- Email: [email protected]
Nonelective Primary Total Hip Arthroplasty: The Effect of Discharge Destination on Postdischarge Outcomes
The Journal of Arthroplasty xxx (2017) 1e7
Contents lists available at ScienceDirect
The Journal of Arthroplasty journal homepage: www.arthroplastyjournal.org
Original Article
Nonelective Primary Total Hip Arthroplasty: The Effect of Discharge Destination on Postdischarge Outcomes Chirag K. Shah, BA a, *, Aakash Keswani, BA a, Debbie Chi, BS a, Alex Sher, BS a, Karl M. Koenig, MD b, Calin S. Moucha, MD a a b
Department of Orthopaedic Surgery, Icahn School of Medicine at Mount Sinai, New York, New York Department of Surgery and Perioperative Care, Dell Medical School, The University of Texas at Austin, Austin, Texas
a r t i c l e i n f o
a b s t r a c t
Article history: Received 19 September 2016 Received in revised form 13 March 2017 Accepted 17 March 2017 Available online xxx
Background: Medicare has enacted a mandatory bundled payment program for primary total joint arthroplasty that includes nonelective primary total hip arthroplasty (THA). Efficient postacute care management has been identified as an opportunity to improve value for patients. We aimed to identify risk factors for and compare rates of complications by discharge destination and then use those factors to risk-stratify non-elective THA patients. Methods: Patients who underwent nonelective primary THA from 2011 to 2014 were identified in the American College of Surgeons National Surgical Quality Improvement Program database and categorized into those discharged to skilled nursing facility or inpatient rehabilitation facility vs home self-managed/ home health (HHH). Bivariate and multivariate analyses of risk factors for postdischarge adverse events were performed using patient characteristics and intraoperative variables. Results: In bivariate analysis, skilled nursing facility or inpatient rehabilitation facility patients compared with HHH patients, had lower rates of postdischarge severe adverse events (SAEs; 49% vs 58%; P < .001) and unplanned 30-day readmissions (71% vs 83%; P < .001). HHH discharged patients with 1 or more of risk factors had a 1.85-6.18 times odds of complications within the first 14 days. Conclusion: The most important risk factors for predicting postdischarge SAE and readmission are predischarge SAE, dependent functional status, body mass index >40 kg/m2, smoking, diabetes, chronic steroid use, and American Society of Anesthesiologists class 3/4. Nonelective THA patients without these risk factors may be safely discharged to home after THA. Orthopedic surgeons and their nonelective THA patients must agree on the most appropriate discharge destination through a shared decision-making process that takes into account these significant risk factors and other psychosocial factors. © 2017 Elsevier Inc. All rights reserved.
Keywords: total hip arthroplasty discharge destination discharge disposition inpatient rehabilitation facility skilled nursing facility
Reimbursement for total joint arthroplasty (TJA) is shifting to value-based models [1]. One increasingly popular example of this is bundled payments, under which providers are paid a single fee for managing all treatments during a defined episode of care (typically 3 days before and 90 days after TJA). In this model, the care team determines which postacute treatment modalities (such as physical
One or more of the authors of this paper have disclosed potential or pertinent conflicts of interest, which may include receipt of payment, either direct or indirect, institutional support, or association with an entity in the biomedical field which may be perceived to have potential conflict of interest with this work. For full disclosure statements refer to http://dx.doi.org/10.1016/j.arth.2017.03.042. * Reprint requests: Chirag K. Shah, BA, Department of Orthopaedic Surgery, Icahn School of Medicine at Mount Sinai, 5 East 98th Street, 9th Floor, New York, NY 10029. http://dx.doi.org/10.1016/j.arth.2017.03.042 0883-5403/© 2017 Elsevier Inc. All rights reserved.
therapy and geriatrician appointments) will best reduce risk of 90day readmissions and complications in their patients and bears financial risk for those decisions. Under Medicare’s voluntary Bundled Payments for Care Improvement program, several pilot demonstrations for primary total hip arthroplasty (THA) and total knee arthroplasty have shown success in improving or maintaining clinical outcomes, while reducing resource utilization and costs [2e4]. Although this model has been successful for elective THA patients with osteoarthritis, its potential is less clear for patients undergoing THA because of hip fracture, acute pain, or any other nonelective reason. Studies reveal that as much as 15% of Medicare THA volume comprised urgent THA patients (mainly hip fracture), and these patients have considerably higher rates of perioperative complications and readmissions along with greater postacute care resource needs [5].
2
C.K. Shah et al. / The Journal of Arthroplasty xxx (2017) 1e7
As shown by Bozic et al [4], postacute care generally accounts for ~40% of episode costs in elective TJA patients, and this number can be up to 40%-50% higher in hip fracture patients [6,7]. Despite this, Medicare included hip fracture patients undergoing TJA in Bundled Payments for Care Improvement and has done so in the Comprehensive Care for Joint Replacement (CJR) programda mandatory bundled payment model applying to 67 geographic areas, ~900 hospitals, and ~25% of national TJA patients [8]. Although the CJR program intends to risk-adjust financial performance targets for hip fracture patients undergoing TJA, surgeons and hospitals must understand the value (health outcomes per health care dollar spent) of nonhome (ie, inpatient rehabilitation facility [IRF], skilled nursing facility [SNF]) vs home (including home self-managed and home health [HHH]) discharge in such patients, as well as how best to risk-stratify them before nonelective THA. Using a high-quality, nationally representative database, this study aimed to compare rates of adverse events in nonelective THA patients by nonhome vs home discharge destinations, identify significant risk factors for postdischarge adverse events in this population, and stratify these patients based on those factors. Methods American College of Surgeons National Surgical Quality Improvement Program Database The American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) is a national surgical database that prospectively collects patient data from over 370 participating institutions. All data are validated with strict adherence guidelines including routine audits to ensure high-quality data. Trained clinical reviewers collect data up to 30 days postoperatively using medical records, operative reports, and patient interviews. In addition, the ACS-NSQIP provides patient demographics such as age, gender, race, smoking status, and functional status among others, as well as patient medical comorbidities including, diabetes, cardiac, pulmonary, renal, cancer, and American Society of Anesthesiologists (ASA) class. Perioperative and intraoperative variables including days from admission to operation, operative time, type of anesthesia, days from operation to discharge, and discharge destination are included as well. Outcomes of Interest Adverse events within 30 days of operation are tracked by the ACS-NSQIP and were classified into the following categories for analysis: severe adverse events (SAEs), minor adverse events, and unplanned readmission [9]. SAEs included death, myocardial infarction, cerebrovascular accident, renal failure, pulmonary embolism, venous thromboembolism, sepsis, septic shock, unplanned intubation, peripheral nerve injury, deep wound infection, organ/space infection, and return to operating room. Minor adverse events included superficial wound infection, urinary tract infection, and pneumonia. SAEs were considered predischarge if they happened on or before the day of nonelective THA, and postdischarge if they happened after the day of surgery. The outcomes of interest for this analysis were postdischarge SAE and unplanned readmission, which by definition are all postdischarge complications.
primary THA from 2011 to 2014. The ACS-NSQIP defines nonelective cases as those with patients who are inpatient at an acute care hospital, transferred from an emergency department (ED), undergo an emergent/urgent surgical case, or are admitted to the hospital on the day(s) before a scheduled procedure for any reason. Primary THA patients were identified using corresponding Current Procedural Terminology code 27130. Patients with incomplete data were removed from the analysis. Based on the discharge destination field, all nonelective THA patients were categorized into nonhome (IRF or SNF) vs home (HHH) discharged cohorts. Patients who passed away before discharge or had “other” discharge destination were removed from the analysis. Although the ACS-NSQIP data collection goes back to 2007, discharge destination data are only available starting from 2011; therefore, only 2011-2014 data were analyzed. Statistical Analysis Statistical analysis was conducted using SAS (version 9.3) software with a 2-tailed alpha of 0.05. Bivariate analysis was conducted to compare demographics, comorbidities, intraoperative variable, predischarge outcomes, and 30-day outcomes between nonhome and home discharge destination THA cohorts. Categorical analysis was conducted with the chi-square and the Fisher exact test where appropriate. Continuous variables were analyzed using the Student t test or the Mann-Whitney U test after testing for normality and equal variance. Multivariate logistic regression models only included predictors which yielded a P value of .20 from bivariate analysis. SAEs predischarge predictors were included in the multivariate logistic regression model regardless of the P value from bivariate analysis. All variables were assessed for confounding and interaction where appropriate. Final models were assessed for goodness of fit using the Hosmer-Lemeshow test and by calculating the area under the receiver-operating characteristics curve (c-statistic). Risk Stratification Analysis Patients in both cohorts were risk stratified into those with 0, 1, and 2 risk factors using 5 significant risk factors for postdischarge SAEs and 6 significant risk factors for unplanned readmission. Rates of postdischarge complications within 0-14 and 15 days postsurgery were compared across these 3 groups for nonhome and home nonelective THA patients. Results Comparison of Patient Characteristics and Comorbidities A total of 3120 nonelective primary TJA patients from 2011 to 2014 were included for analysis. The discharge destinations included home, SNF, and IRF. Compared with those discharged HHH, SNF/IRF-bound patients tended to be older, female, nonsmokers, and less likely to have a body mass index (BMI) >40 kg/m2 (all P .03; Table 1). Nonhome patients were more likely to have diabetes, pulmonary disease, cardiac disease, hypertension, bleeding-causing disorders, ASA class 3/4, and greater days from admission to operation (P .001). Comparison of Adverse Events
Inclusion Population and Categorization A retrospective review of the ACS-NSQIP database was conducted to identify all patients who underwent nonelective
Rates of predischarge SAEs were higher in nonhome patients compared with those of home (61% vs 53%; P ¼ .01). In particular, rates of unplanned intubation (10% vs 3.8%), myocardial infarction
C.K. Shah et al. / The Journal of Arthroplasty xxx (2017) 1e7 Table 1 Comparison of Patient Demographics and Procedure Characteristics Among Nonelective THA Patients. Demographics and Characteristics
Home, 1438 (100%)
SNF/IRF, 1682 (100%)
P Value
Age, mean (SD), y Male gender, n (%) Dependent functional status, n (%) BMI >40 kg/m2, n (%) History of smoking, n (%) History of diabetes, n (%) History of pulmonary disease, n (%) History of chronic heart failure, n (%) Hypertension, n (%) History of renal disease, n (%) Disseminated cancer, n (%) Steroids for chronic condition, n (%) Bleeding-causing disorders, n (%) ASA class 3/4, n (%) Procedure etiology, n (%) Fracture Osteonecrosis Other/unknown Operative time, mean (SD), min Hospital LOS, mean (SD), d
64.0 595 59 69 262 177 65 7 725 13 37 73 82 625
(11.9) (41) (4.1) (4.8) (18) (12) (4.5) (0.5) (50) (0.9) (2.6) (5.1) (5.7) (44)
74.5 543 194 55 219 304 160 39 1107 36 49 129 209 1246
(10.9) (32) (12) (3.3) (13) (18) (9.5) (2.3) (66) (2.1) (2.9) (7.7) (12) (74)
<.001 <.001 <.001 .03 <.001 <.001 <.001 <.001 <.001 .01 .59 .004 <.001 <.001 <.001
684 69 685 105 4.8
(48) (4.8) (48) (54) (5.1)
1166 52 464 102 6.8
(69) (3.1) (28) (54) (11)
.18 <.001
ASA class, American Society of Anesthesiologists Classification System; BMI, body mass index; IRF, inpatient rehabilitation facility; LOS, length of stay; SD, standard deviation; SNF, skilled nursing facility; THA, total hip arthroplasty.
(8.6% vs 5.7%), cardiac arrest requiring cardiopulmonary arrest (5.0% vs 0%), and wound dehiscence (2.2% vs 0%) were drivers for the observed difference (Table 2). Average total length of stay (LOS; 6.8 vs 4.8 days) was higher for nonhome patients (P < .05). Similar analysis revealed lower rates of postdischarge SAEs (49% vs 58%; P < .001) and unplanned 30-day readmissions (71% vs 83%; P < .001) in nonhome vs home patients. This difference was driven by higher
Table 2 Breakdown of Predischarge Severe Adverse Events in Nonelective THA Patients by Discharge Destination. Severe Adverse Events
Home, n (%)
SNF/IRF, n (%)
P Value
Any adverse eventa Severe adverse eventsb Deep wound infection Organ/space infection Wound dehiscence Unplanned intubation Thrombolic event (DVT/PE) Ventilator >48 h Renal insufficiency Renal failure Stroke/CVA Cardiac arrest requiring CPR Myocardial infarction Sepsis Septic shock Return to operating room Minor adverse eventsc Surgical site infection Pneumonia Urinary tract infection
53 28 1 1 0 2 7 2 3 1 1 0 3 5 1 5 30 3 10 17
139 85 4 1 3 14 15 8 2 1 4 7 12 13 3 21 73 6 25 49
<.001 <.001 d d d d d d d d d d d d d d <.001 d d d
(100) (53) (1.9) (1.9) (0) (3.8) (13) (3.8) (5.7) (1.9) (1.9) (0) (5.7) (9.4) (1.9) (15) (57) (5.7) (19) (32)
(100) (61) (2.9) (0.7) (2.2) (10) (11) (5.8) (1.4) (0.7) (2.9) (5.0) (8.6) (9.4) (2.2) (15) (53) (4.3) (18) (35)
CPR, cardiopulmonary arrest; CVA, cerebrovascular accident; DVT, deep vein thrombosis; IRF, inpatient rehabilitation facility; PE, pulmonary embolism; SNF, skilled nursing facility; THA, total hip arthroplasty. a Indicates unique number of patients with any predischarge adverse event (severe or minor). b Indicates unique number of patients with any predischarge severe adverse event. c Indicates unique number of patients with any predischarge minor adverse event.
3
rates of unplanned return to operating room (32% vs 23%), thrombotic events (13% vs 7.7%) and sepsis (10% vs 4.3%) in home patients (P .02 for all; Table 3).
Predictors of Postdischarge Complications In multivariate analysis, SNF/IRF patients were no more likely to suffer an SAE postdischarge or unplanned 30-day readmission than those discharged to HHH. In the same analysis, SAE predischarge, functional status, BMI >40 kg/m2, smoking, and disseminated cancer were identified as predictors for postdischarge SAE (all odds ratio 1.67; P .05). Functional status, BMI >40 kg/m2, and disseminated cancer were also risk factors for unplanned readmission along with diabetes, steroid use for chronic condition, and ASA class (all odds ratio 1.37; P .05) (Table 4). Table 3 Breakdown of Postdischarge Severe Adverse Events in Nonelective THA Patients by Discharge Destination. Severe Adverse Events
Home, n (%)
SNF/IRF, n (%)
P Value
Any adverse event or unplanned readmissiona Any adverse event Severe adverse eventsb Deep wound infection Organ/space infection Wound dehiscence Unplanned intubation Thrombolic event (DVT/PE) Ventilator >48 h Renal insufficiency Renal failure Stroke/CVA Myocardial infarction Sepsis Septic shock Return to operating room Minor adverse eventsc Surgical site infection Pneumonia Urinary tract infection Unplanned readmissiond Medical Pneumonia Thrombolic event (DVT/PE) MI, CHF, or other cardiovascular complications Urinary tract infection Renal insufficiency or failure Ulcer/GI complication Sepsis/septic shock Other/unknown medical reason Surgical Acute pain Dislocation Fracture Postoperative infection or wound complication Hematoma Other/unknown surgical reason
96 (100)
209 (100)
<.001
63 56 9 5 3 1 12 2 2 0 1 0 10 1 31 15 8 1 6 80 36 1 3 3
(66) (58) (9.4) (5.2) (3.1) (1.0) (13) (2.1) (2.1) (0) (1.0) (0) (10) (1.0) (32) (16) (8.3) (1.0) (6.3) (83) (38) (1.0) (3.1) (3.1)
147 103 10 4 3 1 16 0 0 2 3 5 9 4 49 66 19 9 43 149 67 3 1 3
(70) (49) (4.8) (1.9) (1.4) (0.5) (7.7) (0) (0) (1.0) (1.4) (2.4) (4.3) (1.9) (23) (32) (9.1) (4.3) (21) (71) (32) (1.4) (0.5) (1.4)
<.001 .01 d d d d d d d d d d d d d <.001 d d d <.001 d d d d
0 1 2 3 23 44 1 6 4 14
(0) (1.0) (2.1) (3.1) (24) (46) (1.0) (6.3) (4.2) (15)
2 2 6 6 44 82 3 15 5 21
(1.0) (1.0) (2.9) (2.9) (21) (29) (1.4) (7.2) (2.4) (10)
d d d d d d d d d d
1 (1.0) 18 (19)
1 (0.5) 37 (17.8)
d d
CHF, congestive heart failure; CVA, cerebrovascular accident; DVT, deep vein thrombosis; GI, gastrointestinal; IRF, inpatient rehabilitation facility; MI, myocardial infarction; PE, pulmonary embolism; SNF, skilled nursing facility; THA, total hip arthroplasty. a Indicates unique number of patients with any postdischarge adverse event (severe or minor) or unplanned readmission. b Indicates unique number of patients with any postdischarge severe adverse event. c Indicates unique number of patients with any postdischarge minor adverse event. d Indicates unique number of patients with any unplanned readmission.
4
C.K. Shah et al. / The Journal of Arthroplasty xxx (2017) 1e7
Table 4 Risk Factors for Postdischarge Severe Adverse Events and Unplanned Readmission Among Nonelective THA Patients. Risk Factors
SNF/IRF discharge Severe adverse events predischarge Dependent functional status BMI >40 kg/m2 Smoking Diabetes Disseminated cancer Steroids for chronic conditions ASA class 3/4
Postdischarge Severe Adverse Events
Unplanned Readmission
Odds Ratioa (95% CI)
P Value
Odds Ratioa (95% CI)
P Value
1.40 1.85 2.25 3.33 1.67 1.01 2.55 1.22 1.25
.08 .05b <.001b <.001b .02b .96 .01b .51 .28
1.29 0.86 1.66 2.11 1.37 1.37 1.93 1.86 1.71
.11 .67 .01b .01b .10 .05b .04b .01b .01b
(0.96-2.05) (1.01-3.51) (1.44-3.53) (1.83-6.06) (1.09-2.56) (0.65-1.57) (1.27-5.14) (0.68-2.17) (0.83-1.88)
(0.94-1.78) (0.43-1.72) (1.10-2.50) (1.19-3.72) (0.94-1.99) (0.98-1.93) (1.01-3.68) (1.21-2.89) (1.19-2.46)
ASA class, American Society of Anesthesiologists classification system; BMI, body mass index; CI, confidence interval; IRF, inpatient rehabilitation facility; SNF, skilled nursing facility; THA, total hip arthroplasty. a Relative to osteoarthritis etiology. b Indicates significance at P .05.
Risk-Stratified Analysis of Postdischarge Complications For the risk stratification analysis, we used 6 risk factors for unplanned 30-day readmission and 5 risk factors for SAEs postdischarge. We first stratified HHH and nonhome THA patients into 3 categories (0, 1, and 2 risk factors), and then compared rates of adverse events within 0-14 days and 15 days post-THA (Table 5). All significant risk factors for unplanned 30-day readmission and SAE postdischarge were treated with equal weighting for the risk stratification analysis because the odds ratios were all comparable in magnitude (Table 4). Consistent with the multivariate analysis, across all risk levels, there was no difference in the rate of adverse events (0-14 and 15 days postsurgery) between the home and nonhome cohorts. HHH patients with 1 risk factors had a 1.85-6.18 times odds of complications (SAE and readmission) within 14 days postdischarge compared with those with 0 risk factors (P .05 for all). We found no difference in risk beyond 14 days. Among nonhome discharge patients, those with 1 risk factor had a 2.24-4.67 times odds of complications within 14 days postdischarge compared with those with 0 risk factors (P .02 for all). Discussion Care redesign approaches for elective THA bundles have been developed, tested, and validated in multiple health systems, but the same cannot be said for nonelective THA (ie, hip fracture) patients. The goal of this analysis was to use a robust, nationally representative data set to understand the effect of discharge destination on complication risk in nonelective THA patients and to identify other risk factors for postdischarge complications. Comparison of Characteristics and Outcomes of Elective vs Nonelective THA Patients Multiple studies have shown that nonelective THA (mainly hip fracture) patients are a distinct population among those undergoing THA. These patients are more likely to be older, female, and suffer from chronic comorbidities such as hypertension and cognitive disorders compared with those undergoing elective THA [10]. Such patients are also 11 times more likely to suffer 90-day medical (eg, myocardial infarction) and surgical complications (ie, infection, dislocation, and heterotopic ossification) compared with controls [11,12]. A recent study by Yoon et al [13] has shown that hip fracture patients tend to have longer LOS, increased costs, and more readmissions. They have suggested separating patients in bundle
payments based on etiology and comorbidities. Using a prior ACSNSQIP analysis by Keswani et al [14], we found that nonelective THA patients have higher rates of postdischarge SAEs (5.1% vs 2.0%), unplanned 30-day readmissions (7.3% vs 3.5%), and nonhome discharge (54% vs 30%) when compared with elective THA patients (P < .001). Given that (1) complications can cost anywhere from $22,775$36,038, (2) the average readmission for THA patients costs $36,038, and (3) each initial SNF/IRF stay averages $15,000, the average hip fracture patient costs $5000-$7000 more (~17%-23% of the average THA bundle price) than a typical elective THA patient based on postacute care utilization and complications alone [15e17]. This reality combined with the upfront costs of redesigning care for these patients (conservatively approximately ~3% of the average bundle price) suggests that the risk-adjustment multiplier in CJR for hip fracture patients should be at least 1.20-1.25 times the elective THA bundled price. Risk Factors for Postdischarge Complications and Readmissions Identifying risk factors for postdischarge SAEs and unplanned readmission (including SNF/IRF vs home discharge) is necessary for optimizing risk stratification and postdischarge care planning for hip fracture patients. One study by Basques et al [18] identified age, male gender, BMI >35 kg/m2, ASA class, history of pulmonary disease, hypertension, steroid use, and functional status as risk factors for unplanned readmission, which occurred in 10% of the hip €rstedt et al [10] found fracture patients. Another study by Ha ischemic heart disease and malignancy in these patients to be associated with higher rates of mortality. We additionally identified diabetes, disseminated cancer, SAE predischarge, and smoking as risk factors for unplanned readmission and postdischarge SAEs. In contrast to the findings of Keswani et al [14] analysis of elective THA and total knee arthroplasty patients, SNF/IRF discharge was not a significant risk factor for postdischarge SAE or unplanned readmission in the nonelective THA population. This could be because the therapy and close surveillance provided at SNFs/IRFs prevents unplanned readmissions and common complications in hip fracture patients such as falls and periprosthetic fractures [19]. Other risk factors such as age, bleeding-causing disorders, and hypertension were not significant in our nonelective THA cohort while they were among elective patients, but this difference may have been due to the smaller sample size. Among the factors we identified, several (such as smoking, diabetes, and BMI >40 kg/m2) are modifiable, but not in the case of a nonelective surgery. Still, these factors, along with others
5
mentioned previously, can be used to inform postdischarge planning including providing greater care surveillance for high-risk patients.
Odds ratio relative to patient population with 0 risk significant risk factors for 30-d unplanned readmission or SAEs.
Multivariate analysis (Table 4) identified 6 significant (P < .05) risk factors for unplanned readmission: functional status, BMI >40 kg/m2, COPD, renal disease, steroids for chronic conditions, and ASA class 3/4.
c
Multivariate analysis (Table 4) identified 7 significant (P < .05) risk factors for postdischarge SAEs: predischarge SAE, fracture etiology, dependent functional status, BMI >40 kg/m2, smoking, disseminated cancer, and ASA class 3/4. a
b
Risk-Stratified, Time-Based Analysis of Postdischarge Complications
hip arthroplasty.
Asterisks indicate statistical significance at P <.05.
ASA class, American Society of Anesthesiologists classification system; BMI, body mass index; CI, confidence interval; COPD, chronic obstructive pulmonary disease; IRF, inpatient rehabilitation facility; SAEs, severe adverse events; SNF, skilled nursing facility; THA, total
.99 .38 .10 .99 .61 .71 d .12 .01* d 3.55 (0.81-15.54) 6.03 (1.39-26.15) 2 (0.5) 16 (2.0) 18 (3.4) d .02* <.001* d 2.26 (1.09-4.68) 4.67 (2.29-9.55) 9 (2.6) 45 (5.7) 59 (11) 11 (3.2) 61 (7.7) 77 (14) 349 797 536 d .21 .39 d 2.25 (0.63-8.03) 2.11 (0.46-9.48) 4 (0.6) 6 (1.3) 3 (1.2) d .05* <.001* d 1.85 (1.01-3.43) 3.96 (2.14-7.33) 19 (2.7) 23 (4.8) 25 (9.8)
Odds Ratio (95% CI)c 16-30 d, n (%) P Value Odds Ratio (95% CI)c 0-15 d, n (%) Total Readmissions, n (%)
23 (3.2) 29 (6.1) 28 (11)
Number of Patients (100%)
709 475 254
Odds Ratio (95% CI)c 16-30 d, n (%) P Value Odds Ratio (95% CI)c 0-15 d, n (%) Total Readmissions, n (%) SNF/IRF
Number of Patients (100%)
P Value Home
30-d Unplanned Readmission
Number of Significant Risk Factorsb
0 Risk factors 1 Risk factor 2 Risk factors
P Value, 15 d P Value
P Value, 0-15 d
SNF/IRF vs Home
.14 .40 .73 .06 .54 .77 d .68 .01* d 1.15 (0.52-2.54) 4.24 (1.54-11.64) 20 (1.7) 9 (2.0) 5 (6.9) d .003* .001* d 2.24 (1.34-3.75) 4.25 (1.89-9.59) 33 (2.9) 28 (6.2) 8 (11) d .77 .002* d 1.13 (0.35-3.63) 9.94 (2.99-33.03) 10 (1.0) 4 (1.1) 4 (8.9) d <.001* .01* d 3.29 (1.66-6.53) 6.18 (1.98-19.30) 16 (1.6) 18 (5.0) 4 (8.9) 0 Risk factors 1 Risk factor 2 Risk factors
Odds Ratio (95% CI)c 16-30 d, n (%) P Value Odds Ratio (95% CI)c 0-15 d, n (%)
1156 454 72 26 (2.5) 22 (6.0) 8 (18) 1029 364 45
53 (3.2) 37 (8.2) 13 (18)
P Value, 15 d P Value, 0-15 d Number of Patients (100%)
Total SAEs, n (%) SNF/IRF
Number of Patients (100%)
P Value Total SAEs, n (%) Home
Postdischarge Severe Adverse Events
Number of Significant Risk Factorsa
Comparison of Rates of 30-d Unplanned Readmission and SAEs by Timing of Complication and by Discharge Destination in Nonelective THA Patients.
Table 5
0-15 d, n (%)
Odds Ratio (95% CI)c
P Value
16-30 d, n (%)
Odds Ratio (95% CI)c
P Value
SNF/IRF vs Home
C.K. Shah et al. / The Journal of Arthroplasty xxx (2017) 1e7
Risk stratification analysis revealed the incremental increase in complication risk a nonelective THA patient faces from having 1 significant risk factors for 30-day postdischarge complications; specifically, having 1 factor led to a 2.24-3.29 times risk of postdischarge SAE with 15 days, while having 2 factors increased this risk to 4.25-6.18 times compared with having no risk factors. In terms of readmission risk, patients with 1 or 2 risk factors similarly had an increased risk of 1.85 and 4.67 times, respectively, compared with controls. This information is critical for riskstratifying patients before surgery, and tailoring intra-acute and postacute care planning to their level of risk. Examples include specialist consults presurgery for hip fracture patients at high risk of certain complications (eg, cardiology consult for cardiac complications), coordinating with a case manager or social worker to assess the patient's potential for home recovery based on their caregiver support system, engaging those caregivers to increase patient compliance at home, and fracture-specific pain and rehabilitation protocols [20,21]. For the highest risk hip fracture patients, it is now our considered opinion that a longer hospital LOS may often be the best strategy for reducing complications, readmissions, and overall costs of care [22,23]. In addition, Gozalo et al [24] have shown that SNFs/IRFs that care for a higher volume of hip fracture patients tend to have better outcomes although this may be due to “selective referral” as opposed to a “practice makes perfect” effect [25]. The “selective referral” effect suggests that better SNFs would attract most patients because they have shown to have lower readmission rates. Alternatively, the “practice makes perfect” effect occurs when SNFs with larger patient volumes afford their staff greater experience with hip fracture patientespecific needs, rehabilitation protocols, and complications leading to better outcomes per hip fracture patient over time. For those hip fracture patients who are at a particularly higher risk of complications or cannot be feasibly sent home for recovery, identifying and securing beds in high-volume hip fracture SNFs/ IRFs is yet another strategy to improve postdischarge clinical outcomes. For those discharged to HHH, there are several care coordination strategies that have been shown to be effective in orthopedic populations including e-consultations with orthopedic surgeon (or nurse), telemedicine visits with patients, or secure messaging platforms where patients can send pictures of their wounds to the care team [26e28]. Consistent with the multivariate analysis (Table 4), across risk levels, there was no difference in the risk of postdischarge complications of SAE between HHH discharge and SNF/IRF. This information should be presented to patients and their caregivers during the triage process or postoperatively to set expectations that SNF/ IRF discharge destination does not necessarily lead to better outcomes for hip fracture or nonelective THA patients. Comorbidities such as Alzheimer disease or dementia, availability of caregiver support, and family stability have been shown to be critical in the home setting; as such, the appropriate choice of home vs nonhome discharge will differ depending on each patient's circumstances [29e32]. Implications for Care Redesign Beyond improving risk stratification approaches and ensuring appropriately risk adjustment in alternative payment models, our analysis demonstrates the need for a comprehensive nonelective
6
C.K. Shah et al. / The Journal of Arthroplasty xxx (2017) 1e7
joint arthroplasty care pathway to achieve better patient-centered value for this population. Two telling examples are the studies by Lau et al [21] and Folbert et al [20] in which their institutions each developed a multidisciplinary care pathwaydfrom ED through an outpatient follow-updfor geriatric hip fracture patients. These pathways involved fast-tracking patients through the ED, using multidisciplinary teams (comprising social workers, orthopedic surgeons, anesthesiologists, cardiologists, and geriatricians) throughout the process from preoperative treatment decisions through postacute rehabilitation and an “osteophysiotrauma” outpatient follow-up clinic focusing on fall assessment, osteoporosis assessment/treatment, and functional outcomes improvement [20]. These pathways resulted in fewer readmissions (from 12% to 1%), a lower mortality rate (from 2.7 to 1.25), and 2 fewer inpatient days per patient. The risk stratification approach from our analysis could add to these care pathways by enabling the care team to quickly identify high-risk patients within the pathway and use tailored-care approaches described previously. In addition, many techniques that promote early mobilization and successful home discharge for elective TJA patients such as multimodal pain management and rapid recovery protocols can be applied to nonelective THA patients to improve on the current standard of care. Limitations There are several notable limitations to our study. First, we are limited by the size of our sample, as well as the patient characteristics and clinical outcomes captured by the ACS-NSQIP. In addition, the ACS-NSQIP does not report hospital location/region which limits the ability to factor regional variation in rehabilitation facilities use and quality of SNF/IRF in our analysis. Second, the ACSNSQIP does not differentiate between home vs home health discharge destination as these patients were analyzed as one group. Third, we are limited by the data that the ACS-NSQIP collects. It does not provide information about the patient insurance status or the type of hospital providing treatment (ie, regional vs main campus hospital), which, as shown by London et al [33], can significantly affect the proportion of patients discharged to SNF/IRF. Furthermore, the ACS-NSQIP does not collect cost or patientreported outcomes data, both of which would be needed to assess the true impact of discharge destination on patient-centered value. For the purposes of this analysis, we used nonhome discharge destinations (SNF/IRF) as a proxy for greater resource utilization. The ACS-NSQIP also does not contain psychosocial factors (such as sociodemographic status, mental health status, in-home support, and assessments of fall risk at home) that have been shown to explain variation in postacute care utilization and 90-day readmission rates for several conditions and procedures [29]. Such factors could prevent low-risk nonelective THA patients from being discharged home, and worsen short-term clinical outcomes regardless of discharge destination. The ACS-NSQIP database also does not include several orthopedics-specific variables that would be of interest in managing hip fractures, such as the level of mobility after surgery but before discharge, patient-reported outcomes (pain, function, and quality of life), and long-term outcomes (nonunion and malunion) [34]. Fourth, ACS-NSQIP's definition of nonelective THA is broad and further analysis of different subgroups may add insights. Finally, we are limited to a 30-day postoperative follow-up which is likely much shorter than the average recovery period for a typical hip fracture or nonelective THA patient. At minimum, a 90-day follow-up period would be invaluable for analyses aimed at improving performance under bundled payment models.
Conclusion Among nonelective THA patients, significant risk factors for postdischarge SAE or unplanned readmission include predischarge SAE, dependent functional status, BMI >40 kg/m2, smoking, diabetes, chronic steroid use, and ASA class 3/4. Unlike in elective THA patients, SNF/IRF discharge destination was not a significant risk factor for postdischarge complications. Nonelective THA patients without the aforementioned risk factors may be safely discharged to home after THA. Care delivery pathways can be redesigned with these patients in mind to streamline the process and improve patient outcomes while taking into account other rapid ongoing changes in care. These risk factors can be used to identify high-risk patients and to tailor care based on the risk level, especially within the first 2 weeks postdischarge, when the risks of complications are the highest. They should also be included in the algorithm for determining discharge destination; however, it is important to note that these factors are only one input among several (psychosocial factors, insurance status, home safety, and patient/family preference) that will inform where and when the patient is discharged.
References [1] Porter ME. What is value in health care? N Engl J Med 2010;363:2477e81. [2] Iorio R, Clair AJ, Inneh IA, Slover JD, Bosco JA, Zuckerman JD. Early results of Medicare's Bundled Payment Initiative for a 90-day total joint arthroplasty episode of care. J Arthroplasty 2016;31:343e50. [3] Whitcomb WF, Lagu T, Krushell RJ, Lehman AP, Greenbaum J, McGirr J, et al. Experience with designing and implementing a bundled payment program for total hip replacement. Jt Comm J Qual Patient Saf 2015;41:406e13. [4] Bozic KJ, Ward L, Vail TP, Maze M. Bundled payments in total joint arthroplasty: targeting opportunities for quality improvement and cost reduction. Clin Orthop Relat Res 2014;472:188e93. [5] Burgers PT, Hoogendoorn M, Van Woensel EA, Poolman RW, Bhandari M, Patka P, et al. Total medical costs of treating femoral neck fracture patients with hemi- or total hip arthroplasty: a cost analysis of a multicenter prospective study. Osteoporos Int 2016;27:1999e2008. [6] Kivlahan C, Orlowski JM, Pearce J, Walradt J, Baker M, Kirch DG. Taking risk: early results from teaching hospitals' participation in the center for Medicare and Medicaid Innovation Bundled Payments for Care Improvement Initiative. Acad Med 2016;91:936e42. [7] Buntin MB, Colla CH, Deb P, Sood N, Escarce JJ. Medicare spending and outcomes after postacute care for stroke and hip fracture. Med Care 2010;48:776e84. [8] Centers for Medicare & Medicaid Services HHS. Medicare program; Comprehensive Care for Joint Replacement payment model for acute care hospitals furnishing lower extremity joint replacement services. Final rule. Fed Regist 2015;80:73273e554. [9] Pugely AJ, Callaghan JJ, Martin CT, Cram P, Gao Y. Incidence of and risk factors for 30-day readmission following elective primary total joint arthroplasty: analysis from the ACS-NSQIP. J Arthroplasty 2013;28:1499e504. [10] Harstedt M, Rogmark C, Sutton R, Melander O, Fedorowski A. Impact of comorbidity on 6-month hospital readmission and mortality after hip fracture surgery. Injury 2015;46:713e8. [11] Morison Z, Moojen DJ, Nauth A, Hall J, McKee MD, Waddell JP, et al. Total hip arthroplasty after acetabular fracture is associated with lower survivorship and more complications. Clin Orthop Relat Res 2016;474:392e8. [12] Mayers W, Schwartz B, Schwartz A, Moretti V, Goldstein W, Shah R. National trends and in hospital outcomes for total hip arthroplasty in avascular necrosis in the United States. Int Orthop 2016;40:1787e92. [13] Yoon RS, Mahure SA, Hutzler L, Iorio R, Bosco JA. Hip arthroplasty for fracture vs. elective care: one bundle does not fit all. J Arthroplasty 2017 [Epub ahead of print]. [14] Keswani A, Tasi MC, Fields A, Lovy AJ, Moucha CS, Bozic KJ. Discharge destination after total joint arthroplasty: an analysis of postdischarge outcomes, placement risk factors, and recent trends. J Arthroplasty 2016;31: 1155e62. [15] Dobson A, DaVanzo JE, El-Gamil A, Li JW, Manolov N. Assessment of patient outcomes of rehabilitative care provided in inpatient rehabilitation facilities (IRFs) and after discharge. Springfield, IL: American Rehabilitation Association Research Institute; 2014. [16] Clair AJ, Evangelista PJ, Lajam CM, Slover JD, Bosco JA, Iorio R. Cost analysis of total joint arthroplasty readmissions in a Bundled Payment Care Improvement initiative. J Arthroplasty 2016;31:1862e5. [17] Johnson DJ, Greenberg SE, Sathiyakumar V, Thakore R, Ehrenfeld JM, Obremskey WT, et al. Relationship between the Charlson comorbidity index and cost of treating hip fractures: implications for bundled payment. J Orthop Traumatol 2015;16:209e13.
C.K. Shah et al. / The Journal of Arthroplasty xxx (2017) 1e7 [18] Basques BA, Bohl DD, Golinvaux NS, Leslie MP, Baumgaertner MR, Grauer JN. Postoperative length of stay and 30-day readmission after geriatric hip fracture: an analysis of 8434 patients. J Orthop Trauma 2015;29:e115e20. [19] Jung HY, Trivedi AN, Grabowski DC, Mor V. Does more therapy in skilled nursing facilities lead to better outcomes in patients with hip fracture? Phys Ther 2016;96:81e9. [20] Folbert EC, Smit RS, van der Velde D, Regtuijt EM, Klaren MH, Hegeman JH. Geriatric fracture center: a multidisciplinary treatment approach for older patients with a hip fracture improved quality of clinical care and short-term treatment outcomes. Geriatr Orthop Surg Rehabil 2012;3:59e67. [21] Lau TW, Fang C, Leung F. The effectiveness of a geriatric hip fracture clinical pathway in reducing hospital and rehabilitation length of stay and improving short-term mortality rates. Geriatr Orthop Surg Rehabil 2013;4:3e9. [22] Ricciardi BF, Oi KK, Daines SB, Lee YY, Joseph AD, Westrich GH. Patient and perioperative variables affecting 30-day readmission for surgical complications after hip and knee arthroplasties: a matched cohort study. J Arthroplasty 2017;32:1074e9. [23] Nordstrom P, Michaelsson K, Hommel A, Norrman PO, Thorngren KG, Nordstrom A. Geriatric rehabilitation and discharge location after hip fracture in relation to the risks of death and readmission. J Am Med Directors Assoc 2016;17:91.e1e7. [24] Gozalo P, Leland NE, Christian TJ, Mor V, Teno JM. Volume matters: returning home after hip fracture. J Am Geriatr Soc 2015;63:2043e51. [25] Li Y, Cai X, Yin J, Glance LG, Mukamel DB. Is higher volume of postacute care patients associated with a lower rehospitalization rate in skilled nursing facilities? Med Care Res Rev 2012;69:103e18.
7
[26] Haukipuro K, Ohinmaa A, Winblad I, Linden T, Vuolio S. The feasibility of telemedicine for orthopaedic outpatient clinics: a randomized controlled trial. J Telemed Telecare 2000;6:193e8. [27] Semple JL, Sharpe S, Murnaghan ML, Theodoropoulos J, Metcalfe KA. Using a mobile app for monitoring post-operative quality of recovery of patients at home: a feasibility study. JMIR Mhealth Uhealth 2015;3:e18. [28] Wood G, Naudie D, MacDonald S, McCalden R, Bourne R. An electronic clinic for arthroplasty follow-up: a pilot study. Can J Surg 2011;54:381e6. [29] Hospital discharge planning: a guide for families and caregivers. https:// www.caregiver.org/hospital-discharge-planning-guide-families-and-caregivers [accessed 20.07.16]. [30] Discharge planning. https://www.cms.gov/Outreach-and-Education/MedicareLearning-Network-MLN/MLNProducts/Downloads/Discharge-Planning-BookletICN908184.pdf [accessed 20.07.16]. [31] Discharge planning - frequently asked questions. https://www.hss.edu/casemanagement-discharge-planning.asp [accessed 20.07.16]. [32] Bosco J, Shah PC, Slover JD, Torrance A. Optimizing the OR for bundled payments: a case study. Bull Am Coll Surg 2014;99:29e36. [33] London DA, Vilensky S, O'Rourke C, Schill M, Woicehovich L, Froimson MI. Discharge disposition after joint replacement and the potential for cost savings: effect of hospital policies and surgeons. J Arthroplasty 2016;31: 743e8. [34] Molina CS, Thakore RV, Blumer A, Obremskey WT, Sethi MK. Use of the National Surgical Quality Improvement Program in orthopaedic surgery. Clin Orthop Relat Res 2015;473:1574e81.
Contents lists available at ScienceDirect
The Journal of Arthroplasty journal homepage: www.arthroplastyjournal.org
Original Article
Nonelective Primary Total Hip Arthroplasty: The Effect of Discharge Destination on Postdischarge Outcomes Chirag K. Shah, BA a, *, Aakash Keswani, BA a, Debbie Chi, BS a, Alex Sher, BS a, Karl M. Koenig, MD b, Calin S. Moucha, MD a a b
Department of Orthopaedic Surgery, Icahn School of Medicine at Mount Sinai, New York, New York Department of Surgery and Perioperative Care, Dell Medical School, The University of Texas at Austin, Austin, Texas
a r t i c l e i n f o
a b s t r a c t
Article history: Received 19 September 2016 Received in revised form 13 March 2017 Accepted 17 March 2017 Available online xxx
Background: Medicare has enacted a mandatory bundled payment program for primary total joint arthroplasty that includes nonelective primary total hip arthroplasty (THA). Efficient postacute care management has been identified as an opportunity to improve value for patients. We aimed to identify risk factors for and compare rates of complications by discharge destination and then use those factors to risk-stratify non-elective THA patients. Methods: Patients who underwent nonelective primary THA from 2011 to 2014 were identified in the American College of Surgeons National Surgical Quality Improvement Program database and categorized into those discharged to skilled nursing facility or inpatient rehabilitation facility vs home self-managed/ home health (HHH). Bivariate and multivariate analyses of risk factors for postdischarge adverse events were performed using patient characteristics and intraoperative variables. Results: In bivariate analysis, skilled nursing facility or inpatient rehabilitation facility patients compared with HHH patients, had lower rates of postdischarge severe adverse events (SAEs; 49% vs 58%; P < .001) and unplanned 30-day readmissions (71% vs 83%; P < .001). HHH discharged patients with 1 or more of risk factors had a 1.85-6.18 times odds of complications within the first 14 days. Conclusion: The most important risk factors for predicting postdischarge SAE and readmission are predischarge SAE, dependent functional status, body mass index >40 kg/m2, smoking, diabetes, chronic steroid use, and American Society of Anesthesiologists class 3/4. Nonelective THA patients without these risk factors may be safely discharged to home after THA. Orthopedic surgeons and their nonelective THA patients must agree on the most appropriate discharge destination through a shared decision-making process that takes into account these significant risk factors and other psychosocial factors. © 2017 Elsevier Inc. All rights reserved.
Keywords: total hip arthroplasty discharge destination discharge disposition inpatient rehabilitation facility skilled nursing facility
Reimbursement for total joint arthroplasty (TJA) is shifting to value-based models [1]. One increasingly popular example of this is bundled payments, under which providers are paid a single fee for managing all treatments during a defined episode of care (typically 3 days before and 90 days after TJA). In this model, the care team determines which postacute treatment modalities (such as physical
One or more of the authors of this paper have disclosed potential or pertinent conflicts of interest, which may include receipt of payment, either direct or indirect, institutional support, or association with an entity in the biomedical field which may be perceived to have potential conflict of interest with this work. For full disclosure statements refer to http://dx.doi.org/10.1016/j.arth.2017.03.042. * Reprint requests: Chirag K. Shah, BA, Department of Orthopaedic Surgery, Icahn School of Medicine at Mount Sinai, 5 East 98th Street, 9th Floor, New York, NY 10029. http://dx.doi.org/10.1016/j.arth.2017.03.042 0883-5403/© 2017 Elsevier Inc. All rights reserved.
therapy and geriatrician appointments) will best reduce risk of 90day readmissions and complications in their patients and bears financial risk for those decisions. Under Medicare’s voluntary Bundled Payments for Care Improvement program, several pilot demonstrations for primary total hip arthroplasty (THA) and total knee arthroplasty have shown success in improving or maintaining clinical outcomes, while reducing resource utilization and costs [2e4]. Although this model has been successful for elective THA patients with osteoarthritis, its potential is less clear for patients undergoing THA because of hip fracture, acute pain, or any other nonelective reason. Studies reveal that as much as 15% of Medicare THA volume comprised urgent THA patients (mainly hip fracture), and these patients have considerably higher rates of perioperative complications and readmissions along with greater postacute care resource needs [5].
2
C.K. Shah et al. / The Journal of Arthroplasty xxx (2017) 1e7
As shown by Bozic et al [4], postacute care generally accounts for ~40% of episode costs in elective TJA patients, and this number can be up to 40%-50% higher in hip fracture patients [6,7]. Despite this, Medicare included hip fracture patients undergoing TJA in Bundled Payments for Care Improvement and has done so in the Comprehensive Care for Joint Replacement (CJR) programda mandatory bundled payment model applying to 67 geographic areas, ~900 hospitals, and ~25% of national TJA patients [8]. Although the CJR program intends to risk-adjust financial performance targets for hip fracture patients undergoing TJA, surgeons and hospitals must understand the value (health outcomes per health care dollar spent) of nonhome (ie, inpatient rehabilitation facility [IRF], skilled nursing facility [SNF]) vs home (including home self-managed and home health [HHH]) discharge in such patients, as well as how best to risk-stratify them before nonelective THA. Using a high-quality, nationally representative database, this study aimed to compare rates of adverse events in nonelective THA patients by nonhome vs home discharge destinations, identify significant risk factors for postdischarge adverse events in this population, and stratify these patients based on those factors. Methods American College of Surgeons National Surgical Quality Improvement Program Database The American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) is a national surgical database that prospectively collects patient data from over 370 participating institutions. All data are validated with strict adherence guidelines including routine audits to ensure high-quality data. Trained clinical reviewers collect data up to 30 days postoperatively using medical records, operative reports, and patient interviews. In addition, the ACS-NSQIP provides patient demographics such as age, gender, race, smoking status, and functional status among others, as well as patient medical comorbidities including, diabetes, cardiac, pulmonary, renal, cancer, and American Society of Anesthesiologists (ASA) class. Perioperative and intraoperative variables including days from admission to operation, operative time, type of anesthesia, days from operation to discharge, and discharge destination are included as well. Outcomes of Interest Adverse events within 30 days of operation are tracked by the ACS-NSQIP and were classified into the following categories for analysis: severe adverse events (SAEs), minor adverse events, and unplanned readmission [9]. SAEs included death, myocardial infarction, cerebrovascular accident, renal failure, pulmonary embolism, venous thromboembolism, sepsis, septic shock, unplanned intubation, peripheral nerve injury, deep wound infection, organ/space infection, and return to operating room. Minor adverse events included superficial wound infection, urinary tract infection, and pneumonia. SAEs were considered predischarge if they happened on or before the day of nonelective THA, and postdischarge if they happened after the day of surgery. The outcomes of interest for this analysis were postdischarge SAE and unplanned readmission, which by definition are all postdischarge complications.
primary THA from 2011 to 2014. The ACS-NSQIP defines nonelective cases as those with patients who are inpatient at an acute care hospital, transferred from an emergency department (ED), undergo an emergent/urgent surgical case, or are admitted to the hospital on the day(s) before a scheduled procedure for any reason. Primary THA patients were identified using corresponding Current Procedural Terminology code 27130. Patients with incomplete data were removed from the analysis. Based on the discharge destination field, all nonelective THA patients were categorized into nonhome (IRF or SNF) vs home (HHH) discharged cohorts. Patients who passed away before discharge or had “other” discharge destination were removed from the analysis. Although the ACS-NSQIP data collection goes back to 2007, discharge destination data are only available starting from 2011; therefore, only 2011-2014 data were analyzed. Statistical Analysis Statistical analysis was conducted using SAS (version 9.3) software with a 2-tailed alpha of 0.05. Bivariate analysis was conducted to compare demographics, comorbidities, intraoperative variable, predischarge outcomes, and 30-day outcomes between nonhome and home discharge destination THA cohorts. Categorical analysis was conducted with the chi-square and the Fisher exact test where appropriate. Continuous variables were analyzed using the Student t test or the Mann-Whitney U test after testing for normality and equal variance. Multivariate logistic regression models only included predictors which yielded a P value of .20 from bivariate analysis. SAEs predischarge predictors were included in the multivariate logistic regression model regardless of the P value from bivariate analysis. All variables were assessed for confounding and interaction where appropriate. Final models were assessed for goodness of fit using the Hosmer-Lemeshow test and by calculating the area under the receiver-operating characteristics curve (c-statistic). Risk Stratification Analysis Patients in both cohorts were risk stratified into those with 0, 1, and 2 risk factors using 5 significant risk factors for postdischarge SAEs and 6 significant risk factors for unplanned readmission. Rates of postdischarge complications within 0-14 and 15 days postsurgery were compared across these 3 groups for nonhome and home nonelective THA patients. Results Comparison of Patient Characteristics and Comorbidities A total of 3120 nonelective primary TJA patients from 2011 to 2014 were included for analysis. The discharge destinations included home, SNF, and IRF. Compared with those discharged HHH, SNF/IRF-bound patients tended to be older, female, nonsmokers, and less likely to have a body mass index (BMI) >40 kg/m2 (all P .03; Table 1). Nonhome patients were more likely to have diabetes, pulmonary disease, cardiac disease, hypertension, bleeding-causing disorders, ASA class 3/4, and greater days from admission to operation (P .001). Comparison of Adverse Events
Inclusion Population and Categorization A retrospective review of the ACS-NSQIP database was conducted to identify all patients who underwent nonelective
Rates of predischarge SAEs were higher in nonhome patients compared with those of home (61% vs 53%; P ¼ .01). In particular, rates of unplanned intubation (10% vs 3.8%), myocardial infarction
C.K. Shah et al. / The Journal of Arthroplasty xxx (2017) 1e7 Table 1 Comparison of Patient Demographics and Procedure Characteristics Among Nonelective THA Patients. Demographics and Characteristics
Home, 1438 (100%)
SNF/IRF, 1682 (100%)
P Value
Age, mean (SD), y Male gender, n (%) Dependent functional status, n (%) BMI >40 kg/m2, n (%) History of smoking, n (%) History of diabetes, n (%) History of pulmonary disease, n (%) History of chronic heart failure, n (%) Hypertension, n (%) History of renal disease, n (%) Disseminated cancer, n (%) Steroids for chronic condition, n (%) Bleeding-causing disorders, n (%) ASA class 3/4, n (%) Procedure etiology, n (%) Fracture Osteonecrosis Other/unknown Operative time, mean (SD), min Hospital LOS, mean (SD), d
64.0 595 59 69 262 177 65 7 725 13 37 73 82 625
(11.9) (41) (4.1) (4.8) (18) (12) (4.5) (0.5) (50) (0.9) (2.6) (5.1) (5.7) (44)
74.5 543 194 55 219 304 160 39 1107 36 49 129 209 1246
(10.9) (32) (12) (3.3) (13) (18) (9.5) (2.3) (66) (2.1) (2.9) (7.7) (12) (74)
<.001 <.001 <.001 .03 <.001 <.001 <.001 <.001 <.001 .01 .59 .004 <.001 <.001 <.001
684 69 685 105 4.8
(48) (4.8) (48) (54) (5.1)
1166 52 464 102 6.8
(69) (3.1) (28) (54) (11)
.18 <.001
ASA class, American Society of Anesthesiologists Classification System; BMI, body mass index; IRF, inpatient rehabilitation facility; LOS, length of stay; SD, standard deviation; SNF, skilled nursing facility; THA, total hip arthroplasty.
(8.6% vs 5.7%), cardiac arrest requiring cardiopulmonary arrest (5.0% vs 0%), and wound dehiscence (2.2% vs 0%) were drivers for the observed difference (Table 2). Average total length of stay (LOS; 6.8 vs 4.8 days) was higher for nonhome patients (P < .05). Similar analysis revealed lower rates of postdischarge SAEs (49% vs 58%; P < .001) and unplanned 30-day readmissions (71% vs 83%; P < .001) in nonhome vs home patients. This difference was driven by higher
Table 2 Breakdown of Predischarge Severe Adverse Events in Nonelective THA Patients by Discharge Destination. Severe Adverse Events
Home, n (%)
SNF/IRF, n (%)
P Value
Any adverse eventa Severe adverse eventsb Deep wound infection Organ/space infection Wound dehiscence Unplanned intubation Thrombolic event (DVT/PE) Ventilator >48 h Renal insufficiency Renal failure Stroke/CVA Cardiac arrest requiring CPR Myocardial infarction Sepsis Septic shock Return to operating room Minor adverse eventsc Surgical site infection Pneumonia Urinary tract infection
53 28 1 1 0 2 7 2 3 1 1 0 3 5 1 5 30 3 10 17
139 85 4 1 3 14 15 8 2 1 4 7 12 13 3 21 73 6 25 49
<.001 <.001 d d d d d d d d d d d d d d <.001 d d d
(100) (53) (1.9) (1.9) (0) (3.8) (13) (3.8) (5.7) (1.9) (1.9) (0) (5.7) (9.4) (1.9) (15) (57) (5.7) (19) (32)
(100) (61) (2.9) (0.7) (2.2) (10) (11) (5.8) (1.4) (0.7) (2.9) (5.0) (8.6) (9.4) (2.2) (15) (53) (4.3) (18) (35)
CPR, cardiopulmonary arrest; CVA, cerebrovascular accident; DVT, deep vein thrombosis; IRF, inpatient rehabilitation facility; PE, pulmonary embolism; SNF, skilled nursing facility; THA, total hip arthroplasty. a Indicates unique number of patients with any predischarge adverse event (severe or minor). b Indicates unique number of patients with any predischarge severe adverse event. c Indicates unique number of patients with any predischarge minor adverse event.
3
rates of unplanned return to operating room (32% vs 23%), thrombotic events (13% vs 7.7%) and sepsis (10% vs 4.3%) in home patients (P .02 for all; Table 3).
Predictors of Postdischarge Complications In multivariate analysis, SNF/IRF patients were no more likely to suffer an SAE postdischarge or unplanned 30-day readmission than those discharged to HHH. In the same analysis, SAE predischarge, functional status, BMI >40 kg/m2, smoking, and disseminated cancer were identified as predictors for postdischarge SAE (all odds ratio 1.67; P .05). Functional status, BMI >40 kg/m2, and disseminated cancer were also risk factors for unplanned readmission along with diabetes, steroid use for chronic condition, and ASA class (all odds ratio 1.37; P .05) (Table 4). Table 3 Breakdown of Postdischarge Severe Adverse Events in Nonelective THA Patients by Discharge Destination. Severe Adverse Events
Home, n (%)
SNF/IRF, n (%)
P Value
Any adverse event or unplanned readmissiona Any adverse event Severe adverse eventsb Deep wound infection Organ/space infection Wound dehiscence Unplanned intubation Thrombolic event (DVT/PE) Ventilator >48 h Renal insufficiency Renal failure Stroke/CVA Myocardial infarction Sepsis Septic shock Return to operating room Minor adverse eventsc Surgical site infection Pneumonia Urinary tract infection Unplanned readmissiond Medical Pneumonia Thrombolic event (DVT/PE) MI, CHF, or other cardiovascular complications Urinary tract infection Renal insufficiency or failure Ulcer/GI complication Sepsis/septic shock Other/unknown medical reason Surgical Acute pain Dislocation Fracture Postoperative infection or wound complication Hematoma Other/unknown surgical reason
96 (100)
209 (100)
<.001
63 56 9 5 3 1 12 2 2 0 1 0 10 1 31 15 8 1 6 80 36 1 3 3
(66) (58) (9.4) (5.2) (3.1) (1.0) (13) (2.1) (2.1) (0) (1.0) (0) (10) (1.0) (32) (16) (8.3) (1.0) (6.3) (83) (38) (1.0) (3.1) (3.1)
147 103 10 4 3 1 16 0 0 2 3 5 9 4 49 66 19 9 43 149 67 3 1 3
(70) (49) (4.8) (1.9) (1.4) (0.5) (7.7) (0) (0) (1.0) (1.4) (2.4) (4.3) (1.9) (23) (32) (9.1) (4.3) (21) (71) (32) (1.4) (0.5) (1.4)
<.001 .01 d d d d d d d d d d d d d <.001 d d d <.001 d d d d
0 1 2 3 23 44 1 6 4 14
(0) (1.0) (2.1) (3.1) (24) (46) (1.0) (6.3) (4.2) (15)
2 2 6 6 44 82 3 15 5 21
(1.0) (1.0) (2.9) (2.9) (21) (29) (1.4) (7.2) (2.4) (10)
d d d d d d d d d d
1 (1.0) 18 (19)
1 (0.5) 37 (17.8)
d d
CHF, congestive heart failure; CVA, cerebrovascular accident; DVT, deep vein thrombosis; GI, gastrointestinal; IRF, inpatient rehabilitation facility; MI, myocardial infarction; PE, pulmonary embolism; SNF, skilled nursing facility; THA, total hip arthroplasty. a Indicates unique number of patients with any postdischarge adverse event (severe or minor) or unplanned readmission. b Indicates unique number of patients with any postdischarge severe adverse event. c Indicates unique number of patients with any postdischarge minor adverse event. d Indicates unique number of patients with any unplanned readmission.
4
C.K. Shah et al. / The Journal of Arthroplasty xxx (2017) 1e7
Table 4 Risk Factors for Postdischarge Severe Adverse Events and Unplanned Readmission Among Nonelective THA Patients. Risk Factors
SNF/IRF discharge Severe adverse events predischarge Dependent functional status BMI >40 kg/m2 Smoking Diabetes Disseminated cancer Steroids for chronic conditions ASA class 3/4
Postdischarge Severe Adverse Events
Unplanned Readmission
Odds Ratioa (95% CI)
P Value
Odds Ratioa (95% CI)
P Value
1.40 1.85 2.25 3.33 1.67 1.01 2.55 1.22 1.25
.08 .05b <.001b <.001b .02b .96 .01b .51 .28
1.29 0.86 1.66 2.11 1.37 1.37 1.93 1.86 1.71
.11 .67 .01b .01b .10 .05b .04b .01b .01b
(0.96-2.05) (1.01-3.51) (1.44-3.53) (1.83-6.06) (1.09-2.56) (0.65-1.57) (1.27-5.14) (0.68-2.17) (0.83-1.88)
(0.94-1.78) (0.43-1.72) (1.10-2.50) (1.19-3.72) (0.94-1.99) (0.98-1.93) (1.01-3.68) (1.21-2.89) (1.19-2.46)
ASA class, American Society of Anesthesiologists classification system; BMI, body mass index; CI, confidence interval; IRF, inpatient rehabilitation facility; SNF, skilled nursing facility; THA, total hip arthroplasty. a Relative to osteoarthritis etiology. b Indicates significance at P .05.
Risk-Stratified Analysis of Postdischarge Complications For the risk stratification analysis, we used 6 risk factors for unplanned 30-day readmission and 5 risk factors for SAEs postdischarge. We first stratified HHH and nonhome THA patients into 3 categories (0, 1, and 2 risk factors), and then compared rates of adverse events within 0-14 days and 15 days post-THA (Table 5). All significant risk factors for unplanned 30-day readmission and SAE postdischarge were treated with equal weighting for the risk stratification analysis because the odds ratios were all comparable in magnitude (Table 4). Consistent with the multivariate analysis, across all risk levels, there was no difference in the rate of adverse events (0-14 and 15 days postsurgery) between the home and nonhome cohorts. HHH patients with 1 risk factors had a 1.85-6.18 times odds of complications (SAE and readmission) within 14 days postdischarge compared with those with 0 risk factors (P .05 for all). We found no difference in risk beyond 14 days. Among nonhome discharge patients, those with 1 risk factor had a 2.24-4.67 times odds of complications within 14 days postdischarge compared with those with 0 risk factors (P .02 for all). Discussion Care redesign approaches for elective THA bundles have been developed, tested, and validated in multiple health systems, but the same cannot be said for nonelective THA (ie, hip fracture) patients. The goal of this analysis was to use a robust, nationally representative data set to understand the effect of discharge destination on complication risk in nonelective THA patients and to identify other risk factors for postdischarge complications. Comparison of Characteristics and Outcomes of Elective vs Nonelective THA Patients Multiple studies have shown that nonelective THA (mainly hip fracture) patients are a distinct population among those undergoing THA. These patients are more likely to be older, female, and suffer from chronic comorbidities such as hypertension and cognitive disorders compared with those undergoing elective THA [10]. Such patients are also 11 times more likely to suffer 90-day medical (eg, myocardial infarction) and surgical complications (ie, infection, dislocation, and heterotopic ossification) compared with controls [11,12]. A recent study by Yoon et al [13] has shown that hip fracture patients tend to have longer LOS, increased costs, and more readmissions. They have suggested separating patients in bundle
payments based on etiology and comorbidities. Using a prior ACSNSQIP analysis by Keswani et al [14], we found that nonelective THA patients have higher rates of postdischarge SAEs (5.1% vs 2.0%), unplanned 30-day readmissions (7.3% vs 3.5%), and nonhome discharge (54% vs 30%) when compared with elective THA patients (P < .001). Given that (1) complications can cost anywhere from $22,775$36,038, (2) the average readmission for THA patients costs $36,038, and (3) each initial SNF/IRF stay averages $15,000, the average hip fracture patient costs $5000-$7000 more (~17%-23% of the average THA bundle price) than a typical elective THA patient based on postacute care utilization and complications alone [15e17]. This reality combined with the upfront costs of redesigning care for these patients (conservatively approximately ~3% of the average bundle price) suggests that the risk-adjustment multiplier in CJR for hip fracture patients should be at least 1.20-1.25 times the elective THA bundled price. Risk Factors for Postdischarge Complications and Readmissions Identifying risk factors for postdischarge SAEs and unplanned readmission (including SNF/IRF vs home discharge) is necessary for optimizing risk stratification and postdischarge care planning for hip fracture patients. One study by Basques et al [18] identified age, male gender, BMI >35 kg/m2, ASA class, history of pulmonary disease, hypertension, steroid use, and functional status as risk factors for unplanned readmission, which occurred in 10% of the hip €rstedt et al [10] found fracture patients. Another study by Ha ischemic heart disease and malignancy in these patients to be associated with higher rates of mortality. We additionally identified diabetes, disseminated cancer, SAE predischarge, and smoking as risk factors for unplanned readmission and postdischarge SAEs. In contrast to the findings of Keswani et al [14] analysis of elective THA and total knee arthroplasty patients, SNF/IRF discharge was not a significant risk factor for postdischarge SAE or unplanned readmission in the nonelective THA population. This could be because the therapy and close surveillance provided at SNFs/IRFs prevents unplanned readmissions and common complications in hip fracture patients such as falls and periprosthetic fractures [19]. Other risk factors such as age, bleeding-causing disorders, and hypertension were not significant in our nonelective THA cohort while they were among elective patients, but this difference may have been due to the smaller sample size. Among the factors we identified, several (such as smoking, diabetes, and BMI >40 kg/m2) are modifiable, but not in the case of a nonelective surgery. Still, these factors, along with others
5
mentioned previously, can be used to inform postdischarge planning including providing greater care surveillance for high-risk patients.
Odds ratio relative to patient population with 0 risk significant risk factors for 30-d unplanned readmission or SAEs.
Multivariate analysis (Table 4) identified 6 significant (P < .05) risk factors for unplanned readmission: functional status, BMI >40 kg/m2, COPD, renal disease, steroids for chronic conditions, and ASA class 3/4.
c
Multivariate analysis (Table 4) identified 7 significant (P < .05) risk factors for postdischarge SAEs: predischarge SAE, fracture etiology, dependent functional status, BMI >40 kg/m2, smoking, disseminated cancer, and ASA class 3/4. a
b
Risk-Stratified, Time-Based Analysis of Postdischarge Complications
hip arthroplasty.
Asterisks indicate statistical significance at P <.05.
ASA class, American Society of Anesthesiologists classification system; BMI, body mass index; CI, confidence interval; COPD, chronic obstructive pulmonary disease; IRF, inpatient rehabilitation facility; SAEs, severe adverse events; SNF, skilled nursing facility; THA, total
.99 .38 .10 .99 .61 .71 d .12 .01* d 3.55 (0.81-15.54) 6.03 (1.39-26.15) 2 (0.5) 16 (2.0) 18 (3.4) d .02* <.001* d 2.26 (1.09-4.68) 4.67 (2.29-9.55) 9 (2.6) 45 (5.7) 59 (11) 11 (3.2) 61 (7.7) 77 (14) 349 797 536 d .21 .39 d 2.25 (0.63-8.03) 2.11 (0.46-9.48) 4 (0.6) 6 (1.3) 3 (1.2) d .05* <.001* d 1.85 (1.01-3.43) 3.96 (2.14-7.33) 19 (2.7) 23 (4.8) 25 (9.8)
Odds Ratio (95% CI)c 16-30 d, n (%) P Value Odds Ratio (95% CI)c 0-15 d, n (%) Total Readmissions, n (%)
23 (3.2) 29 (6.1) 28 (11)
Number of Patients (100%)
709 475 254
Odds Ratio (95% CI)c 16-30 d, n (%) P Value Odds Ratio (95% CI)c 0-15 d, n (%) Total Readmissions, n (%) SNF/IRF
Number of Patients (100%)
P Value Home
30-d Unplanned Readmission
Number of Significant Risk Factorsb
0 Risk factors 1 Risk factor 2 Risk factors
P Value, 15 d P Value
P Value, 0-15 d
SNF/IRF vs Home
.14 .40 .73 .06 .54 .77 d .68 .01* d 1.15 (0.52-2.54) 4.24 (1.54-11.64) 20 (1.7) 9 (2.0) 5 (6.9) d .003* .001* d 2.24 (1.34-3.75) 4.25 (1.89-9.59) 33 (2.9) 28 (6.2) 8 (11) d .77 .002* d 1.13 (0.35-3.63) 9.94 (2.99-33.03) 10 (1.0) 4 (1.1) 4 (8.9) d <.001* .01* d 3.29 (1.66-6.53) 6.18 (1.98-19.30) 16 (1.6) 18 (5.0) 4 (8.9) 0 Risk factors 1 Risk factor 2 Risk factors
Odds Ratio (95% CI)c 16-30 d, n (%) P Value Odds Ratio (95% CI)c 0-15 d, n (%)
1156 454 72 26 (2.5) 22 (6.0) 8 (18) 1029 364 45
53 (3.2) 37 (8.2) 13 (18)
P Value, 15 d P Value, 0-15 d Number of Patients (100%)
Total SAEs, n (%) SNF/IRF
Number of Patients (100%)
P Value Total SAEs, n (%) Home
Postdischarge Severe Adverse Events
Number of Significant Risk Factorsa
Comparison of Rates of 30-d Unplanned Readmission and SAEs by Timing of Complication and by Discharge Destination in Nonelective THA Patients.
Table 5
0-15 d, n (%)
Odds Ratio (95% CI)c
P Value
16-30 d, n (%)
Odds Ratio (95% CI)c
P Value
SNF/IRF vs Home
C.K. Shah et al. / The Journal of Arthroplasty xxx (2017) 1e7
Risk stratification analysis revealed the incremental increase in complication risk a nonelective THA patient faces from having 1 significant risk factors for 30-day postdischarge complications; specifically, having 1 factor led to a 2.24-3.29 times risk of postdischarge SAE with 15 days, while having 2 factors increased this risk to 4.25-6.18 times compared with having no risk factors. In terms of readmission risk, patients with 1 or 2 risk factors similarly had an increased risk of 1.85 and 4.67 times, respectively, compared with controls. This information is critical for riskstratifying patients before surgery, and tailoring intra-acute and postacute care planning to their level of risk. Examples include specialist consults presurgery for hip fracture patients at high risk of certain complications (eg, cardiology consult for cardiac complications), coordinating with a case manager or social worker to assess the patient's potential for home recovery based on their caregiver support system, engaging those caregivers to increase patient compliance at home, and fracture-specific pain and rehabilitation protocols [20,21]. For the highest risk hip fracture patients, it is now our considered opinion that a longer hospital LOS may often be the best strategy for reducing complications, readmissions, and overall costs of care [22,23]. In addition, Gozalo et al [24] have shown that SNFs/IRFs that care for a higher volume of hip fracture patients tend to have better outcomes although this may be due to “selective referral” as opposed to a “practice makes perfect” effect [25]. The “selective referral” effect suggests that better SNFs would attract most patients because they have shown to have lower readmission rates. Alternatively, the “practice makes perfect” effect occurs when SNFs with larger patient volumes afford their staff greater experience with hip fracture patientespecific needs, rehabilitation protocols, and complications leading to better outcomes per hip fracture patient over time. For those hip fracture patients who are at a particularly higher risk of complications or cannot be feasibly sent home for recovery, identifying and securing beds in high-volume hip fracture SNFs/ IRFs is yet another strategy to improve postdischarge clinical outcomes. For those discharged to HHH, there are several care coordination strategies that have been shown to be effective in orthopedic populations including e-consultations with orthopedic surgeon (or nurse), telemedicine visits with patients, or secure messaging platforms where patients can send pictures of their wounds to the care team [26e28]. Consistent with the multivariate analysis (Table 4), across risk levels, there was no difference in the risk of postdischarge complications of SAE between HHH discharge and SNF/IRF. This information should be presented to patients and their caregivers during the triage process or postoperatively to set expectations that SNF/ IRF discharge destination does not necessarily lead to better outcomes for hip fracture or nonelective THA patients. Comorbidities such as Alzheimer disease or dementia, availability of caregiver support, and family stability have been shown to be critical in the home setting; as such, the appropriate choice of home vs nonhome discharge will differ depending on each patient's circumstances [29e32]. Implications for Care Redesign Beyond improving risk stratification approaches and ensuring appropriately risk adjustment in alternative payment models, our analysis demonstrates the need for a comprehensive nonelective
6
C.K. Shah et al. / The Journal of Arthroplasty xxx (2017) 1e7
joint arthroplasty care pathway to achieve better patient-centered value for this population. Two telling examples are the studies by Lau et al [21] and Folbert et al [20] in which their institutions each developed a multidisciplinary care pathwaydfrom ED through an outpatient follow-updfor geriatric hip fracture patients. These pathways involved fast-tracking patients through the ED, using multidisciplinary teams (comprising social workers, orthopedic surgeons, anesthesiologists, cardiologists, and geriatricians) throughout the process from preoperative treatment decisions through postacute rehabilitation and an “osteophysiotrauma” outpatient follow-up clinic focusing on fall assessment, osteoporosis assessment/treatment, and functional outcomes improvement [20]. These pathways resulted in fewer readmissions (from 12% to 1%), a lower mortality rate (from 2.7 to 1.25), and 2 fewer inpatient days per patient. The risk stratification approach from our analysis could add to these care pathways by enabling the care team to quickly identify high-risk patients within the pathway and use tailored-care approaches described previously. In addition, many techniques that promote early mobilization and successful home discharge for elective TJA patients such as multimodal pain management and rapid recovery protocols can be applied to nonelective THA patients to improve on the current standard of care. Limitations There are several notable limitations to our study. First, we are limited by the size of our sample, as well as the patient characteristics and clinical outcomes captured by the ACS-NSQIP. In addition, the ACS-NSQIP does not report hospital location/region which limits the ability to factor regional variation in rehabilitation facilities use and quality of SNF/IRF in our analysis. Second, the ACSNSQIP does not differentiate between home vs home health discharge destination as these patients were analyzed as one group. Third, we are limited by the data that the ACS-NSQIP collects. It does not provide information about the patient insurance status or the type of hospital providing treatment (ie, regional vs main campus hospital), which, as shown by London et al [33], can significantly affect the proportion of patients discharged to SNF/IRF. Furthermore, the ACS-NSQIP does not collect cost or patientreported outcomes data, both of which would be needed to assess the true impact of discharge destination on patient-centered value. For the purposes of this analysis, we used nonhome discharge destinations (SNF/IRF) as a proxy for greater resource utilization. The ACS-NSQIP also does not contain psychosocial factors (such as sociodemographic status, mental health status, in-home support, and assessments of fall risk at home) that have been shown to explain variation in postacute care utilization and 90-day readmission rates for several conditions and procedures [29]. Such factors could prevent low-risk nonelective THA patients from being discharged home, and worsen short-term clinical outcomes regardless of discharge destination. The ACS-NSQIP database also does not include several orthopedics-specific variables that would be of interest in managing hip fractures, such as the level of mobility after surgery but before discharge, patient-reported outcomes (pain, function, and quality of life), and long-term outcomes (nonunion and malunion) [34]. Fourth, ACS-NSQIP's definition of nonelective THA is broad and further analysis of different subgroups may add insights. Finally, we are limited to a 30-day postoperative follow-up which is likely much shorter than the average recovery period for a typical hip fracture or nonelective THA patient. At minimum, a 90-day follow-up period would be invaluable for analyses aimed at improving performance under bundled payment models.
Conclusion Among nonelective THA patients, significant risk factors for postdischarge SAE or unplanned readmission include predischarge SAE, dependent functional status, BMI >40 kg/m2, smoking, diabetes, chronic steroid use, and ASA class 3/4. Unlike in elective THA patients, SNF/IRF discharge destination was not a significant risk factor for postdischarge complications. Nonelective THA patients without the aforementioned risk factors may be safely discharged to home after THA. Care delivery pathways can be redesigned with these patients in mind to streamline the process and improve patient outcomes while taking into account other rapid ongoing changes in care. These risk factors can be used to identify high-risk patients and to tailor care based on the risk level, especially within the first 2 weeks postdischarge, when the risks of complications are the highest. They should also be included in the algorithm for determining discharge destination; however, it is important to note that these factors are only one input among several (psychosocial factors, insurance status, home safety, and patient/family preference) that will inform where and when the patient is discharged.
References [1] Porter ME. What is value in health care? N Engl J Med 2010;363:2477e81. [2] Iorio R, Clair AJ, Inneh IA, Slover JD, Bosco JA, Zuckerman JD. Early results of Medicare's Bundled Payment Initiative for a 90-day total joint arthroplasty episode of care. J Arthroplasty 2016;31:343e50. [3] Whitcomb WF, Lagu T, Krushell RJ, Lehman AP, Greenbaum J, McGirr J, et al. Experience with designing and implementing a bundled payment program for total hip replacement. Jt Comm J Qual Patient Saf 2015;41:406e13. [4] Bozic KJ, Ward L, Vail TP, Maze M. Bundled payments in total joint arthroplasty: targeting opportunities for quality improvement and cost reduction. Clin Orthop Relat Res 2014;472:188e93. [5] Burgers PT, Hoogendoorn M, Van Woensel EA, Poolman RW, Bhandari M, Patka P, et al. Total medical costs of treating femoral neck fracture patients with hemi- or total hip arthroplasty: a cost analysis of a multicenter prospective study. Osteoporos Int 2016;27:1999e2008. [6] Kivlahan C, Orlowski JM, Pearce J, Walradt J, Baker M, Kirch DG. Taking risk: early results from teaching hospitals' participation in the center for Medicare and Medicaid Innovation Bundled Payments for Care Improvement Initiative. Acad Med 2016;91:936e42. [7] Buntin MB, Colla CH, Deb P, Sood N, Escarce JJ. Medicare spending and outcomes after postacute care for stroke and hip fracture. Med Care 2010;48:776e84. [8] Centers for Medicare & Medicaid Services HHS. Medicare program; Comprehensive Care for Joint Replacement payment model for acute care hospitals furnishing lower extremity joint replacement services. Final rule. Fed Regist 2015;80:73273e554. [9] Pugely AJ, Callaghan JJ, Martin CT, Cram P, Gao Y. Incidence of and risk factors for 30-day readmission following elective primary total joint arthroplasty: analysis from the ACS-NSQIP. J Arthroplasty 2013;28:1499e504. [10] Harstedt M, Rogmark C, Sutton R, Melander O, Fedorowski A. Impact of comorbidity on 6-month hospital readmission and mortality after hip fracture surgery. Injury 2015;46:713e8. [11] Morison Z, Moojen DJ, Nauth A, Hall J, McKee MD, Waddell JP, et al. Total hip arthroplasty after acetabular fracture is associated with lower survivorship and more complications. Clin Orthop Relat Res 2016;474:392e8. [12] Mayers W, Schwartz B, Schwartz A, Moretti V, Goldstein W, Shah R. National trends and in hospital outcomes for total hip arthroplasty in avascular necrosis in the United States. Int Orthop 2016;40:1787e92. [13] Yoon RS, Mahure SA, Hutzler L, Iorio R, Bosco JA. Hip arthroplasty for fracture vs. elective care: one bundle does not fit all. J Arthroplasty 2017 [Epub ahead of print]. [14] Keswani A, Tasi MC, Fields A, Lovy AJ, Moucha CS, Bozic KJ. Discharge destination after total joint arthroplasty: an analysis of postdischarge outcomes, placement risk factors, and recent trends. J Arthroplasty 2016;31: 1155e62. [15] Dobson A, DaVanzo JE, El-Gamil A, Li JW, Manolov N. Assessment of patient outcomes of rehabilitative care provided in inpatient rehabilitation facilities (IRFs) and after discharge. Springfield, IL: American Rehabilitation Association Research Institute; 2014. [16] Clair AJ, Evangelista PJ, Lajam CM, Slover JD, Bosco JA, Iorio R. Cost analysis of total joint arthroplasty readmissions in a Bundled Payment Care Improvement initiative. J Arthroplasty 2016;31:1862e5. [17] Johnson DJ, Greenberg SE, Sathiyakumar V, Thakore R, Ehrenfeld JM, Obremskey WT, et al. Relationship between the Charlson comorbidity index and cost of treating hip fractures: implications for bundled payment. J Orthop Traumatol 2015;16:209e13.
C.K. Shah et al. / The Journal of Arthroplasty xxx (2017) 1e7 [18] Basques BA, Bohl DD, Golinvaux NS, Leslie MP, Baumgaertner MR, Grauer JN. Postoperative length of stay and 30-day readmission after geriatric hip fracture: an analysis of 8434 patients. J Orthop Trauma 2015;29:e115e20. [19] Jung HY, Trivedi AN, Grabowski DC, Mor V. Does more therapy in skilled nursing facilities lead to better outcomes in patients with hip fracture? Phys Ther 2016;96:81e9. [20] Folbert EC, Smit RS, van der Velde D, Regtuijt EM, Klaren MH, Hegeman JH. Geriatric fracture center: a multidisciplinary treatment approach for older patients with a hip fracture improved quality of clinical care and short-term treatment outcomes. Geriatr Orthop Surg Rehabil 2012;3:59e67. [21] Lau TW, Fang C, Leung F. The effectiveness of a geriatric hip fracture clinical pathway in reducing hospital and rehabilitation length of stay and improving short-term mortality rates. Geriatr Orthop Surg Rehabil 2013;4:3e9. [22] Ricciardi BF, Oi KK, Daines SB, Lee YY, Joseph AD, Westrich GH. Patient and perioperative variables affecting 30-day readmission for surgical complications after hip and knee arthroplasties: a matched cohort study. J Arthroplasty 2017;32:1074e9. [23] Nordstrom P, Michaelsson K, Hommel A, Norrman PO, Thorngren KG, Nordstrom A. Geriatric rehabilitation and discharge location after hip fracture in relation to the risks of death and readmission. J Am Med Directors Assoc 2016;17:91.e1e7. [24] Gozalo P, Leland NE, Christian TJ, Mor V, Teno JM. Volume matters: returning home after hip fracture. J Am Geriatr Soc 2015;63:2043e51. [25] Li Y, Cai X, Yin J, Glance LG, Mukamel DB. Is higher volume of postacute care patients associated with a lower rehospitalization rate in skilled nursing facilities? Med Care Res Rev 2012;69:103e18.
7
[26] Haukipuro K, Ohinmaa A, Winblad I, Linden T, Vuolio S. The feasibility of telemedicine for orthopaedic outpatient clinics: a randomized controlled trial. J Telemed Telecare 2000;6:193e8. [27] Semple JL, Sharpe S, Murnaghan ML, Theodoropoulos J, Metcalfe KA. Using a mobile app for monitoring post-operative quality of recovery of patients at home: a feasibility study. JMIR Mhealth Uhealth 2015;3:e18. [28] Wood G, Naudie D, MacDonald S, McCalden R, Bourne R. An electronic clinic for arthroplasty follow-up: a pilot study. Can J Surg 2011;54:381e6. [29] Hospital discharge planning: a guide for families and caregivers. https:// www.caregiver.org/hospital-discharge-planning-guide-families-and-caregivers [accessed 20.07.16]. [30] Discharge planning. https://www.cms.gov/Outreach-and-Education/MedicareLearning-Network-MLN/MLNProducts/Downloads/Discharge-Planning-BookletICN908184.pdf [accessed 20.07.16]. [31] Discharge planning - frequently asked questions. https://www.hss.edu/casemanagement-discharge-planning.asp [accessed 20.07.16]. [32] Bosco J, Shah PC, Slover JD, Torrance A. Optimizing the OR for bundled payments: a case study. Bull Am Coll Surg 2014;99:29e36. [33] London DA, Vilensky S, O'Rourke C, Schill M, Woicehovich L, Froimson MI. Discharge disposition after joint replacement and the potential for cost savings: effect of hospital policies and surgeons. J Arthroplasty 2016;31: 743e8. [34] Molina CS, Thakore RV, Blumer A, Obremskey WT, Sethi MK. Use of the National Surgical Quality Improvement Program in orthopaedic surgery. Clin Orthop Relat Res 2015;473:1574e81.