.^23-31 Despite these proposed advantages, there is limited literature to prove the superiority of this implant. In fact, randomized trials have described similar outcomes in treatment with IMN when compared to SHS.^25-29,31-37

Current American Academy of Orthopedic Surgeons (AAOS) recommendations include both SHS and IMN as suitable implants for treating stable IT fractures, and IMN as superior in managing unstable fracture patterns due to the aforementioned biomechanical advantages.^20,28,38-40 However, the implant cost of IMN remains approximately 2-3 times higher than SHS, when both devices are appropriate options for management of stable fractures, information that may not be common knowledge of the operating surgeon.^30,38,41,42

Given current implant trends, cost considerations, and changing patient population factors, it is prudent to continue to further investigate outcomes surrounding the two treatment options. Therefore, the purpose of this study was to analyze a comprehensive administrative database and investigate patients undergoing SHS or IMN for intertrochanteric fractures. Specifically, we evaluated: 1) annual utilization trends; 2) patient-demographics; and 3) complications including mortality.

Methods

Database

A query from January 1st, 2005 to March 31st, 2014 was performed using the Medicare Standard Analytical Files from the PearlDiver (PearlDiver Technologies, Fort Wayne, Indiana) supercomputer. PearlDiver is a commercially available database which has been used extensively for orthopedic-related research as it contains the records of over 100 million patients from the Humana and Medicare claims databases. The database relies on International Classification of Disease, ninth revision (ICD-9), ICD-10, and Current Procedural Terminology (CPT) coding to query complications, diagnoses, reimbursements, discharge dispositions, in addition to other metrics. Since the database provides anonymous patient information the study was exempt from the institution’s Institutional Review Board (IRB) approvals.

Study Group

The database was first queried for all patients who sustained closed intertrochanteric fractures and were queried using ICD-9 diagnosis code 801.20 to 801.22. The database was then queried for all patients having undergone SHS or IMN fixation using CPT codes 27244 and 27245, respectively. The query yielded a total of 37,929 patients utilizing SHS (n = 11,665) or IMN (n = 26,264) for intertrochanteric fractures during the study interval. Patients were then matched on a 1:1 basis based on medical comorbidities, which included age, sex, tobacco use, hypertension, chronic obstructive pulmonary disease, body mass index, hyperlipidemia and major depressive disorder. After 1:1 matching, the query yielded 23,236 total patients with an equal number in each group (n=11,618).

Outcomes Assessed

Outcomes analyzed in this study included annual utilization trends, patient demographics, 90-day medical complications, and 90-day readmission rates between the two cohorts. Patient demographics analyzed and compared included: age, sex, and medical comorbidities – body mass index (BMI), congestive heart failure, chronic obstructive pulmonary disease, coagulopathies, depression, diabetes mellitus, electrolyte/fluid imbalance, hyperlipidemia, hypothyroidism, peptic ulcer disease, peripheral vascular disease, opioid use disorder, renal failure, rheumatoid arthritis, sleep apnea, and tobacco use between the two cohorts.

Ninety-day medical complications analyzed and compared included post-operative infections, irrigation and debridement, myocardial infarction, respiratory failure, urinary tract infections, pneumonia, cholecystitis, cerebrovascular accidents, and transfusions of blood products. Ninety-day medical post-operative complications were also analyzed for patients that were matched 1:1, which also included the same variables discussed above in addition to 90-day mortality.

Data Analyses

Statistical analyses were performed using the open programming language R (R Foundation for Statistical Computation, Vienna, Austria). Linear regression analyses were used to compare changes in rates of the two procedures for intertrochanteric fractures through the study interval. Patient demographics of age, sex, and medical comorbidities were analyzed and compared using Pearson’s c² analyses. Medical complications and 90-day readmission rates were also analyzed utilizing Pearson’s c² analyses. A p-value less than 0.05 was considered statistically significant.

Results

Patient Demographics

Intramedullary nail for intertrochanteric fractures were utilized more often across all age groups compared to sliding hip screw in a statistically significant manner (p<0.0001), however distribution of intramedullary nail and sliding hip screw were similar with respect to sex (p=0.316). Patients with intramedullary nail for intertrochanteric fractures were found to have significantly higher frequency of comorbid conditions compared to SHS, except for congestive heart failure and renal failure; however, the frequency of the latter comorbid conditions did not reach statistical significance. Comorbid conditions with the greatest prevalence in the IMN cohort included hypertension (95.0%), electrolyte/fluid imbalance (81.4%), and hyperlipidemia (70.6%); whereas the greatest comorbid conditions seen in the SHS cohort were hypertension (94.4%), electrolyte/fluid imbalance (80.5%), and peripheral vascular disease (65.9%) (Table 1).

Table 1.

Patient Demographics

	Intramedullary Nail		Sliding Hip Screw
Demographics	n	%	n	%	p-valueD
Age (Years)					<0.0001
<64	1,071	4.1	514	4.4
65 – 69	1,484	5.7	615	5.3
70 – 74	2,200	8.4	886	7.6
75 – 79	3,486	13.3	1,633	14.0
80 – 84	5,616	21.4	2,554	21.9
85>	12,337	47.0	5,422	46.5
Unknown	70	0.3	41	0.4
Sex					0.316
Female	19,142	72.9	8,463	72.6
Male	7,052	26.9	3,161	27.1
Unknown	70	0.3	41	0.4
Comorbidities
Alcohol Abuse	1,612	6.1	705	6.0	0.741
BMI <19kg/m2	1,703	6.5	617	5.3	<0.0001
BMI 19 – 24kg/m2	1,206	4.6	367	3.1	<0.0001
BMI 25 – 29kg/m2	778	3.0	214	1.8	<0.0001
BMI 30 – 39kg/m2	776	3.0	257	2.2	<0.0001
BMI 40 – 70kg/m2	254	1.0	87	0.7	0.04
CHF	15,745	59.9	7,099	60.9	0.09
Coagulopathies	7,933	30.2	3,263	28.0	<0.0001
COPD	1,444	5.5	556	4.8	0.003
Depression	13,428	51.1	5,699	48.9	<0.0001
Diabetes Mellitus	13,350	50.8	5,625	48.2	<0.0001
Electrolyte/Fluid Imbalance	21,380	81.4	9,389	80.5	0.03
Hyperlipidemia	18,532	70.6	7,308	62.6	<0.0001
Hypertension	24,950	95.0	11,007	94.4	0.01
Hypothyroidism	12,756	48.6	5,336	45.7	<0.0001
Opioid Use Disorder	461	1.8	189	1.6	0.37
Peptic Ulcer Disease	3,219	12.3	1,329	11.4	0.01
Peripheral Vascular Disease	17,830	67.9	7,683	65.9	<0.0001
Renal Failure	5,334	20.3	2,400	20.6	0.56
Rheumatoid Arthritis	5,159	19.6	2,000	17.1	<0.0001
Sleep Apnea	2,139	8.1	793	6.8	<0.0001
Tobacco Use	4,471	17.0	1,736	14.9	<0.0001

Demographics of Patients Utilizing Either Intramedullary Nail or Sliding Hip Screw for Intertrochanteric Fractures within the Medicare Population from 2005 to 2014. BMI = Body Mass Index; CHF = Congestive Heart Failure; COPD = Chronic Obstructive Pulmonary Disease; D = Assessed by Pearson’s c²; * = <11 Patients.

Annual Utilization

The number of intertrochanteric fractures decreased through the study interval from 122,481 in 2005 to 99,607 by first quarter of 2014 (p<0.0001). During this same time, annual utilization rates of intramedullary nail for intertrochanteric fractures increased 21.19% through the study interval (2005 = 2,137; 2014 = 2,590; p=0.59), but failed to reach statistical significance. Annual utilization rates of SHS for intertrochanteric fractures decreased 80.81% during the study interval (2005 = 2,544; 2014 = 488; p<0.0001) (Figure 1).

Figure 1.

Annual Utilization Trends of Intramedullary Nails vs Sliding Hip Screws. Linear representation of the annual utilization rates (2005-2014)0of intramedullary nail or sliding hip screw for treated patients with intertrochanteric fractures within the sampled Medicare population.

90-Day Complications

Patients in the IMN cohort were found to have significantly higher incidence of medical complications compared to patients within the SHS cohort. IMN cohort patients were found to have significantly higher incidences of cerebrovascular accidents (6.4 vs. 5.3%, p<0.0001), cholecystitis (0.2 vs. 0.2%; p=0.005), deep vein thromboses (3.8 vs. 2.8%; p<0.0001), myocardial infarctions (3.0 vs. 2.3%; p<0.0001), pneumonias (3.1 vs. 2.5%; p<0.0001), pulmonary embolisms (1.0 vs. 0.8%, p<0.0001), requiring transfusion of blood products (7.0 vs. 5.9%; p<0.0001) and urinary tract infections (27.5 vs. 23.3%, p<0.0001) within 90-days following the index procedure (Table 2).

Table 2.

Medical Complications at 90-Days

	Intramedullary Nail		Sliding Hip Screw
Medical Complications	n	%	n	%	p-valueD
Cerebrovascular Accidents	1,679	6.4	613	5.3	<0.0001
Cholecystitis	64	0.2	26	0.2	0.005
Deep Vein Thromboses	1,011	3.8	325	2.8	<0.0001
Irrigation and Debridement	102	0.4	58	0.5	<0.0001
Myocardial Infarction	780	3.0	268	2.3	<0.0001
Pneumonias	803	3.1	286	2.5	<0.0001
Post-operative Infections	392	1.5	177	1.5	<0.0001
Pulmonary Embolisms	266	1.0	93	0.8	<0.0001
Respiratory Insufficiencies	3,093	11.8	964	8.3	<0.0001
Transfusions	1,839	7.0	693	5.9	<0.0001
Urinary Tract Infections	7,225	27.5	2,722	23.3	<0.0001

Comparison of 90-Day Medical Complications in Patients Undergoing Intramedullary Nail or Sliding Hip Screw for Intertrochanteric Fractures within the Medicare Population. D = Assessed by Pearson’s c².

Patients undergoing IMN for stable IT fractures were found to have higher incidence of cerebrovascular accidents (6.4 vs 5.3%, p<0.0001), DVT (3.8 vs 2.8% p<0.0001), MI (3.0 vs 2.3%, p<0.0001), pneumonia (3.1 vs 2.5%, p<0.0001), post-operative infection (1.5 vs 1.5%, p<0.0001), PE (1.0 vs 0.8%, p<0.0001), respiratory insufficiencies (11.8 vs 8.3%, p<0.0001), transfusions (7.0 vs 5.9%, p<0.0001), and UTI (27.5 vs 23.3%, p<0.0001). However, patients undergoing SHS had significantly higher incidence of irrigation and debridement (0.5 vs 0.4%, p<0.0001).

90-day medical complications and 90-day mortality was also identified after 1:1 matching based on medical comorbidities. Patients undergoing IMN had statistically higher incidence of DVT (3.20 vs 2.33%, p<0.0001), MI (2.82 vs 2.27%, p=0.007), VTE (4.18 vs 3.24%, p<0.0001), respiratory insufficiencies (11.02 vs 8.14%, p<0.0001), and UTIs (26.8 vs 22.3%, p<0.0001). Overall, patients in the IMN cohort had significantly greater incidence of 90-day complications (60.71 vs 50.56%, p<0.0001) (Table 3). There was no significant difference in 90-day mortality between the two groups (0.19 vs 0.13%, p=0.249) (Table 3).

Table 3.

Medical Complications at 90-Days (Matched)

	Intramedullary Nail		Sliding Hip Screw
Medical Complications	n	%	n	%	p-valueD
Post-Operative Infection	164	1.41	173	1.49	0.0621
Irrigation and Debridement	45	0.39	56	0.48	0.272
Deep Vein Thromboses	372	3.20	271	2.33	<0.0001
Cerebrovascular Accidents	720	6.20	633	5.45	0.0148
Myocardial Infarc- tion	328	2.82	264	2.27	0.007
Pneumonias	333	2.87	279	2.40	0.026
Venous Thrombus Embolism	486	4.18	376	3.24	<0.0001
Pulmonary Embolisms	117	1.01	97	0.83	0.169
Respiratory Insufficiencies	1280	11.02	946	8.14	<0.0001
Transfusions	90	0.77	69	0.59	0.094
Urinary Tract Infections	3118	26.8	2710	22.3	<0.0001
Mortality	22	0.19	15	0.13	0.249
Total	7053	60.71	5478	50.56	<0.0001

Comparison of 90-Day Medical Complications in patients matched 1:1 for comorbidities undergoing Intramedullary Nail or Sliding Hip Screw for Intertrochanteric Fractures within the Medicare Population. D = Assessed by Pearson’s c².

Discussion

Intertrochanteric femur fractures are commonly encountered by orthopedic surgeons in all subspecialties with an annual incidence of 150,000 in the United States alone1. With an aging population and increased rates of osteoporosis, it is projected that the incidence of IT fractures continues to grow, though these projections have not fully been realized as recent studies have conflicting evidence for increased incidence of IT fractures.^2,3,9,22 Treatment for these fractures remains operative, with surgeon choice of sliding hip screw versus intramedullary nail.²⁰ The main difference between implant choices remains theoretic superiority of IMN in unstable patterns as well as cost, with IMN costing roughly 2-3 times that of SHS.^30,38,41 Despite this cost difference and lack of clinical evidence of superiority, the rates of IMN usage continue to increase.^10,25,28,37 The current study demonstrates the continued preference of orthopedic surgeons for use of intramedullary nail for treatment of intertrochanteric femur fractures over sliding hip screws. Additionally, when accounting for underlying comorbid conditions, this study found patients treated with IMN had higher 90 day complication rates as compared to those treated with SHS; however, there was no significant difference in 90-day mortality rates between the two groups. To our knowledge, this is the first study to compare outcomes in matched patients undergoing IMN compared to SHS for stable IT fractures in the Medicare population with a large sample size in each group (n=11,618).

The trend toward use of IMN for IT fractures was first identified by Anglen et al. who identified a dramatic increase of IMN usage from 3% in 1999 to 67% in 2006 using the ABOS Part II database.¹⁰ Smith et al. confirmed the continuation of this trend in young orthopedic surgeons through 2017, at which time 92.4% of IT femur fractures were treated with IMN.²² This trend has not only been observed in young orthopedic surgeons. Studies by Radcliff et al. and Werner et al. demonstrated a very similar trend in patients at Veteran’s Affairs hospitals as well as the Medicare patient population.^9,43 The study presented here serves to demonstrate the continuation of the trend in Medicare patients as first demonstrated by Werner et al. The final five years of study seem to indicate a plateau reached in the use of IMN, which may be the result of hitting a “floor” in the use of SHS as suggested by Smith et al.

There are several hypothesized reasons for the increased incidence in the use of IMN. Forte et al. demonstrated younger surgeons and those that work at teaching hospitals were more likely to opt for IMN.⁴⁴ Further studies have shown that surgeons who worked at their institution for shorter periods, specialties outside of lower limb trauma, and those who had exposure to IMN in training were more likely to use IMN.^45,46 Given these findings, it is not surprising that young surgeons coming out of training are more likely to use intramedullary devices as this is a familiar device. This may explain the trend noted in the ABOS part II.^10,22 No current study is available that reviews current orthopedic surgery resident case numbers when comparing these treatment options, but may be an interesting area of future study.

This study identified that 90 day medical complication rates are increased in the IMN group. This is in agreement with the previous studies by Smith et al. and Werner et al.^9,22 In addition to medical complications, some studies have also shown increases in 30 day mortality for patients treated with IMN.⁴⁷ These findings may be multifactorial and could be related to patient selection, as this study demonstrated increased co-morbidities among the IMN group. Although, when controlling for these variables, complication rates remained significantly elevated. There may be characteristics inherent to the device that contribute to this as well. Reaming of the intramedullary canal introduces increased intramedullary pressure that has been demonstrated to increase fat intravasation which may lead to pulmonary emboli.^48,49 This may explain the higher rates of pulmonary emboli and cerebrovascular accidents, but does not fully explain the remainder of the medical complications. Increased operative time and blood loss have also been implicated as a reason for the increase in medical complications9. However, the current literature does not seem to demonstrate a clear increase in blood loss or operative time when comparing these two treatment options.^50,51

The 90-mortality rate between the IMN and SHS cohorts were also compared in this study. The findings in this study were consistent with Werner et al., where there was no statistical difference in the 90-day mortality.⁵² However, there does not appear to be a consensus in the literature regarding mortality rates in patients undergoing IMN vs SHS for IT fractures. Whitehouse et al. analyzed the 82,990 patients from the National Hip Fracture database and found a 12.5% increase in 30 day mortality for patients undergoing IMN compared to SHS.⁴⁷ On the other hand, Grønhaug et al. demonstrated significantly lower one year mortality rates in patients undergoing treatment of unstable and stable IT fractures with IMN using the Norwegian Hip Fracture Register.⁵³

While the large number of patients ensures the study was well-powered, the current study is not without limitations; most of which are inherent to the use of an administrative database. We analyzed only a single insurance database, and the results of the study might not be a true cross-sectional representation on the trends of treating intertrochanteric fractures in the United States. Additionally, the study is reliant on accurate diagnostic and procedural coding, and it is currently estimated that there are up to 1.3% of coding errors within the Medicare database. Furthermore, the database was unable to provide additional granular detail such as radiographic imaging to determine whether patients had stable or unstable intertrochanteric fractures, as the fracture classification is vital in driving management on treating these fracture patterns. Finally, clinical outcome data is not available through the database that was utilized, which limits the ability to comment on outcomes.

Conclusion

There remains a trend toward the use of IMN over SHS for the treatment of intertrochanteric femur fractures despite evidence to support superiority of the implant. Given the increased cost associated with the implant and increased rates of 90 day medical complications, this study can assist surgeons and consoling patients on possible outcomes depending on treatment modality of stable intertrochanteric femur fractures.