Admission to rehab increases risk of postoperative venous thromboembolism and bleeding after operative fixation of femoral neck fractures

Hannah I. Travers; Gloria S. Coden; Samantha Simon; Mikhail Kuznetsov; Eric L. Smith; Rushad Patell; Brian L. Hollenbeck

doi:10.1016/j.jor.2025.01.032

View/Download PDF

Buy Reprints

PDF

Translate this page into:

68 (); 114-120

doi:

10.1016/j.jor.2025.01.032

Admission to rehab increases risk of postoperative venous thromboembolism and bleeding after operative fixation of femoral neck fractures

Hannah I. Travers^a,, Gloria S. Coden^a, Samantha Simon^a, Mikhail Kuznetsov^a, Eric L. Smith^a, Rushad Patell^b, Brian L. Hollenbeck^a

a Department of Orthopedic Surgery, New England Baptist Hospital, MA, 02120, Boston, USA

b Department of Hematology-Oncology, Beth Israel Deaconess Medical Center, MA, 02120, Boston, USA

⁎Corresponding author: Hannah I. Travers. travers.hannah@gmail.com

Received: 2025-1-6, Accepted: 2025-1-26, Published: 2025-2-13

Disclaimer:
This article was originally published by Reed Elsevier India Pvt. Ltd. and was migrated to Scientific Scholar after the change of Publisher.

Abstract

The risk of postoperative venous thromboembolism (VTE) and bleeding after operative management of femoral neck fractures (FNF) is well established. It is common for FNF patients to be discharged to rehabilitation facilities, despite higher complication rates. It was hypothesized that discharge to rehabilitation after operative fixation of FNF was associated with increased risk of VTE and bleeding.

Retrospective review of 7483 patients who sustained FNF and underwent operative treatment between 1/1/2019-12/31/2021 was conducted using a commercial claims database. The primary outcome was VTE and bleeding post-discharge within 30 days of surgery. Risk factors for VTE and bleeding were evaluated by logistic regression analysis.

The cumulative incidence was 3.77 % (95 % confidence interval (CI) = 3.33–4.21) for VTE and 4.01 % (CI = 3.56–4.46) for bleeding. There were 56.7 % of patients who were discharged to rehabilitation. Multivariable analysis demonstrated hereditary hypercoagulable diagnosis (odds ratio (OR) = 2.86, CI = 1.33–6.16), discharge to rehabilitation (OR = 2.08, CI = 1.47–2.92), discharge to another location (OR = 1.94, CI = 1.15–3.27), and length of stay (LOS) ≥5 days (OR = 1.69, CI = 1.15–2.50) increased risk for postoperative VTE. Variables that increased risk of bleeding included discharge to rehabilitation (OR = 2.20, CI = 1.55–3.12), discharge to other (OR = 1.92, CI = 1.12–3.27), and chronic anticoagulation (OR = 1.58, CI = 1.19–2.10). Analysis of patients prescribed thromboprophylactic medications demonstrated cumulative incidence of VTE 3.46 % (CI = 2.24–4.68) and 4.47 % (CI = 3.09–5.85) for bleeding at 30 days.

In patients with FNF, discharge to inpatient rehabilitation was associated with increased risk of postoperative VTE and bleeding. LOS ≥5 days was also associated with increased VTE risk, demonstrating the importance of timely hospital discharge.

1 Introduction

Femoral neck fracture (FNF) is a significant driver of morbidity and mortality in the aging population,¹ with postoperative venous thromboembolism (VTE) following surgical intervention a well-established and potentially devastating complication.²^,³ Both the American Academy of Orthopedic Surgeons and American College of Chest Physicians (ACCP) recommend VTE prophylaxis for FNF patients undergoing surgery.⁴^,⁵ In the absence of VTE prophylaxis, the ACCP estimated that fatal pulmonary embolism will occur in 7.5 % of patients with hip fracture.⁶ In addition, FNF is also associated with increased risk of future falls and bleeding,⁷ making the need to carefully balance anticoagulation measures with bleeding precautions paramount.

Precise pharmacological recommendations based on patient-specific risk factors can be key to addressing this clinical challenge. The ACCP recommends 10–14 days of low-molecular-weight-heparin for patients undergoing surgical treatment of FNF.⁵ Alternatively, some studies propose aspirin as an effective prophylactic option for patients undergoing total hip arthroplasty (THA) for FNF,⁸^,⁹ but this recommendation has not been corroborated by prospective studies or major clinical guidelines.¹⁰ Mechanical prophylaxis offers adjunct to chemoprophylaxis and is recommended in the form of fitted compression hosiery, intermittent pneumatic compression devices, and early mobilization.¹¹ Furthermore, safe discharge disposition and patient education may additionally contribute to multimodal VTE prevention.

It is common for FNF patients to be discharged to a rehabilitation facility following surgery, despite the higher incidence of complications reported in such facilities compared to discharge to home.¹²^,¹³ Alternatively in the hip fracture population, discharge to inpatient rehabilitation facility was previously associated with increased early mobilization and decreased short term mortality.¹⁴ While early mobilization is protective for VTE, it may contribute to increased fall risk. Still, VTE is a major source of morbidity and mortality in the acute and chronic rehabilitation populations, where rates of VTE have been reported to be as high as 10 %.¹⁵ While some studies attribute this risk to comorbidities including older age, prior history of VTE, and diabetes diagnosis in the inpatient rehabilitation population,¹⁶ others question whether discharge home may be safer. It was hypothesized that discharge to an inpatient rehabilitation facility after operative fixation of FNF may be associated with increased the risk of postoperative VTE and bleeding.

2 Methods

2.1

2.1 Data source

This was a retrospective observational cohort study using the Merative MarketScan Commercial Claims Database, which captures insurance and procedure related data on over 273 million unique patients annually, including insured employees, retirees, and their dependents covered by employer-sponsored private health insurance in the United States.¹⁷ The database enables longitudinal tracking of patients across multiple institutions and healthcare providers. Additionally, it incorporates medical and pharmacy claims from employers and health plans through the Commercial Claims and Encounters Database.

2.2

2.2 Study population

Following approval by the New England Baptist Hospital Institutional Review Board, the Merative MarketScan Commercial Claims Databases were queried from January 1, 2019 to December 31, 2021. Patients who underwent a THA, hemiarthroplasty, or closed reduction and percutaneous pinning (CRPP) after sustaining a femoral neck fracture using International Classification of Diseases, Tenth Revision (ICD-10) codes (Supplemental Table 1) were included. Patients were excluded if they sustained a periprosthetic or pathologic fracture (Supplemental Table 1). Those with more than one postoperative National Drug Code (NDC) for a thromboprophylactic medication, those who were on chronic anticoagulant medication, those with no postoperative thromboprophylactic medication NDC code, and those with no continuous insurance enrollment three months prior to and following arthroplasty were excluded from the study. Chronic anticoagulant medication was defined as having a prescription for an anticoagulant within 90 days prior to surgery. For each patient basic demographic and surgical data including age, sex, surgery type, surgery date, and length of stay (LOS) were collected.

2.3

2.3 Comorbidities

ICD-10 codes were used to identify the following patient comorbidities: obesity, chronic kidney disease (CKD), antiphospholipid antibody syndrome, lupus anticoagulant, history of VTE, factor V leiden, antithrombin gene mutation, prothrombin gene mutation, protein C or S deficiency, and cancer (Supplemental Table 2).¹⁸ A hereditary hypercoagulable diagnosis was defined as having factor V leiden, antithrombin gene mutation, prothrombin gene mutation, protein C deficiency, or protein S deficiency.

2.4

2.4 Pharmacological thromboprophylaxis

Data on thromboprophylactic medications using NDCs for prescription aspirin, rivaroxaban, apixaban, enoxaparin, and warfarin (Supplemental Table 3) were collected. The pharmacological thromboprophylaxis was assigned by taking the first prescription for a thromboprophylactic medication up to 45 days following the index surgery. All medication prescriptions occurring after a VTE diagnosis were removed.

2.5

2.5 Outcomes

The primary outcome of thirty-day cumulative incidence of postoperative VTE was identified with ICD-10 and Current Procedural Terminology (CPT) codes as previously described (Supplemental Table 2).¹⁹^,²⁰ Secondary outcomes of VTE at ninety-days and bleeding at thirty and ninety-days were also defined with previously validated ICD-10 codes.²¹^,²² Bleeding was analyzed in two groups: post-operative surgical bleeding events (ie. hematomas and events directly following procedure) and all post-operative bleeding events (Supplemental Table 2).²³

2.6

2.6 Sub-analysis of thromboprophylaxis

The preoperative use of thromboprophylactic medication in certain patients was identified as a potentially vulnerable population warranting study. Patients with a chronic thromboprophylactic medication prescription were identified using ICD-10 codes (Supplemental Table 3). For each patient, basic demographic and surgical data including age, sex, surgery type, surgery date, and length of stay (LOS) were collected. Medical comorbidities as described above were additionally tracked. The collected medical comorbidities were grouped together and labeled as “hypercoagulable state” for analysis due to lack of power and numbers, as the collected comorbidities including chronic kidney disease are known predictors of hypercoagulable state. Both the primary outcome of thirty-day cumulative incidence of VTE and the secondary outcome of postoperative bleeding were assessed in this group.

2.7

2.7 Statistical analysis

Cumulative incidences of VTE and bleeding were calculated using SAS (version 9.4). A univariate chi-squared and regression analyses to identify risk factors for 30-day VTE events was performed. Multivariable logistic regression was performed to identify independent risk factors for the 30-day VTE or bleeding events, adjusting for age, sex, number of days between admission and surgery, LOS, obesity, cancer, CKD, history of VTE, hereditary hypercoagulable state, discharge disposition, chronic anticoagulation, and VTE prophylaxis. Data is presented as odds ratios (OR) with 95 % confidence intervals (CI). Finally, the VTE and bleeding cumulative incidences over time were compared.

3 Results

3.1

3.1 Patient overview

Of the patients who underwent THA/TKA between 2019 and 2021, a final cohort of 6731 patients with complete data was analyzed. The median age was 74.7 ± 15 years and 67.0 % of the patients were female. Approximately 98.3 % of patients underwent surgery 0–1 days following admission. 50.3 % of patients had a diagnosis of polytrauma, and 16.4 % had a history of FNF. 0.3 % of patients had a history of previous VTE. The majority were discharged to an inpatient rehabilitation facility (56.6 %) and the mean LOS was 4.5 ± 2.8 days. Mean Charlson Comorbidity Index (CCI) was 5.2 ± 3.3 and common comorbidities included obesity (1.5 %), hereditary hypercoagulable state (0.9 %), CKD (22.5 %), and cancer (10.4 %) (Table 1).

Table 1 Univariate analysis of femoral neck fracture patient and surgical variables.

Variables	TotalNo. (%)	No VTENo. (%)	VTENo. (%)	p-value
Total	6844	6582 (96.2)	262 (3.8)
Age, Mean (SD)	74.7 (15.0)	74.6 (15.1)	77 (13.5)	0.005
Gender
Male	2259 (33.0)	2166 (95.9)	93 (4.1)	0.382
Female	4585 (67.0)	4416 (96.3)	169 (3.7)
Time from admission to Surgery
Surgery 0–1 day after admission	6724 (98.3)	6471 (96.2)	253 (3.8)	0.034
Surgery ≥2 days after admission	120 (1.8)	111 (92.5)	9 (7.5)
Length of Stay, Mean (SD)	4.5 (2.8)	4.5 (2.7)	5.3 (3)	<0.001
Year of Surgery
2019	1591 (23.3)	1529 (96.1)	62 (3.9)	0.881
2020	2495 (36.5)	2405 (96.4)	90 (3.6)	0.471
2021	2758 (40.3)	2648 (96.0)	110 (4.0)	ref
Polytrauma	3442 (50.3)	3304 (96.0)	138 (4.0)	0.432
History of Femoral Neck Fracture	1122 (16.4)	1074 (95.7)	48 (4.3)	0.390
History of VTE	22 (0.3)	18 (81.8)	4 (18.2)	<0.001
Obesity	100 (1.5)	94 (94.0)	6 (6.0)	0.254
Cancer	710 (10.4)	678 (95.5)	32 (4.5)	0.319
Chronic Kidney Disease	1538 (22.5)	1468 (95.5)	70 (4.6)	0.093
Hereditary Hypercoagulable State	64 (0.9)	56 (87.5)	8 (12.5)	<0.001
Antiphospholipid Antibody or Lupus Anticoagulant	103 (1.5)	100 (97.1)	3 (2.9)	0.626
Discharge Location
Home	382 (5.6)	362 (94.8)	20 (5.2)	ref
Other	2591 (37.9)	2532 (97.7)	59 (2.3)	0.001
Rehab	3871 (56.6)	3688 (95.3)	183 (4.7)	<0.001
CCI, Mean (SD)	5.2 (3.3)	5.2 (3.3)	5.7 (3.3)	0.022
Chronic Anticoagulation	968 (14.1)	915 (94.5)	53 (5.5)	0.004
VTE prophylaxis	998 (14.6)	962 (96.4)	36 (3.6)	0.694

3.2

3.2 Risk factors for VTE

There were 262 VTE events at 30 days and 390 at 90 days. The cumulative incidence was 3.81 % (95 % CI = 3.35–4.27) for VTE at 30 days postoperatively (Fig. 1). Univariate analysis demonstrated that older age (p = 0.005), surgery ≥2 days after admission (p = 0.034), longer LOS (p < 0.001), history of VTE (p < 0.001), hereditary hypercoagulable state (p < 0.001), discharge to a location other than home (p < 0.001), higher CCI (p = 0.022), and chronic anticoagulation (p = 0.004) were associated with increased risk of postoperative VTE. On multivariate analysis, factors associated with significant higher odds of 30-day VTE included longer length of stay (OR = 1.6, 95 % CI = 1.02–1.11), hereditary hypercoagulable state (OR = 3.09, 95 % CI = 1.36–7.02), discharge to other vs home (OR = 2.17, 95 % CI = 1.27–1.70), and discharge to inpatient rehabilitation facility vs home (OR = 1.84, 95 % CI = 1.30–2.61).

Cumulative incidence graph of venous thromboembolism and bleed after discharge for femoral neck fracture patients. Red shows the bleeding events. Blue shows the venous thromboembolisms.

3.3

3.3 Risk factors for bleeding

There were 378 bleeding (Table 2) events at 30 days and 656 at 90 days. The cumulative incidence was 5.42 % (CI = 4.87–5.97) for bleeding at 30 days postoperatively (Fig. 1). On multivariate analysis, factors associated with significant higher odds of 30-day bleeding events included male sex (OR = 1.25, CI = 1.01–1.56), longer length of stay (OR = 1.07, CI = 1.04–1.11), CKD (OR = 1.28, CI = 1.01–1.62), discharge other vs home (OR = 2.68, 95 % CI = 1.69–4.26), discharge to inpatient rehabilitation facility vs home (OR = 2.26, 95 % CI = 1.66–3.09), and chronic anticoagulation (OR = 1.58, 95 % CI = 1.21–2.05).

Table 2 Multivariable Analysis of 30-day venous thromboembolism and 30-day bleeds after discharge for femoral neck fracture patients.

Variable	VTE		Bleeding
Variable	OR (95 % CI)	P-value	OR (95 % CI)	P-value
Age	1.00 (0.99, 1.01)	0.884	1.01 (1.00, 1.02)	0.241
Female vs. Male	0.90 (0.69, 1.17)	0.440	0.80 (0.64, 0.99)	0.043
Days between admission and surgery 0–1 vs ≥ 2	1.42 (0.67, 2.98)	0.358	1.25 (0.65, 2.38)	0.506
Length of Stay	1.06 (1.02, 1.11)	0.003	1.07 (1.04, 1.11)	<0.001
Obesity	1.46 (0.62, 3.45)	0.392	0.25 (0.06, 1.05)	0.058
Cancer	1.05 (0.71, 1.54)	0.826	1.00 (0.72, 1.39)	1.000
Chronic Kidney Disease	1.01 (0.75, 1.36)	0.946	1.28 (1.01, 1.62)	0.041
History of VTE	3.09 (0.92, 10.34)	0.067	1.40 (0.39, 5.08)	0.606
APLA or Lupus Anticoagulant	0.54 (0.17, 1.76)	0.310	1.21 (0.57, 2.56)	0.614
Hereditary Hypercoagulable state	3.09 (1.36, 7.02)	0.007	1.55 (0.63, 3.80)	0.341
Discharged to Other vs. Home	2.17 (1.27, 3.70)	0.005	2.68 (1.69, 4.26)	<0.001
Discharged to Rehab vs Home	1.84 (1.30, 2.61)	0.001	2.26 (1.66, 3.09)	<0.001
Chronic Anticoagulation	1.26 (0.90, 1.75)	0.177	1.58 (1.21, 2.05)	0.001
VTE prophylaxis	1.06 (0.72, 1.54)	0.774	1.01 (0.73, 1.40)	0.952

3.4

3.4 Sub-analysis of thromboprophylactic medication use

There were 31 VTE events in the patients who were previously prescribed thromboprophylaxis, resulting in a cumulative incidence of 3.46 % (CI = 2.24–4.68) (Fig. 2). There were 34 events of bleeding at thirty-days in this group, with a cumulative incidence of 4.47 % (CI = 3.09–5.85) (Fig. 2). On univariate analysis, incidence of VTE was higher in patients undergoing surgery ≥2 days after admission (p = 0.003), in patients with history of VTE (p = 0.006), for patients with chronic kidney disease (p = 0.028), and for patients who were prescribed other thromboprophylaxis compared to patients prescribed with aspirin (p = 0.046) (Table 3). On multivariate analysis controlling for age, sex, surgical variables, and discharge disposition in patients prescribed thromboprophylactic medication ≥2 days between admission and surgery was a predictor of VTE (OR = 4.95, 95 % CI = 1.26–19.44). Predictors of bleeding events at thirty days were female sex (OR = 2.27, 95 % CI = 1.16–4.35), discharge to inpatient rehabilitation facility (OR = 6.46, 95 % CI = 1.98–21.00), and discharge to other vs home (OR = 3.63, 95 % CI = 1.60–8.25) (Table 4).

Cumulative incidence graph of venous thromboembolisms and bleeds after discharge for patients prescribed anticoagulants. Red shows the bleeding events. Blue shows the venous thromboembolisms.

Table 3 Univariate analysis of patient and surgical variables for femoral neck fracture patients with anticoagulant prescriptions.

Variables	Total	No VTENo. (%)	VTENo. (%)	P-Value
Total	895	862 (96.31)	33 (3.69)
Age, Mean (SD)	69.4 (15.8)	69.2 (15.9)	73.1 (13.5)	0.128
Male	295 (33.0)	284 (96.3)	11 (3.7)	0.963
Female	600 (67.0)	578 (96.3)	22 (3.7)
Surgery 0–1 day after admission	877 (98.0)	847 (96.6)	30 (3.4)	0.003
Surgery ≥2 days after admission	18 (2.0)	15 (83.3)	3 (16.7)
Length of Stay, Mean (SD)	4 (2.5)	4 (2.5)	4.4 (2.6)	0.356
Year of Surgery
2019	175 (19.6)	169 (96.6)	6 (3.4)	0.667
2020	314 (35.1)	304 (96.8)	10 (3.2)	0.484
2021	406 (45.4)	389 (95.8)	17 (4.2)	ref
Polytrauma	412 (46.0)	400 (97.1)	12 (2.9)	0.256
No Polytrauma		462 (95.7)	21 (4.4)
History of Femoral Neck Fracture	104 (11.6)	98 (94.2)	6 (5.8)	0.231
No History of Femoral Neck Fracture		764 (96.6)	27 (3.4)
History of Venous Thromboembolism	3 (0.3)	2 (66.7)	1 (33.3)	0.006
No History of Venous Thromboembolism		860 (96.4)	32 (3.6)
Obesity	14 (1.6)	13 (92.9)	1 (7.1)	0.489
No Obesity		849 (96.4)	32 (3.6)
Cancer	83 (9.3)	82 (98.8)	1 (1.2)	0.208
No Cancer		780 (96.1)	32 (3.9)
Chronic Kidney Disease	147 (16.4)	137 (93.2)	10 (6.8)	0.028
No Chronic Kidney Disease		725 (96.9)	23 (3.1)
Hereditary Hypercoagulable State	11 (1.2)	10 (90.9)	1 (9.1)	0.339
No hereditary Hypercoagulable State		852 (96.4)	32 (3.6)
Antiphospholipid Antibody or Lupus Anticoagulant	9 (1.0)	9 (100.0)	0 (0.0)	0.555
No Antiphospholipid Antibody or Lupus Anticoagulant		853 (96.3)	33 (3.7)
Discharge Location
Home	49 (5.5)	46 (93.9)	3 (6.1)	ref
Other	582 (65.0)	566 (97.3)	16 (2.8)	0.197
Rehab	264 (29.5)	250 (94.7)	14 (5.3)	0.067
Charleston Comorbidity Index, Mean (SD)	4.4 (3.2)	4.4 (3.2)	5 (3.8)	0.369
Chronic Anticoagulation	198 (22.1)	187 (94.4)	11 (5.6)	0.114
No Chronic Anticoagulation		675 (96.8)	22 (3.2)
Anticoagulant	758 (84.7)	726 (95.8)	32 (4.2)	0.046
Aspirin	137 (15.3)	136 (99.3)	1 (0.7)

Table 4 Multivariable Analysis for 30-day Venous Thromboembolism and Bleeding in femoral neck fracture patients with anticoagulant prescriptions.

Variable	Venous Thromboembolism		Bleed
Variable	Odds Ratio (95 % CI)	P-value	Odds Ratio (95 % CI)	P-value
Age	1.00 (0.98, 1.03)	0.770	1.01 (0.99, 1.04)	0.352
Male vs. Female	0.84 (0.39, 1.82)	0.653	0.44 (0.23, 0.86)	0.017
Days between admission and surgery ≥ 2 vs. 0–1	4.95 (1.26, 19.44)	0.022	2.45 (0.58, 10.30)	0.222
Length of Stay	1.00 (0.86, 1.16)	0.966	1.11 (0.99, 1.24)	0.087
Hypercoagulable state	1.13 (0.49, 2.57)	0.779	0.89 (0.43, 1.83)	0.750
Discharged to Other vs. Home	2.57 (0.68, 9.73)	0.165	6.46 (1.98, 21.00)	0.002
Discharged to Rehab vs Home	1.48 (0.62, 3.51)	0.376	3.63 (1.60, 8.25)	0.002
Anticoagulant vs Aspirin	4.71 (0.61, 36.48)	0.138	4.33 (0.55, 33.98)	0.163

4 Dicussion

Femoral neck fracture is a common consequence of trauma, particularly among the elderly, with VTE being a frequent complication that requires careful surveillance. The absence of proper prophylaxis in hospitalized patients leads to occurrence in 10–80 % cases.²⁴ In the present study, the overall thirty-day cumulative incidence for VTE following FNF was 3.81 % (Fig. 1). Compared to previous studies assessing the epidemiology of VTE following FNF, this estimate is within the wide reported range of 1.6–18.9 %.^25–28 In multivariate analysis, LOS (OR = 1.6, 95 % CI = 1.02–1.11), hereditary hypercoagulability diagnosis (OR = 3.09, 95 % CI = 1.36–7.02), and discharge disposition (OR = 1.84, 95 % CI = 1.30–2.61 to inpatient rehabilitation facility) were independent predictors of increased incidence of VTE. Acquired and hereditary conditions can increase thrombotic risk in patients through a variety of factors including increased blood viscosity, impaired endothelial function, and inflammatory pathways.²⁹ Alternatively, LOS and discharge disposition are two modifiable risk factors that clinicians should consider prioritizing in the prevention of postoperative VTE.

The cumulative incidence of bleeding at 30 days postoperatively was 5.42 % in this study, which was greater than that of VTE. As the cumulative incidence of bleeding was greater than that of VTE, the importance of proper bleeding precautions in postsurgical patients was underscored, especially during early mobilization. Longer length of stay (OR = 1.07, CI = 1.04–1.11) and discharge disposition (OR = 2.26, 95 % CI = 1.66–3.09 to inpatient rehabilitation facility) were both predictors of bleeding events on multivariate analysis, similar to the factors implicated in VTE. Additionally, male sex (OR = 1.25, CI = 1.01–1.56), CKD diagnosis (OR = 1.28, CI = 1.01–1.62), and chronic anticoagulation (OR = 1.58, 95 % CI = 1.21–2.05) were associated with increased risk of bleeding. Again, LOS and discharge disposition are highlighted as potential targets for hospital teams to pursue when reducing postoperative complications.

Patients who were discharged to home were found to have lower risk of both VTE and bleeding when controlling for patient specific factors including age, gender, multiple comorbidities, and prophylactic measures. Approximately 5.6 % of FNF patients in this study were discharged home. This rate is relatively low compared to that of Bentler et al. who reported that approximately 14 % of hip fracture patients were discharged home with 58 % discharged to inpatient nursing facilities.³⁰ There are many factors that contribute to the preference of inpatient rehabilitation facilities in the FNF population, namely the difficulty in achieving pre-fracture functional and social state in the elderly. One study remarked that when assessing benchmarks including balance, muscle strength, and pace; rehabilitation conducted at specialized facilities was superior.³¹ Alternatively, it has been reported that the quality of postsurgical rehabilitation in frail patients relies on a personalized schedule, given that quality of sessions should be valued over quantity in this population.³² With the availability of at home nursing care and physical therapy programs, patients can receive personalized care without relying on an outpatient facility. Such programs can ensure that proper rehabilitation resources are available despite the at home setting.

Safe discharge disposition requires careful consideration of a variety of factors. As temporary lower limb immobilization inherently increases risk of VTE, the postoperative period for patients following surgical treatment of FNF is particularly vulnerable.³³ To mitigate this increased risk, patients must be aware of the importance of early mobilization and chemoprophylaxis. Primarily, patient education is vital before safe discharge home. Many studies have investigated the relationship between preoperative education and surgical outcomes, finding that more well-informed patients tend to have better surgical outcomes. Effective discharge planning involves educational programs that are concise, patient-centered, and multimodal.³⁴ Additional support through online portals and applications have also risen in efficacy in recent years, as patients can refer back to physician specific and reliable information.

Given that the use of thromboprophylactic medications is associated with a 2x risk of bleeding,³⁸ a sub-analysis of patients who were prescribed thromboprophylaxis preoperatively was conducted. Patients in this group were found to have increased risk of bleeding when discharged to an inpatient rehabilitation facility (OR = 6.46, 95 % CI = 1.98–21.00) (Table 4). The increased odds of a bleeding event seen in the inpatient discharge group underscores how home discharge may be particularly valuable in FNF patients taking thromboprophylaxis. Surgeons may consider the home discharge over discharge to an outside facility in FNF patients given the >6x increase in bleeding risk. Furthermore, FNF patients taking thromboprophylaxis who waited ≥2 days between admission and surgery has increased risk of VTE (OR = 4.95, 95 % CI = 1.26–19.44). These increased odds are likely related to the immobilization associated with hip fracture and emphasizes how early treatment can prevent complications. The importance of prompt surgical repair of FNF also relates to the importance of minimizing LOS for improved overall outcomes.

The average LOS was 4.5 days for FNF patients included in this study. This is well below the benchmark of 14 days, which has been associated in previous reports with increased 30-day mortality.³⁵ As length of stay was found to be predictive of both increased risk of VTE and increased risk of bleeding, surgeons are incentivized to keep LOS to a minimum. The caveat to emphasizing early discharge is the challenge of meeting patient needs following departure from the hospital, as they lose the around-the-clock supervision available through in-house physical therapy, on-call physicians, and nursing. In parallel to minimizing LOS, surgical teams must also ensure that patients are able to individually manage their care and achieve daily activities once discharged. In FNF patients, who tend to be elderly, this may be a barrier to home discharge. In previous studies of VTE following hip fracture, the majority of patients sustained VTE after discharge from the hospital.³⁶ Furthermore, ACCP guidelines for chemoprophylaxis regimens recommend prolonging regimens 35 days postoperatively, which extends well beyond average day of discharge.³⁷ Therefore, complications are occurring at higher rates once the patient has left continual monitoring by the surgical team. As such, patients and their caretakers must understand their increased risk of complications and be prepared to take proper precautions to mitigate their susceptibility, while also understanding the signs and symptoms of possible VTE to obtain proper care.

The cost of inpatient rehabilitation may also be a factor that motivates safe home discharge following FNF. Some parallels in analysis of cost can be seen between FNF and elective total joint arthroplasty (TJA). In TJA, bundled payments have been widely implemented and have resulted in cost savings through a variety of initiatives. A major method has been the reduction of discharges to post-acute care facilities, and, as a result, more patients are being discharged home. While this model may be enticing to policy makers for cost savings in the post hip fracture population, the heterogeneity of the post-fracture patient population makes the bundled payment less effective. The FNF patients are at increased risk of complications compared to elective TJA and have an overall higher length of stay, morbidity, and mortality. This means that while the surgical treatment is often similar, the rehabilitation needs are different. Therefore, a comprehensive assessment of patient capabilities is necessary to ensure that discharge home is appropriate. Even so, the lower rates of complications and lower cost associated with discharge home are compelling motivators to increase leverage of safe home discharge following FNF.

There are several strengths to this study's approach, including the use of a national commercial claims database. This database allows for longitudinal tracking of a diverse set of patients across multiple facilities and providers. A limitation of this study is that it is retrospective in nature. A prospectively designed randomized study studying discharge disposition would provide more definitive evidence on patient outcomes. Furthermore, the use of this database relied on the proper documentation and coding of patient specific factors and results. Additionally, the presence of a thromboprophylaxis medication prescription may not be a true indicator of proper patient adherence and may over estimate the number of patients taking these medications as prescribed. While errors in documentation are likely to be randomly distributed throughout the dataset, there is a possibility that improper documentation could skew the observed results.

5 Conclusion

In patients with FNF, discharge to inpatient rehabilitation was associated with increased risk of both postoperative VTE and bleeding. This risk was underscored in patients prescribed thromboprophylaxis medications. LOS ≥5 days was also associated with increased VTE risk, demonstrating the importance of timely hospital discharge. Surgeons should carefully consider whether patients truly require inpatient rehabilitation or can be safely managed at home with services.

CRediT authorship contribution statement

Hannah I. Travers: Conceptualization, Methodology, Formal analysis, Investigation, Validation, Resources, Writing – original draft, Writing – review & editing, Visualization, Project administration. Gloria S. Coden: Conceptualization, Methodology, Software, Validation, Formal analysis, Investigation, Resources, Data curation, Writing – original draft, Writing – review & editing, Visualization, Project administration. Samantha Simon: Conceptualization, Methodology, Software, Validation, Formal analysis, Investigation, Resources, Data curation, Writing – original draft, Writing – review & editing, Visualization. Mikhail Kuznetsov: Writing – review & editing. Eric L. Smith: Conceptualization, Methodology, Data curation, Validation, Supervision, Writing – review & editing. Rushad Patell: Conceptualization, Methodology, Data curation, Validation, Supervision, Writing – review & editing. Brian L. Hollenbeck: Conceptualization, Methodology, Data curation, Validation, Supervision, Writing – review & editing, All authors must have made substantial contributions.

Ethical approval and patient consent

This submission has been reviewed and approved by the NEBH IRB, Assurance # FWA 00009165. During the review, the IRB specifically considered (i) the risks and anticipated benefits, if any, to subjects; (ii) the selection of subjects; (iii) the procedures for securing and documenting informed consent; (iv) the safety of subjects; and (v) the privacy of subjects and confidentiality of the data. (2023 (2021-26 MS)).

Ethics committee approval

IRB approval has been obtained for this study.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial or not-for-profit sectors.

References

Florschutz A.V., Langford J.R., Haidukewych G.J., Koval K.J., . Femoral neck fractures: current management. J Orthop Trauma. 2015;29(3):121-129.
[Google Scholar]
Anderson F.A., Spencer F.A., . Risk factors for venous thromboembolism. Circulation. 2003;107(23_suppl l_1)
[Google Scholar]
Haake D.A., Berkman S.A., . Venous thromboembolic disease after hip surgery. Risk factors, prophylaxis, and diagnosis. Clin Orthop. 1989;242:212-231.
[Google Scholar]
Brox W.T., Roberts K.C., Taksali S., et al . The American Academy of orthopaedic surgeons evidence-based guideline on management of hip fractures in the elderly. J Bone Joint Surg Am. 2015;97(14):1196-1199.
[Google Scholar]
Falck-Ytter Y., Francis C.W., Johanson N.A., et al . Prevention of VTE in orthopedic surgery patients. Chest. 2012;141(2):e278S-e325S.
[Google Scholar]
Geerts W.H., Pineo G.F., Heit J.A., et al . Prevention of venous thromboembolism. Chest. 2004;126(3):338S-400S.
[Google Scholar]
Iolascon G., de Sire A., Calafiore D., et al . Multifactorial assessment of risk of falling in 753 post-menopausal women: a multicenter cross-sectional study by the Italian group for the study of metabolic bone diseases. Clin Interv Aging. 2020;15:1077-1084.
[Google Scholar]
Chisari E., Tan T.L., Shah R., et al . Aspirin is an effective prophylaxis for venous thromboembolism in ambulatory patients with femoral neck fracture undergoing hip arthroplasty. J Bone Jt Surg. 2022;104(7):603-609.
[Google Scholar]
Habibi A.A., Brash A., Rozell J.C., Ganta A., Schwarzkopf R., Arshi A., . Aspirin prophylaxis is not associated with increased risk of venous thromboembolism in arthroplasty for femoral neck fractures: a non-inferiority study. Eur J Orthop Surg Traumatol. 2024;34(3):1405-1411.
[Google Scholar]
Recommendations from the ICM-VTE: trauma. J Bone Jt Surg. 2022;104(Suppl 1):280-308.
[Google Scholar]
Flevas D.A., Megaloikonomos P.D., Dimopoulos L., Mitsiokapa E., Koulouvaris P., Mavrogenis A.F., . Thromboembolism prophylaxis in orthopaedics: an update. EFORT Open Rev. 2018;3(4):136-148.
[Google Scholar]
Malik A.T., Jain N., Frantz T.L., et al . Discharge to inpatient care facilities following hip fracture surgery: incidence, risk factors, and 30-day post-discharge outcomes. Hip Int J Clin Exp Res Hip Pathol Ther. 2022;32(1):131-139.
[Google Scholar]
Benz T., Angst F., Oesch P., et al . Comparison of patients in three different rehabilitation settings after knee or hip arthroplasty: a natural observational, prospective study. BMC Muscoskelet Disord. 2015;16(1):317.
[Google Scholar]
Pfeufer D., Kammerlander C., Stadler C., et al . Multidisciplinary inpatient rehabilitation improves the long-term functional status of geriatric hip-fracture patients. Eur J Med Res. 2020;25(1):31.
[Google Scholar]
Kelly B.M., Yoder B.M., Tang C., Wakefield T.W., . Venous thromboembolic events in the rehabilitation setting. PM&R. 2010;2(7):647-663.
[Google Scholar]
Sikora-Klak J., Zarling B., Bergum C., Flynn J.C., Markel D.C., . The effect of comorbidities on discharge disposition and readmission for total joint arthroplasty patients. J Arthroplast. 2017;32(5):1414-1417.
[Google Scholar]
Merative marketscan research databases.
[Google Scholar]
Friedman R.J., Sengupta N., Lees M., . Economic impact of venous thromboembolism after hip and knee arthroplasty: potential impact of rivaroxaban. Expert Rev Pharmacoecon Outcomes Res. 2011;11(3):299-306.
[Google Scholar]
Lanes S., Fraeman K., Meyers A., Ives J.W., Huang H.‐Y., . Incidence rates for thromboembolic, bleeding and hepatic outcomes in patients undergoing hip or knee replacement surgery. J Thromb Haemostasis. 2011;9(2):325-332.
[Google Scholar]
Agaba P., Kildow B.J., Dhotar H., Seyler T.M., Bolognesi M., . Comparison of postoperative complications after total hip arthroplasty among patients receiving aspirin, enoxaparin, warfarin, and factor Xa inhibitors. J Orthop. 2017;14(4):537-543.
[Google Scholar]
Lawrence K., Joos C., Jones A.E., Johnson S.A., Witt D.M., . Assessing the accuracy of ICD-10 codes for identifying acute thromboembolic events among patients receiving anticoagulation therapy. J Thromb Thrombolysis. 2019;48(2):181-186.
[Google Scholar]
Lutsey P.L., Zakai N.A., MacLehose R.F., et al . Risk of hospitalised bleeding in comparisons of oral anticoagulant options for the primary treatment of venous thromboembolism. Br J Haematol. 2019;185(5):903-911.
[Google Scholar]
Shah S., Norby F.L., Datta Y.H., et al . Comparative effectiveness of direct oral anticoagulants and warfarin in patients with cancer and atrial fibrillation. Blood Adv. 2018;2(3):200-209.
[Google Scholar]
Nkoke C., Tchinde Ngueping M.J., Atemkeng F., Teuwafeu D., Boombhi J., Menanga A., . Incidence of venous thromboembolism, risk factors and prophylaxis in hospitalized patients in the south west region of Cameroon. Vasc Health Risk Manag. 2020;16:317-324.
[Google Scholar]
Hitos K., Fletcher J., . Venous thromboembolism and fractured neck of femur. Thromb Haemost. 2005;94(11):991-996.
[Google Scholar]
Faroug R., Konnuru S., Min S.S., Hussain F., Ampat G., . Venous thromboembolism prevention post neck of femur fractures – does it make a difference? Thromb J. 2008;6(1):8.
[Google Scholar]
Rosencher N., Vielpeau C., Emmerich J., Fagnani F., Samama C.M., . Venous thromboembolism and mortality after hip fracture surgery: the ESCORTE study. J Thromb Haemostasis. 2005;3(9):2006-2014.
[Google Scholar]
Xia Z.N., Xiao K., Zhu W., et al . Risk assessment and management of preoperative venous thromboembolism following femoral neck fracture. J Orthop Surg. 2018;13(1):291.
[Google Scholar]
Xia Z., Chen W., Wang Q., . Risk factors for venous thromboembolism following surgical treatment of fractures: a systematic review and meta‐analysis. Int Wound J. 2023;20(4):995-1007.
[Google Scholar]
Bentler S.E., Liu L., Obrizan M., et al . The aftermath of hip fracture: discharge placement, functional status change, and mortality. Am J Epidemiol. 2009;170(10):1290-1299.
[Google Scholar]
Auais M.A., Eilayyan O., Mayo N.E., . Extended exercise rehabilitation after hip fracture improves patients' physical function: a systematic review and meta-analysis. Phys Ther. 2012;92(11):1437-1451.
[Google Scholar]
Contro D., Elli S., Si Castaldi, et al . Continuity of care for patients with hip fracture after discharge from rehabilitation facility. Acta Bio-Medica Atenei Parm. 2019;90(3):385-393.
[Google Scholar]
Horner D., Goodacre S., Pandor A., et al . Thromboprophylaxis in lower limb immobilisation after injury (TiLLI) Emerg Med J. 2020;37(1):36-41.
[Google Scholar]
Edwards P.K., Mears S.C., Lowry Barnes C., . Preoperative education for hip and knee replacement: never stop learning. Curr Rev Musculoskelet Med. 2017;10(3):356-364.
[Google Scholar]
Nikkel L.E., Kates S.L., Schreck M., Maceroli M., Mahmood B., Elfar J.C., . Length of hospital stay after hip fracture and risk of early mortality after discharge in New York state: retrospective cohort study. Br Med J. 2015;10:h6246.
[Google Scholar]
Dwyer E.P., Moed B.R., . Venous thromboembolism after hospital discharge in pelvic and acetabular fracture patients treated operatively. J Orthop Surg. 2019;27(1)
[Google Scholar]
Geerts W.H., Bergqvist D., Pineo G.F., et al . Prevention of venous thromboembolism. Chest. 2008;133(6):381S-453S.
[Google Scholar]
Shoeb M., Fang M.C., . Assessing bleeding risk in patients taking anticoagulants. J Thromb Thrombolysis. 2013;35(3):312-319.
[Google Scholar]

Show Sections

[1] Florschutz A.V., Langford J.R., Haidukewych G.J., Koval K.J., . Femoral neck fractures: current management. J Orthop Trauma. 2015;29(3):121-129.
[Google Scholar]

[2] Anderson F.A., Spencer F.A., . Risk factors for venous thromboembolism. Circulation. 2003;107(23_suppl l_1)
[Google Scholar]

[3] Haake D.A., Berkman S.A., . Venous thromboembolic disease after hip surgery. Risk factors, prophylaxis, and diagnosis. Clin Orthop. 1989;242:212-231.
[Google Scholar]

[4] Brox W.T., Roberts K.C., Taksali S., et al . The American Academy of orthopaedic surgeons evidence-based guideline on management of hip fractures in the elderly. J Bone Joint Surg Am. 2015;97(14):1196-1199.
[Google Scholar]

[5] Falck-Ytter Y., Francis C.W., Johanson N.A., et al . Prevention of VTE in orthopedic surgery patients. Chest. 2012;141(2):e278S-e325S.
[Google Scholar]

[6] Geerts W.H., Pineo G.F., Heit J.A., et al . Prevention of venous thromboembolism. Chest. 2004;126(3):338S-400S.
[Google Scholar]

[7] Iolascon G., de Sire A., Calafiore D., et al . Multifactorial assessment of risk of falling in 753 post-menopausal women: a multicenter cross-sectional study by the Italian group for the study of metabolic bone diseases. Clin Interv Aging. 2020;15:1077-1084.
[Google Scholar]

[8] Chisari E., Tan T.L., Shah R., et al . Aspirin is an effective prophylaxis for venous thromboembolism in ambulatory patients with femoral neck fracture undergoing hip arthroplasty. J Bone Jt Surg. 2022;104(7):603-609.
[Google Scholar]

[9] Habibi A.A., Brash A., Rozell J.C., Ganta A., Schwarzkopf R., Arshi A., . Aspirin prophylaxis is not associated with increased risk of venous thromboembolism in arthroplasty for femoral neck fractures: a non-inferiority study. Eur J Orthop Surg Traumatol. 2024;34(3):1405-1411.
[Google Scholar]

[10] Recommendations from the ICM-VTE: trauma. J Bone Jt Surg. 2022;104(Suppl 1):280-308.
[Google Scholar]

[11] Flevas D.A., Megaloikonomos P.D., Dimopoulos L., Mitsiokapa E., Koulouvaris P., Mavrogenis A.F., . Thromboembolism prophylaxis in orthopaedics: an update. EFORT Open Rev. 2018;3(4):136-148.
[Google Scholar]

[12] Malik A.T., Jain N., Frantz T.L., et al . Discharge to inpatient care facilities following hip fracture surgery: incidence, risk factors, and 30-day post-discharge outcomes. Hip Int J Clin Exp Res Hip Pathol Ther. 2022;32(1):131-139.
[Google Scholar]

[13] Benz T., Angst F., Oesch P., et al . Comparison of patients in three different rehabilitation settings after knee or hip arthroplasty: a natural observational, prospective study. BMC Muscoskelet Disord. 2015;16(1):317.
[Google Scholar]

[14] Pfeufer D., Kammerlander C., Stadler C., et al . Multidisciplinary inpatient rehabilitation improves the long-term functional status of geriatric hip-fracture patients. Eur J Med Res. 2020;25(1):31.
[Google Scholar]

[15] Kelly B.M., Yoder B.M., Tang C., Wakefield T.W., . Venous thromboembolic events in the rehabilitation setting. PM&R. 2010;2(7):647-663.
[Google Scholar]

[16] Sikora-Klak J., Zarling B., Bergum C., Flynn J.C., Markel D.C., . The effect of comorbidities on discharge disposition and readmission for total joint arthroplasty patients. J Arthroplast. 2017;32(5):1414-1417.
[Google Scholar]

[17] Merative marketscan research databases.
[Google Scholar]

[18] Friedman R.J., Sengupta N., Lees M., . Economic impact of venous thromboembolism after hip and knee arthroplasty: potential impact of rivaroxaban. Expert Rev Pharmacoecon Outcomes Res. 2011;11(3):299-306.
[Google Scholar]

[19] Lanes S., Fraeman K., Meyers A., Ives J.W., Huang H.‐Y., . Incidence rates for thromboembolic, bleeding and hepatic outcomes in patients undergoing hip or knee replacement surgery. J Thromb Haemostasis. 2011;9(2):325-332.
[Google Scholar]

[20] Agaba P., Kildow B.J., Dhotar H., Seyler T.M., Bolognesi M., . Comparison of postoperative complications after total hip arthroplasty among patients receiving aspirin, enoxaparin, warfarin, and factor Xa inhibitors. J Orthop. 2017;14(4):537-543.
[Google Scholar]

[21] Lawrence K., Joos C., Jones A.E., Johnson S.A., Witt D.M., . Assessing the accuracy of ICD-10 codes for identifying acute thromboembolic events among patients receiving anticoagulation therapy. J Thromb Thrombolysis. 2019;48(2):181-186.
[Google Scholar]

[22] Lutsey P.L., Zakai N.A., MacLehose R.F., et al . Risk of hospitalised bleeding in comparisons of oral anticoagulant options for the primary treatment of venous thromboembolism. Br J Haematol. 2019;185(5):903-911.
[Google Scholar]

[23] Shah S., Norby F.L., Datta Y.H., et al . Comparative effectiveness of direct oral anticoagulants and warfarin in patients with cancer and atrial fibrillation. Blood Adv. 2018;2(3):200-209.
[Google Scholar]

[24] Nkoke C., Tchinde Ngueping M.J., Atemkeng F., Teuwafeu D., Boombhi J., Menanga A., . Incidence of venous thromboembolism, risk factors and prophylaxis in hospitalized patients in the south west region of Cameroon. Vasc Health Risk Manag. 2020;16:317-324.
[Google Scholar]

[25] Hitos K., Fletcher J., . Venous thromboembolism and fractured neck of femur. Thromb Haemost. 2005;94(11):991-996.
[Google Scholar]

[26] Faroug R., Konnuru S., Min S.S., Hussain F., Ampat G., . Venous thromboembolism prevention post neck of femur fractures – does it make a difference? Thromb J. 2008;6(1):8.
[Google Scholar]

[27] Rosencher N., Vielpeau C., Emmerich J., Fagnani F., Samama C.M., . Venous thromboembolism and mortality after hip fracture surgery: the ESCORTE study. J Thromb Haemostasis. 2005;3(9):2006-2014.
[Google Scholar]

[28] Xia Z.N., Xiao K., Zhu W., et al . Risk assessment and management of preoperative venous thromboembolism following femoral neck fracture. J Orthop Surg. 2018;13(1):291.
[Google Scholar]

[29] Xia Z., Chen W., Wang Q., . Risk factors for venous thromboembolism following surgical treatment of fractures: a systematic review and meta‐analysis. Int Wound J. 2023;20(4):995-1007.
[Google Scholar]

[30] Bentler S.E., Liu L., Obrizan M., et al . The aftermath of hip fracture: discharge placement, functional status change, and mortality. Am J Epidemiol. 2009;170(10):1290-1299.
[Google Scholar]

[31] Auais M.A., Eilayyan O., Mayo N.E., . Extended exercise rehabilitation after hip fracture improves patients' physical function: a systematic review and meta-analysis. Phys Ther. 2012;92(11):1437-1451.
[Google Scholar]

[32] Contro D., Elli S., Si Castaldi, et al . Continuity of care for patients with hip fracture after discharge from rehabilitation facility. Acta Bio-Medica Atenei Parm. 2019;90(3):385-393.
[Google Scholar]

[33] Horner D., Goodacre S., Pandor A., et al . Thromboprophylaxis in lower limb immobilisation after injury (TiLLI) Emerg Med J. 2020;37(1):36-41.
[Google Scholar]

[34] Edwards P.K., Mears S.C., Lowry Barnes C., . Preoperative education for hip and knee replacement: never stop learning. Curr Rev Musculoskelet Med. 2017;10(3):356-364.
[Google Scholar]

[35] Nikkel L.E., Kates S.L., Schreck M., Maceroli M., Mahmood B., Elfar J.C., . Length of hospital stay after hip fracture and risk of early mortality after discharge in New York state: retrospective cohort study. Br Med J. 2015;10:h6246.
[Google Scholar]

[36] Dwyer E.P., Moed B.R., . Venous thromboembolism after hospital discharge in pelvic and acetabular fracture patients treated operatively. J Orthop Surg. 2019;27(1)
[Google Scholar]

[37] Geerts W.H., Bergqvist D., Pineo G.F., et al . Prevention of venous thromboembolism. Chest. 2008;133(6):381S-453S.
[Google Scholar]

[38] Shoeb M., Fang M.C., . Assessing bleeding risk in patients taking anticoagulants. J Thromb Thrombolysis. 2013;35(3):312-319.
[Google Scholar]

Admission to rehab increases risk of postoperative venous thromboembolism and bleeding after operative fixation of femoral neck fractures

Abstract

Abstract

1 Introduction