- Email: [email protected]
Preoperative Predictors of Discharge Destination after Endovascular Repair of Abdominal Aortic Aneurysms
Clinical Research Preoperative Predictors of Discharge Destination after Endovascular Repair of Abdominal Aortic Aneurysms Laura T. Boitano,1 James C. Iannuzzi,1 Adam Tanious,1 Jahan Mohebali,1 Samuel I. Schwartz,1 David C. Chang,2 W. Darrin Clouse,1 and Mark F. Conrad,1 Boston, Massachusetts
Background: There is a paucity of data guiding preoperative counseling on the need for discharge to a facility or nonhome discharge (NHD) following elective endovascular repair of abdominal aortic aneurysms (endovascular aneurysm repair [EVAR]). This study seeks to determine the preoperative predictors of NHD following EVAR in baseline home-dwelling patients and to determine whether NHD is associated with major postdischarge complications and readmission. Methods: This retrospective cohort study utilized the National Surgical Quality Improvement Program Vascular Procedure Targeted database to identify elective EVAR cases admitted from home (2011 to 2015). The primary end point was NHD. A multivariable logistic regression model was used to determine predictive preoperative factors for NHD and to determine whether NHD predicted major postdischarge complications and readmission. Results: Overall 6,276 cases were included; 291 (4.6%) required NHD. NHD were more frequently female, anemic, functionally dependent, nonsmokers, had chronic obstructive pulmonary disease, recent congestive heart failure exacerbation, and open baseline wounds. NHD was associated with complex surgery, indicated by operative time more than the median, 2.5 hr. Significant predictors for NHD on multivariable analysis included female sex (odds ratio [OR]: 2.2, confidence interval [CI]: 1.7e2.9, P < 0.001), octogenarians (OR: 5.7 CI: 2.3e14.1; P < 0.001) and nonagenarians (OR: 14.6, CI: 5.4e39.2; P < 0.001), dependent functional status (OR: 5.4, CI: 3.3e8.8; P < 0.001), preoperative open wound (OR: 3.5, CI: 1.4e8.9; P ¼ 0.006), high operative time (OR: 2.7, CI: 2.0e3.6; P < 0.001), and hypogastric embolization (OR: 1.6, CI: 1.1e2.1 P ¼ 0.022), C-statistic ¼ 0.780. On adjusted analysis, NHD did not independently predict major postdischarge complication (OR: 1.0 CI: 0.6e1.9; P ¼ 0.875) or unplanned readmission (OR 1.0, CI: 0.6e1.5, P ¼ 0.842). Conclusions: Discharge to skilled facility following EVAR can be predicted using preoperative factors. Future studies should seek to validate these findings in a prospective manner. Identifying high-risk patients’ NHD can help define expectations and facilitate early referral to skilled facilities that may reduce hospital length of stay, reducing health-care costs.
Presented at the 42nd Annual Meeting of the Vascular and Endovascular Surgery Society, February 1e4, 2018, Vail, CO.
Massachusetts General Hospital, 15 Parkman Street WACC 440, Boston, MA 02114 USA; E-mail: [email protected]
1 Division of Vascular and Endovascular Surgery, Department of Surgery, Massachusetts General Hospital, Boston, MA.
Ann Vasc Surg 2019; -: 1–9 https://doi.org/10.1016/j.avsg.2018.12.058 Ó 2019 Elsevier Inc. All rights reserved. Manuscript received: July 16, 2018; manuscript accepted: December 6, 2018; published online: - - -
2
Department of Surgery, Massachusetts General Hospital, Boston,
MA. Correspondence to: Laura T. Boitano, MD, Vascular Surgery Research Fellow, Division of Vascular and Endovascular Surgery,
1
2 Boitano et al.
INTRODUCTION Endovascular aneurysm repair (EVAR) has become the predominant treatment modality for infrarenal abdominal aortic aneurysms (AAAs), and as of 2006, 60e70% of all AAA repairs in the United States are being done with this approach.1e4 Randomized controlled trials and observational studies have demonstrated better perioperative outcomes but similar long-term survival compared with open aortic repair (OAR).4 With a decrease in perioperative morbidity, EVAR is associated with enhanced recovery and shorter length of stay compared with OAR.4e8 However, in this modern era of endovascular repair, less is known about those at risk for functional decline and risk factors for rehabilitation following EVAR. Preoperative counseling often focuses on the major morbidity and mortality associated with surgical repair. However, a previous study has found that, regardless of the treatment approach, the ability to perform basic everyday activities even 1 day earlier was viewed as important to patients.9 Interestingly, these same patients did not give this functional outcome much value preoperatively. This suggests that although patients may not prompt this discussion in the preoperative period while focusing on the surgery itself and other major postoperative morbidities, functional outcomes are an important aspect of recovery and should be part of preoperative counseling.9e12 A handful of studies have investigated the risk factors for discharge to additional skilled facilities following vascular surgery; however, there has yet to be an investigation specific to EVAR.10,11,13 A better understanding of baseline preoperative predictors of decline in functional status and need for inpatient rehabilitation is essential for guiding this preoperative conversation. Thus, our aim was to assess the independent preoperative predictors of discharge to a skilled nursing facility or rehab, nonhome discharge (NHD), following elective infrarenal EVAR. In addition, we sought to quantify any downstream impact after NHD including whether there was increased risk of readmission or postdischarge complications.
METHODS Data Source A retrospective cohort study was performed using the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) and the Vascular Targeted Participant Use Data Files for EVAR for years 2011e2015. The
Annals of Vascular Surgery
procedure-targeted EVAR database includes an additional 25 vascular specific variables, and by 2015, 89 participating sites contribute to this module. Further details regarding data collection, auditing, and quality are described in the ACS-NSQIP User Guide.14 The data included in the ACS-NSQIP Vascular Targeted Participant Use Data Files are retrospective in nature and deidentified; thus, they are not considered for human research, and the Massachusetts General Hospital Institutional Review Board waived the need for informed consent. Patient Cohort and Variables All patients who underwent elective infrarenal aortic repair using EVAR were included. Exclusion criteria were fenestrated devices, emergency cases, ruptured or symptomatic aneurysms, admission from any facility other than home, in-hospital deaths, and discharge to unknown destination. Preoperative demographic and clinical information were obtained. Functional status reflected the patient’s ability to perform activities of daily living (ADLs) in the 30 days before surgery.14 Dependent functional status was considered partial or total dependence on others for ADLs. Anemia was defined as hematocrit less than 36%.15 Operative characteristics included groin access (percutaneous versus open femoral access [groin cutdown]), lower extremity revascularization, hypogastric revascularization, and hypogastric embolization. High operative time (>2.5 hr) was evaluated as a possible covariate, considered as a potential marker for a more complex repair, and defined as greater than the median in the cohort. Outcome The primary end point was NHD including discharge to rehabilitation unit, separate acute care, skilled care, or unskilled facility. Secondary outcomes included postoperative complications. These were divided into predischarge, postdischarge, and 30-day outcomes. Outcomes missing day of event were only included as part of the 30-day outcomes. Notably, unplanned reoperation was missing day of event in almost 50% of cases. Major complications included organ space infection, pulmonary complications (pneumonia, reintubation, failure to wean from ventilator [>48 hr]), acute renal failure, stroke, cardiac complications (cardiac arrest, myocardial infarction), bleeding requiring transfusion, venous thromboembolism (deep venous thrombosis and pulmonary embolism), sepsis/septic shock, ischemic colitis, graft failure, return to the operating room, postoperative lower extremity
Volume
-, -
2019
revascularization, and rupture.16e18 Minor complications included superficial wound infection, superficial wound dehiscence, urinary tract infection, and Clostridium difficile infection. Statistical Analysis Categorical variables are presented as counts with percentages, and continuous variables are presented as mean ± standard deviation. Chi-squared test and Fisher’s exact test were used for categorical variables, and t-test was used for continuous variables, where appropriate. Manual stepwise multivariable binary logistic regression determined the independent predictors for NHD. Inclusion criterion was P < 0.100 bivariate, and exclusion criterion was P > 0.05. Model performance was assessed using the concordance statistic (C-statistic) of the receiver-operating curve, and the final model covariates were included to optimize parsimony and the C-statistic. Manual stepwise multivariable binary logistic regression was also used to determine whether NHD predicted unplanned readmission or major postdischarge complications. All analysis was performed with STATA, version 14.1, statistical software (StataCorp LP, College Station, TX).
RESULTS Baseline Covariates and Operative Characteristics From 2011 to 2015, 9,598 cases were entered into the ACS-NSQIP Vascular Targeted Participant Use Data File for EVAR. After inclusion and exclusion criteria were applied, 6,276 cases were evaluated, and of these, 4.6% (291) required NHD. Baseline demographics, comorbidities, and operative characteristics differ as follows: NHD were more frequently female (36.1% vs. 16.7%, P < 0.001), older (80.0 ± 7.7 vs. 73.4 ± 8.6 years, P < 0.001), nonsmokers (75.3% vs. 69.4%, P ¼ 0.037), anemic (26.5% vs. 12.2%, P < 0.001), and underweight (body mass index <18.5) (3.1% vs. 1.3%, P ¼ 0.020), Table I. In addition, they had chronic obstructive pulmonary disease (22.0% vs. 17.1%, P ¼ 0.039), recent congestive heart failure exacerbations (2.8% vs. 1.0%, P ¼ 0.010), open wounds (2.1% vs. 0.7%, P ¼ 0.024), and were more frequently dependent at the baseline (9.3% vs. 1.3%, P < 0.001). The NHD cohort had a more complex repair with a greater proportion of cases requiring high operative time (75.6% vs. 51.3%, P < 0.001).
Predictors of discharge destination after EVAR 3
Furthermore, NHD cases had significantly more bilateral groin cutdowns (63.6% vs. 55.1%, P ¼ 0.004), access vessel repair (8.9% vs. 5.4%, P ¼ 0.011), lower extremity revascularization (6.6% vs. 3.3%, P ¼ 0.003), and hypogastric embolization (10.3% vs. 6.4%, 0.008). Overall 30-Day Complications At 30 days, the NHD cohort had more major complications (37.8% vs. 9.4%, P < 0.001) but had similar rates of minor complications (3.8% vs. 2.3%, P ¼ 0.119). Specific complications are presented in Table II. Notably, for NHD versus home discharge, death (1.4% vs. 0.3%, P ¼ 0.011), unplanned return to the operating room (13.4% vs. 2.9%, P < 0.001), and postoperative lower extremity revascularization (4.1% vs. 1.0%, P < 0.001) were more frequent in NHD. However, unplanned readmission did not significantly differ between the 2 groups (9.3% vs. 6.6%, P ¼ 0.070). Predischarge Complications During the index hospitalization, all complications assessed on unadjusted analyses were more frequent in the NHD cohort except for venous thromboembolism and rupture, Table II. These included more major complications (32.7% vs. 6.7%, P < 0.001), minor complications (2.8% vs. 0.4%, P < 0.001), wound complications (1.0% vs. 0.1%, P ¼ 0.001), pulmonary complications (5.8% vs. 0.5%, P < 0.001), cardiac complications (5.2% vs. 0.4%, P < 0.001), and renal complications (3.8% vs. 0.1%, P < 0.001). The NHD cohort suffered more strokes (1.7% vs. 0.1%, P < 0.001), were more likely to require blood transfusion (22.3% vs. 5.5%, P < 0.001), and develop a sepsis/ shock (2.4% vs. 0.1%, P < 0.001) before discharge. Urinary tract infections were more common in NHD before discharge (2.1% vs. 0.3%, P < 0.001) as was ischemic colitis (2.1% vs. 0.2%, P < 0.001). Patients requiring NHD more likely had an unplanned return to the operating room (0.7% vs. 0.1%, P < 0.001), required lower extremity revascularization postoperatively (3.1% vs. 0.6%, P < 0.001), and had a longer overall (5 ± 2 vs. 1 ± 1 days, P < 0.001) and intensive care unit (ICU) (0 ± 2 vs. 0 ± 1 days, P < 0.001) length of stay. Postdischarge Complications In contrast, postdischarge complications were overall similar between the 2 groups, Table II. Major complications were more frequent in the NHD cohort (5.2% vs. 3.0%, P ¼ 0.037); however, minor complications did
4 Boitano et al.
Annals of Vascular Surgery
Table I. Baseline demographics and comorbidities Covariates a
Age (years) Female White race ASA class 3 or higher BMI <18.5 >10% weight lossb Albumin <3.0 Dependent functional status Current steroid use Current smoker COPD HTN CHF exacerbationc Diabetes mellitus ESRD on dialysis Bleeding disorder Anemia (hct<36%) Disseminated cancer Open wound Aneurysm diameter (cm)a High operative time Bilateral open femoral access Access vesselsdconduit, repair Lower extremity revascularization Hypogastric revascularization Hypogastric embolization Convert to open procedure
Total (n ¼ 6,476)
Discharge home (n ¼ 6,185)
NHD (n ¼ 291)
P
73.8 ± 8.7 1,137 (17.6%) 5376 (92.8%) 6,064 (93.6%) 91 (1.4%) 42 (0.7%) 3801 (58.7%) 106 (1.6%) 277 (4.3%) 1,962 (30.3%) 1,124 (17.4%) 5187 (80.1%) 67 (1.0%) 1,048 (16.2%) 48 (0.7%) 732 (11.3%) 832 (12.9%) 31 (0.5%) 50 (0.8%) 5.6 ± 1.4 3,393 (53.4%) 3,574 (55.5%) 361 (5.8%) 207 (3.4%) 289 (4.5%) 424 (6.6%) 31 (0.5%)
73.4 ± 8.6 1,032 (16.7%) 5124 (92.9%) 5784 (93.5%) 82 (1.3%) 40 (0.7%) 3640 (58.9%) 79 (1.3%) 262 (4.2%) 1,890 (30.6%) 1,060 (17.1%) 4949 (80.0%) 59 (1.0%) 995 (16.1%) 48 (0.8%) 692 (11.2%) 755 (12.2%) 31 (0.5%) 44 (0.7%) 5.6 ± 1.4 3,173 (51.3%) 3,389 (55.1%) 335 (5.4%) 189 (3.3%) 273 (4.4%) 394 (6.4%) 30 (0.5%)
80.0 ± 7.7 105 (36.1%) 252 (91.0%) 280 (96.2%) 9 (3.1%) 3 (1.0%) 162 (55.6%) 27 (9.3%) 15 (5.2%) 72 (24.7%) 64 (22.0%) 238 (81.8%) 8 (2.8%) 53 (18.2%) 0 (0%) 40 (13.8%) 77 (26.5%) 0 (0%) 6 (2.1%) 5.8 ± 1.2 220 (75.6%) 185 (63.6%) 26 (8.9%) 18 (6.6%) 16 (5.5%) 30 (10.3%) 2 (0.7%)
<0.001 <0.001 0.223 0.065 0.020 0.442 0.281 <0.001 0.449 0.037 0.039 0.460 0.010 0.336 0.277 0.184 <0.001 0.401 0.024 0.015 <0.001 0.004 0.011 0.003 0.382 0.008 0.654
ASA, American Society of Anesthesiologists; BMI, body mass index; COPD, chronic obstructive pulmonary disease; HTN, hypertension; CHF, congestive heart failure; ESRD, end-stage renal disease; Hct, hematocrit. a Mean ± standard deviation. b Within last 6 months. c Within last 30 days.
not differ significantly (1.0% vs. 2.0%, P ¼ 0.380). Pulmonary complications (1.4% vs. 0.2%, P ¼ 0.002) and death (1.4% vs. 0.3%, P ¼ 0.011) were the only postdischarge complication seen more frequently in the NHD cohort. Multivariable Model for NHD Multivariable binary logistic regression demonstrated the independent predictors for discharge to NHD, Table III. Octogenarians(oddsratio[OR]: 5.7, 95%confidenceinterval [CI], 2.3e14.1]; P < 0.001) and nonagenarians (OR: 14.6, 95% CI: 5.4e39.2; P < 0.001) and dependent functional status (OR: 5.4, 95% CI: 3.3e8.8; P < 0.001) were the strongest predictors of NHD. Other independent predictors of NHD identified in the model included female sex (OR: 2.2, 95% CI: 1.7e2.8, P < 0.001), open wound at baseline (OR: 3.5, 95% CI: 1.4e8.9; P ¼ 0.006), high operative time (OR: 2.7, 95% CI: 2.0e3.6; P < 0.001), and hypogastric embolization (OR: 1.6, 95% CI: 1.1e2.1 P ¼ 0.022). The final model had a C-statistic of 0.780. We included only preoperative covariates in the final model; however,
on sensitivity analysis using inpatient covariates, major predischarge complications (OR: 4.1, 95% CI: 3.1e5.6; P < 0.001) and ICU stay (OR: 1.7, 95% CI: 1.3e2.2; P < 0.001) were independently associated with NHD. On multivariable analysis, NHD did not independently predict unplanned readmission (OR 1.0, 95% CI: 0.6e1.5, P ¼ 0.842) after adjusting for female sex (OR 1.6, 95% CI: 1.3e2.0; P < 0.001), anemia (OR: 1.5, 95% CI: 1.1e1.9; P ¼ 0.003), current steroid use (OR: 2.0, 95% CI: 1.4e3.0; P < 0.001), hypertension (OR: 1.4, 95% CI: 1.0e1.8; P ¼ 0.030), high operative time (OR: 1.4, 95% CI: 1.1e1.7; P ¼ 0.002), bilateral groin cutdown (OR: 1.2, 95% CI: 1.0e 1.5; P ¼ 0.049), admission to the ICU (OR: 1.5, 95% CI: 1.2e1.8; P < 0.001), and major inpatient complications (OR: 1.4, 95% CI: 1.0e1.9, P ¼ 0.027). Similarly, NHD did not independently predict major postdischarge complication (OR: 1.0 95% CI: 0.6e1.9; P ¼ 0.875) after adjusting for preoperative anemia (OR: 1.6, 95% CI: 1.1e2.4; P ¼ 0.008), preoperative open wound (OR: 3.5,
Volume
-, -
2019
Predictors of discharge destination after EVAR 5
Table II. Thirty-day, predischarge, and postdischarge outcomes Outcome
Wound complications 30 day Predischarge Postdischarge Pulmonary complications 30 day Predischarge Postdischarge Cardiac complications 30 day Predischarge Postdischarge Stroke 30 day Predischarge Postdischarge Renal complications 30 day Predischarge Postdischarge Urinary tract infection 30 day Predischarge Postdischarge Venous thromboembolism 30 day Predischarge Postdischarge Bleeding requiring transfusion 30 day Predischarge Postdischarge Sepsis/septic shock 30 day Predischarge Postdischarge Ischemic colitis 30 day Predischarge Postdischarge Postoperative lower extremity revascularization 30 day Predischarge Postdischarge Unplanned reoperationa 30 day Predischarge Postdischarge Rupture 30 day Predischarge Postdischarge
Total (n ¼ 6,476)
Discharge home (n ¼ 6,185)
NHD (n ¼ 291)
P
84 (1.3%) 7 (0.1%) 77 (1.2%)
79 (1.3%) 4 (0.1%) 75 (1.2%)
5 (1.7%) 3 (1.0%) 2 (0.7%)
0.428 0.001 0.584
56 (0.9%) 45 (0.7%) 13 (0.2%)
37 (0.6%) 28 (0.5%) 9 (0.2%)
19 (6.5%) 17 (5.8%) 4 (1.4%)
<0.001 <0.001 0.002
57 (0.9%) 39 (6.0%) 18 (0.3%)
41 (0.7%) 24 (0.4%) 17 (0.3%)
16 (5.5%) 15 (5.2%) 1 (0.3%)
<0.001 <0.001 0.563
15 (0.2%) 7 (0.1%) 8 (0.1%)
10 (0.2%) 2 (0.1%) 8 (0.1%)
5 (1.7%) 5 (1.7%) 0 (0%)
<0.001 <0.001 >0.999
29 (0.5%) 19 (0.3%) 10 (0.2%)
17 (0.5%) 8 (0.1%) 9 (0.2%)
12 (4%) 11 (3.8%) 1 (0.3%)
<0.001 <0.001 0.369
79 (1.2%) 24 (0.4%) 55 (0.9%)
72 (1.2%) 18 (0.3%) 54 (0.9%)
7 (2.4%) 6 (2.1%) 1 (0.3%)
0.089 <0.001 0.518
28 (0.4%) 6 (0.1%) 22 (0.3%)
24 (0.4%) 5 (0.1%) 19 (0.3%)
4 (1.4%) 1 (0.3%) 3 (1.0%)
0.035 0.241 0.074
405 (6.3%) 405 (6.3%) 0 (0%)
340 (5.5%) 340 (5.5%) 0 (0%)
65 (22.3%) 65 (22.3%) 0 (0%)
<0.001 <0.001 >0.999
43 (0.7%) 13 (0.2 %) 30 (0.5%)
35 (0.6%) 6 (0.1%) 29 (0.5%)
8 (2.8%) 7 (2.4%) 1 (0.3%)
0.001 <0.001 >0.999
24 (0.4%) 17 (0.3%) 7 (0.1%)
17 (0.3%) 10 (0.2%) 6 (0.1%)
7 (2.4%) 6 (2.1%) 1 (0.3%)
<0.001 <0.001 0.275
76 (1.2%) 44 (0.7%) 30 (0.5%)
64 (1.0%) 35 (0.6%) 28 (0.5%)
12 (4.1%) 9 (3.1%) 2 (0.7%)
<0.001 <0.001 0.393
219 (3.8%) 7 (0.1%) 123 (1.9%)
180 (2.9%) 5 (0.1%) 115 (1.9%)
39 (13.4%) 2 (0.7%) 8 (2.8%)
<0.001 0.036 0.268
2 (0.1%) 1 (0.1%) 1 (0.1%)
1 (0.1%) 1 (0.1%) 1 (0.1%)
0 (0%) 0 (0%) 0 (0%)
>0.999 >0.999 >0.999 (Continued)
6 Boitano et al.
Annals of Vascular Surgery
Table II. Continued Outcome
Total (n ¼ 6,476)
Discharge home (n ¼ 6,185)
NHD (n ¼ 291)
P
ICU length of stay (days)b Length of stay (days)b Unplanned readmission Death Minor complications 30 day Predischarge Postdischarge Major complications 30 day Predischarge Postdischarge
0±1 1±1 433 (6.7%) 20 (0.3%)
0±1 1±1 406 (6.6%) 15 (0.3%)
0±1 5±2 27 (9.3%) 5 (1.4%)
<0.001 <0.001 0.070 0.011
156 (2.4%) 30 (0.5%) 126 (2.0%)
145 (2.3%) 22 (0.4%) 123 (2.0%)
11 (3.8%) 8 (2.8%) 3 (1.0%)
0.119 <0.001 0.380
694 (10.7%) 509 (7.9%) 200 (3.1%)
584 (9.4%) 414 (6.7%) 185 (3.0%)
110 (37.8%) 95 (32.7%) 15 (5.2%)
<0.001 <0.001 0.037
ICU, intensive care unit. a Missing dates are only included in overall 30-day reoperation. b Median ± interquartile range.
Table III. Logistic regression for nonhome discharge Covariate
Odds ratio
95% confidence interval
P value
Female Age <60 Age 60e69 Age 70e79 Age 80e89 Age 90+ Dependent functional status Open wound High operative time Hypogastric embolization C-statistic
2.2 Reference 1.2 1.8 5.7 14.6 5.4 3.5 2.7 1.6
1.7e2.9
<0.001
0.4e3.1 0.7e4.6 2.3e14.1 5.4e39.2 3.3e8.8 1.4e8.9 2.0e3.6 1.1e2.1
0.754 0.193 <0.001 <0.001 <0.001 0.006 <0.001 0.022 0.780
95% CI: 1.3e9.2; P ¼ 0.012), index procedural lower extremity revascularization (OR: 2.0, 95% CI: 1.1e3.5; P ¼ 0.015), admission to the ICU (OR: 1.5, 95% CI: 1.1e2.0; P ¼ 0.013), and major inpatient complications (OR: 2.3, 95% CI: 1.5e3.4; P < 0.001).
DISCUSSION This is one of the first studies to evaluate preoperative predictors of discharge to rehab after EVAR with procedure-specific data.10,11,13 The present study found 4.6% of previously home-dwelling patients require NHD following EVAR, and preoperative covariates predicted NHD with good discriminatory power. Independent predictors included female sex, age 80 years, dependent functional status, baseline open wound, high operative time, and hypogastric embolization. Age was the strongest predictor of discharge destination, and dependent
functional status was an important predictor with 5 times the odds of discharge to NHD compared with independent functioning patients. Not surprisingly, NHD patients experienced significantly more predischarge complications, including major complications, and were more likely to be admitted to the ICU. Although it has previously been shown that discharge to NHD is associated with worse postdischarge outcomes, NHD was not an independent predictor of readmission or postdischarge 30-day complications in the present study.11,19e23 In the present study, advanced age was the strongest predictor of discharge destination. Nine percent of octogenarians and 19.6% of nonagenarians required NHD as compared with 1.5% of patients younger than 60 years. Age is a well-known risk factor for worse outcome following EVAR.4,8,19,24,25 Gupta et al.24 found increasing age was associated with postdischarge 30-day morbidity and mortality in patients undergoing
Volume
-, -
2019
EVAR. Eslami et al.25 evaluated all patients undergoing elective AAA repair, both OAR and EVAR, and again older age (>70 years) was an independent predictor of 30-day mortality. In addition, our findings are consistent with other reports that advanced age is associated with NHD following vascular operations. In a recent study by Schermerhorn et al. evaluating the Medicare population, 15% of patients 85 years required NHD as compared with 2% of patients aged 67e69 years.4 Furthermore, in a study evaluating predictors of NHD following elective vascular surgery, patients 85 years had a 5-fold increased odds of NHD as compared with patients aged 65 years and younger.11 Despite the known importance of age on outcomes, there are some experts who suggest, regardless of chronological age, frailty is the most important determinant of outcome.26e28 There is currently no gold standard for measuring frailty, but regardless of how it is measured, studies have described worse outcomes in frail patients.11,27e30 Most frailty indices include variables evaluating preoperative nutritional status, weight loss, functional status, and baseline comorbid conditions.11,28e30 In the present study, dependent functional status was an important predictor of NHD, demonstrating 5 times the odds of NHD compared with patients independent at baseline. However, after risk adjustment, other indictors of frailty were not independent predictors of NHD. Gender differences in outcomes of aneurysm repair been reported previously, and female sex has been associated with poor outcomes after both EVAR and OAR.11,24,31e37 Consistent with other reports, this study identified female sex as an independent predictor of NHD.10,11 Several studies have postulated explanations for such differences and suggest that women have more complex anatomy, smaller vessel diameter, and untreated comorbid conditions.10,11,36,37 In addition, studies suggest that women are often in the caregiver role and might have a sicker family that is unable to care for them. Several reports have stated that advanced age at the time of treatment also contributes to worse outcomes and discharge to a skilled facility as they are likely to have outlived their spouse or caregiver.24,31e35 In accordance with this, female patients in this study were on average older by 3 years as compared with men (76 ± 8.7 vs. 73 ± 8.5 years, P < 0.001). Without socioeconomic factors, there is limited understanding of how social support impacted discharge status. Even with the advent of iliac branched devices and revascularization options to address perfusion
Predictors of discharge destination after EVAR 7
to the internal iliac artery, endovascular treatment of internal iliac artery aneurysms is most commonly performed by coil embolization with extension of the graft limb into the external iliac artery.32 A recent study evaluating concomitant hypogastric artery embolization during EVAR found that overall patients had increased operative time, more likely to require blood transfusion, and had longer ICU and hospital stay as well as a higher likelihood of discharge to NHD (12% vs. 7.8%, P ¼ 0.008).32 Our findings are consistent with this study, showing hypogastric embolization was also an independent predictor of discharge to NHD and was associated with a 60% increased risk of NHD (OR: 1.6, CI: 1.1e2.1, P ¼ 0.022). The inclusion of hypogastric coiling demonstrates the importance of procedurespecific variables and evaluating each procedure individually. High operative time was an independent predictor of discharge to NHD (OR: 2.7, CI: 2.0e3.6, P < 0.001) and was associated with bilateral groin cutdown on bivariate analysis (59.9% vs. 43.0%, P < 0.001), attempted percutaneous access but converted to groin cutdown (52.0% vs. 84.3%, P < 0.001), iliac stent placement (61.9% vs. 51.6%, P < 0.001), hypogastric revascularization (68.5% vs. 51.6%, P < 0.001), hypogastric embolization (78.1 vs. 50.6, P < 0.001), and lower extremity revascularization (88.4% vs. 50.0%, P < 0.001). In a recent study evaluating outcomes after EVAR with or without hypogastric coiling, longer-than-average operative repair was associated with increased risk of 30-day mortality, ischemic colitis, and renal failure requiring dialysis.32 In addition, Arya et al.11 found for every increase in an hour of operative time, there was a 26% increased odds of NHD in patients undergoing nonspecific vascular procedures. This increase in operative time could have been more attributed to differences in the vascular procedures evaluated. The present study demonstrates that within the same procedure type, EVAR, this is an important variable to evaluate. Interestingly, open wound at baseline was associated with 3 times the odds of discharge to NHD. Crouch et al. evaluated risk factors associated with decline in independent living status after vascular surgery and found open wound at discharge, primarily related to the surgical procedure, independently predicted such decline.13 However, this study uniquely identifies the importance of preoperative wounds on need for NHD after EVAR. Finally, this study did not find an association between NHD and readmission or postdischarge complication. This finding is in contrast to other reports, across surgical subspecialties, which have
8 Boitano et al.
found that NHD is associated with increased morbidity, mortality, and readmission.11,19e23 Arya et al.11 reported a 6-fold increase in postdischarge mortality and 2-fold increase in readmission rate in vascular surgery patients requiring NHD. However, this study failed to perform an adjusted analysis to evaluate this association, which may contribute to the difference in our findings. Greenblatt et al. found that after EVAR and OAR, discharge to destination other than home predicted readmission, and it was the most powerful predictor of early readmission.19 The difference in our findings with those of Greenblatt et al. may be related to the inclusion of OAR in their study. The present study did find an association on unadjusted analysis between NHD and 30-day death; multivariable analysis could not be conducted to determine whether this association held after risk adjustment due to small outcome numbers. In light of trial data from the EVAR 2 study, careful consideration and patient counseling based on individual risk factors are warranted as EVAR may not provide adequate risk reduction in high-risk populations, including those at high risk for NHD.38
Study Limitations This study is limited in that it utilizes a large registry such that the analysis is limited to the variables which are captured by the NSQIP Vascular Targeted EVAR database. There is limited information regarding socioeconomic status and social support, which may be important in the prediction of discharge destination. Furthermore, there were a number of variables with a large amount of missingness and thus could not be included as potential predictors of NHD. These included impaired sensorium, preoperative myocardial infarction, and history of hemiplegia or paraplegia. In addition, the NSQIP database does not include hospital identification such that a mixed-effects model could not be constructed to account for individual hospital practices. Finally, the analysis is limited to 30 days so this study was unable to evaluate whether NHD has implications on long-term survival or functional status. This study identifies independent risk factors for NHD. However, these predictors have not been prospectively validated. Thus, future studies should focus on validation of these findings. Furthermore, future studies should seek to determine a risk prediction score such that patients can be counseled preoperatively on the risk for NHD. In addition, studies should look at the cost savings and hospital
Annals of Vascular Surgery
length of stay associated with early identification of patients who are likely to be discharged to NHD.
CONCLUSION This is one of the first studies evaluating preoperative predictors of NHD following elective EVAR. In this study, using 6 preoperative variables on multivariable analysis, we were able to have high predictability of discharge destination. Age greater than 80 years and dependent functional status were the most important predictors of discharge destination. Future studies should seek to validate these findings in a prospective manner. Patients discharged to rehab did not experience increased postdischarge perioperative major complications and are not at higher risk for readmission following EVAR. These data are important for understanding the risk of NHD following elective EVAR and can help to better counsel patients with such characteristics who are at higher risk for NHD. REFERENCES 1. Levin DC, Rao VM, Parker L, et al. Endovascular repair vs open surgical repair of abdominal aortic aneurysms: comparative utilization trends from 2001 to 2006. J Am Coll Radiol 2009;6:506e9. 2. Schwarze ML, Shen Y, Hemmerich J, et al. Age-related trends in utilization and outcome of open and endovascular repair for abdominal aortic aneurysm in the United States, 2001-2006. J Vasc Surg 2009;50:722e729.e2. 3. Brewster DC, Jones JE, Chung TK, et al. Long-term outcomes after endovascular abdominal aortic aneurysm repair: the first decade. Ann Surg 2006;244:426e38. 4. Schermerhorn ML, Buck DB, O’Malley AJ, et al. Long-term outcomes of abdominal aortic aneurysm in the Medicare population. N Engl J Med 2015;373:328e38. 5. Prinssen M, Verhoeven EL, Buth J, et al. A randomized trial comparing conventional and endovascular repair of abdominal aortic aneurysms. N Engl J Med 2004;351:1607e18. 6. Participants Et.. Endovascular aneurysm repair versus open repair in patients with abdominal aortic aneurysm (EVAR trial 1): randomised controlled trial. Lancet 2005;365: 2179e86. 7. Lederle FA, Freischlag JA, Kyriakides TC, et al. Outcomes following endovascular vs open repair of abdominal aortic aneurysm: a randomized trial. JAMA 2009;302:1535e42. 8. Lederle FA, Freischlag JA, Kyriakides TC, et al. Long-term comparison of endovascular and open repair of abdominal aortic aneurysm. N Engl J Med 2012;367:1988e97. 9. Faggioli G, Scalone L, Mantovani LG, et al., Group Ps. Preferences of patients, their family caregivers and vascular surgeons in the choice of abdominal aortic aneurysms treatment options: the PREFER study. Eur J Vasc Endovasc Surg 2011;42:26e34. 10. Vogel TR, Nackman GB, Crowley JG, et al. Factors impacting functional health and resource utilization following abdominal aortic aneurysm repair by open and endovascular techniques. Ann Vasc Surg 2005;19:641e7.
Volume
-, -
2019
11. Arya S, Long CA, Brahmbhatt R, et al. Preoperative frailty increases risk of Nonhome discharge after elective vascular surgery in home-dwelling patients. Ann Vasc Surg 2016;35:19e29. 12. Aljabri B, Al Wahaibi K, Abner D, et al. Patient-reported quality of life after abdominal aortic aneurysm surgery: a prospective comparison of endovascular and open repair. J Vasc Surg 2006;44:1182e7. 13. Crouch DS, McLafferty RB, Karch LA, et al. A prospective study of discharge disposition after vascular surgery. J Vasc Surg 2001;34:62e8. 14. User Guide for the 2015 ACS NSQIP Procedure Targeted Participant Use Data Gile (PUF). Available at: https://www. facs.org/w/media/files/quality%20programs/nsqip/pt._nsqip_ puf_user_guide_2015.ashx. Accessed May 1, 2018. 15. Iannuzzi JC, Chandra A, Kelly KN, et al. Risk score for unplanned vascular readmissions. J Vasc Surg 2014;59:1340e1347.e1. 16. Shields E, Iannuzzi JC, Thorsness R, et al. Postoperative morbidity by procedure and patient factors Influencing major complications within 30 Days following shoulder surgery. Orthop J Sports Med 2014;2:2325967114553164. 17. Iannuzzi JC, Chandra A, Rickles AS, et al. Resident involvement is associated with worse outcomes after major lower extremity amputation. J Vasc Surg 2013;58:827e831.e1. 18. Iannuzzi JC, Rickles AS, Deeb AP, et al. Outcomes associated with resident involvement in partial colectomy. Dis Colon Rectum 2013;56:212e8. 19. Greenblatt DY, Greenberg CC, Kind AJ, et al. Causes and implications of readmission after abdominal aortic aneurysm repair. Ann Surg 2012;256:595e605. 20. Williamson WK, Nicoloff AD, Taylor LM Jr, et al. Functional outcome after open repair of abdominal aortic aneurysm. J Vasc Surg 2001;33:913e20. 21. Legner VJ, Massarweh NN, Symons RG, et al. The significance of discharge to skilled care after abdominopelvic surgery in older adults. Ann Surg 2009;249:250e5. 22. Ottenbacher KJ, Karmarkar A, Graham JE, et al. Thirty-day hospital readmission following discharge from postacute rehabilitation in fee-for-service Medicare patients. JAMA 2014;311:604e14. 23. Palmieri TL, Molitor F, Chan G, et al. Long-term functional outcomes in the elderly after burn injury. J Burn Care Res 2012;33:497e503. 24. Gupta PK, Engelbert TL, Ramanan B, et al. Postdischarge outcomes after endovascular abdominal aortic aneurysm repair. J Vasc Surg 2014;59:903e8. 25. Eslami MH, Rybin DV, Doros G, et al. Description of a risk predictive model of 30-day postoperative mortality after elective abdominal aortic aneurysm repair. J Vasc Surg 2017;65:65e74.e2.
Predictors of discharge destination after EVAR 9
26. Rockwood K, Howlett SE, MacKnight C, et al. Prevalence, attributes, and outcomes of fitness and frailty in community-dwelling older adults: report from the Canadian study of health and aging. J Gerontol A Biol Sci Med Sci 2004;59:1310e7. 27. Lin HS, Watts JN, Peel NM, et al. Frailty and post-operative outcomes in older surgical patients: a systematic review. BMC Geriatr 2016;16:157. 28. Makary MA, Segev DL, Pronovost PJ, et al. Frailty as a predictor of surgical outcomes in older patients. J Am Coll Surg 2010;210:901e8. 29. Ambler GK, Brooks DE, Al Zuhir N, et al. Effect of frailty on short- and mid-term outcomes in vascular surgical patients. Br J Surg 2015;102:638e45. 30. Arya S, Kim SI, Duwayri Y, et al. Frailty increases the risk of 30-day mortality, morbidity, and failure to rescue after elective abdominal aortic aneurysm repair independent of age and comorbidities. J Vasc Surg 2015;61:324e31. 31. Zettervall SL, Schermerhorn ML, Soden PA, et al. The effect of surgeon and hospital volume on mortality after open and endovascular repair of abdominal aortic aneurysms. J Vasc Surg 2017;65:626e34. 32. Farivar BS, Kalsi R, Drucker CB, et al. Implications of concomitant hypogastric artery embolization with endovascular repair of infrarenal abdominal aortic aneurysms. J Vasc Surg 2017;66:95e101. 33. Dillavou ED, Muluk SC, Makaroun MS. A decade of change in abdominal aortic aneurysm repair in the United States: have we improved outcomes equally between men and women? J Vasc Surg 2006;43:230e8. discussion 8. 34. Egorova NN, Vouyouka AG, McKinsey JF, et al. Effect of gender on long-term survival after abdominal aortic aneurysm repair based on results from the Medicare national database. J Vasc Surg 2011;54:1e12.e6. discussion 1e2. 35. Lo RC, Bensley RP, Hamdan AD, et al. Gender differences in abdominal aortic aneurysm presentation, repair, and mortality in the Vascular Study Group of New England. J Vasc Surg 2013;57:1261e8.e1e5. 36. Deery SE, Soden PA, Zettervall SL, et al. Sex differences in mortality and morbidity following repair of intact abdominal aortic aneurysms. J Vasc Surg 2017;65:1006e13. 37. Deery SE, Shean KE, Wang GJ, et al. Female sex independently predicts mortality after thoracic endovascular aortic repair for intact descending thoracic aortic aneurysms. J Vasc Surg 2017;66:2e8. 38. participants Et.. Endovascular aneurysm repair and outcome in patients unfit for open repair of abdominal aortic aneurysm (EVAR trial 2): randomised controlled trial. Lancet 2005;365:2187e92.
Background: There is a paucity of data guiding preoperative counseling on the need for discharge to a facility or nonhome discharge (NHD) following elective endovascular repair of abdominal aortic aneurysms (endovascular aneurysm repair [EVAR]). This study seeks to determine the preoperative predictors of NHD following EVAR in baseline home-dwelling patients and to determine whether NHD is associated with major postdischarge complications and readmission. Methods: This retrospective cohort study utilized the National Surgical Quality Improvement Program Vascular Procedure Targeted database to identify elective EVAR cases admitted from home (2011 to 2015). The primary end point was NHD. A multivariable logistic regression model was used to determine predictive preoperative factors for NHD and to determine whether NHD predicted major postdischarge complications and readmission. Results: Overall 6,276 cases were included; 291 (4.6%) required NHD. NHD were more frequently female, anemic, functionally dependent, nonsmokers, had chronic obstructive pulmonary disease, recent congestive heart failure exacerbation, and open baseline wounds. NHD was associated with complex surgery, indicated by operative time more than the median, 2.5 hr. Significant predictors for NHD on multivariable analysis included female sex (odds ratio [OR]: 2.2, confidence interval [CI]: 1.7e2.9, P < 0.001), octogenarians (OR: 5.7 CI: 2.3e14.1; P < 0.001) and nonagenarians (OR: 14.6, CI: 5.4e39.2; P < 0.001), dependent functional status (OR: 5.4, CI: 3.3e8.8; P < 0.001), preoperative open wound (OR: 3.5, CI: 1.4e8.9; P ¼ 0.006), high operative time (OR: 2.7, CI: 2.0e3.6; P < 0.001), and hypogastric embolization (OR: 1.6, CI: 1.1e2.1 P ¼ 0.022), C-statistic ¼ 0.780. On adjusted analysis, NHD did not independently predict major postdischarge complication (OR: 1.0 CI: 0.6e1.9; P ¼ 0.875) or unplanned readmission (OR 1.0, CI: 0.6e1.5, P ¼ 0.842). Conclusions: Discharge to skilled facility following EVAR can be predicted using preoperative factors. Future studies should seek to validate these findings in a prospective manner. Identifying high-risk patients’ NHD can help define expectations and facilitate early referral to skilled facilities that may reduce hospital length of stay, reducing health-care costs.
Presented at the 42nd Annual Meeting of the Vascular and Endovascular Surgery Society, February 1e4, 2018, Vail, CO.
Massachusetts General Hospital, 15 Parkman Street WACC 440, Boston, MA 02114 USA; E-mail: [email protected]
1 Division of Vascular and Endovascular Surgery, Department of Surgery, Massachusetts General Hospital, Boston, MA.
Ann Vasc Surg 2019; -: 1–9 https://doi.org/10.1016/j.avsg.2018.12.058 Ó 2019 Elsevier Inc. All rights reserved. Manuscript received: July 16, 2018; manuscript accepted: December 6, 2018; published online: - - -
2
Department of Surgery, Massachusetts General Hospital, Boston,
MA. Correspondence to: Laura T. Boitano, MD, Vascular Surgery Research Fellow, Division of Vascular and Endovascular Surgery,
1
2 Boitano et al.
INTRODUCTION Endovascular aneurysm repair (EVAR) has become the predominant treatment modality for infrarenal abdominal aortic aneurysms (AAAs), and as of 2006, 60e70% of all AAA repairs in the United States are being done with this approach.1e4 Randomized controlled trials and observational studies have demonstrated better perioperative outcomes but similar long-term survival compared with open aortic repair (OAR).4 With a decrease in perioperative morbidity, EVAR is associated with enhanced recovery and shorter length of stay compared with OAR.4e8 However, in this modern era of endovascular repair, less is known about those at risk for functional decline and risk factors for rehabilitation following EVAR. Preoperative counseling often focuses on the major morbidity and mortality associated with surgical repair. However, a previous study has found that, regardless of the treatment approach, the ability to perform basic everyday activities even 1 day earlier was viewed as important to patients.9 Interestingly, these same patients did not give this functional outcome much value preoperatively. This suggests that although patients may not prompt this discussion in the preoperative period while focusing on the surgery itself and other major postoperative morbidities, functional outcomes are an important aspect of recovery and should be part of preoperative counseling.9e12 A handful of studies have investigated the risk factors for discharge to additional skilled facilities following vascular surgery; however, there has yet to be an investigation specific to EVAR.10,11,13 A better understanding of baseline preoperative predictors of decline in functional status and need for inpatient rehabilitation is essential for guiding this preoperative conversation. Thus, our aim was to assess the independent preoperative predictors of discharge to a skilled nursing facility or rehab, nonhome discharge (NHD), following elective infrarenal EVAR. In addition, we sought to quantify any downstream impact after NHD including whether there was increased risk of readmission or postdischarge complications.
METHODS Data Source A retrospective cohort study was performed using the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) and the Vascular Targeted Participant Use Data Files for EVAR for years 2011e2015. The
Annals of Vascular Surgery
procedure-targeted EVAR database includes an additional 25 vascular specific variables, and by 2015, 89 participating sites contribute to this module. Further details regarding data collection, auditing, and quality are described in the ACS-NSQIP User Guide.14 The data included in the ACS-NSQIP Vascular Targeted Participant Use Data Files are retrospective in nature and deidentified; thus, they are not considered for human research, and the Massachusetts General Hospital Institutional Review Board waived the need for informed consent. Patient Cohort and Variables All patients who underwent elective infrarenal aortic repair using EVAR were included. Exclusion criteria were fenestrated devices, emergency cases, ruptured or symptomatic aneurysms, admission from any facility other than home, in-hospital deaths, and discharge to unknown destination. Preoperative demographic and clinical information were obtained. Functional status reflected the patient’s ability to perform activities of daily living (ADLs) in the 30 days before surgery.14 Dependent functional status was considered partial or total dependence on others for ADLs. Anemia was defined as hematocrit less than 36%.15 Operative characteristics included groin access (percutaneous versus open femoral access [groin cutdown]), lower extremity revascularization, hypogastric revascularization, and hypogastric embolization. High operative time (>2.5 hr) was evaluated as a possible covariate, considered as a potential marker for a more complex repair, and defined as greater than the median in the cohort. Outcome The primary end point was NHD including discharge to rehabilitation unit, separate acute care, skilled care, or unskilled facility. Secondary outcomes included postoperative complications. These were divided into predischarge, postdischarge, and 30-day outcomes. Outcomes missing day of event were only included as part of the 30-day outcomes. Notably, unplanned reoperation was missing day of event in almost 50% of cases. Major complications included organ space infection, pulmonary complications (pneumonia, reintubation, failure to wean from ventilator [>48 hr]), acute renal failure, stroke, cardiac complications (cardiac arrest, myocardial infarction), bleeding requiring transfusion, venous thromboembolism (deep venous thrombosis and pulmonary embolism), sepsis/septic shock, ischemic colitis, graft failure, return to the operating room, postoperative lower extremity
Volume
-, -
2019
revascularization, and rupture.16e18 Minor complications included superficial wound infection, superficial wound dehiscence, urinary tract infection, and Clostridium difficile infection. Statistical Analysis Categorical variables are presented as counts with percentages, and continuous variables are presented as mean ± standard deviation. Chi-squared test and Fisher’s exact test were used for categorical variables, and t-test was used for continuous variables, where appropriate. Manual stepwise multivariable binary logistic regression determined the independent predictors for NHD. Inclusion criterion was P < 0.100 bivariate, and exclusion criterion was P > 0.05. Model performance was assessed using the concordance statistic (C-statistic) of the receiver-operating curve, and the final model covariates were included to optimize parsimony and the C-statistic. Manual stepwise multivariable binary logistic regression was also used to determine whether NHD predicted unplanned readmission or major postdischarge complications. All analysis was performed with STATA, version 14.1, statistical software (StataCorp LP, College Station, TX).
RESULTS Baseline Covariates and Operative Characteristics From 2011 to 2015, 9,598 cases were entered into the ACS-NSQIP Vascular Targeted Participant Use Data File for EVAR. After inclusion and exclusion criteria were applied, 6,276 cases were evaluated, and of these, 4.6% (291) required NHD. Baseline demographics, comorbidities, and operative characteristics differ as follows: NHD were more frequently female (36.1% vs. 16.7%, P < 0.001), older (80.0 ± 7.7 vs. 73.4 ± 8.6 years, P < 0.001), nonsmokers (75.3% vs. 69.4%, P ¼ 0.037), anemic (26.5% vs. 12.2%, P < 0.001), and underweight (body mass index <18.5) (3.1% vs. 1.3%, P ¼ 0.020), Table I. In addition, they had chronic obstructive pulmonary disease (22.0% vs. 17.1%, P ¼ 0.039), recent congestive heart failure exacerbations (2.8% vs. 1.0%, P ¼ 0.010), open wounds (2.1% vs. 0.7%, P ¼ 0.024), and were more frequently dependent at the baseline (9.3% vs. 1.3%, P < 0.001). The NHD cohort had a more complex repair with a greater proportion of cases requiring high operative time (75.6% vs. 51.3%, P < 0.001).
Predictors of discharge destination after EVAR 3
Furthermore, NHD cases had significantly more bilateral groin cutdowns (63.6% vs. 55.1%, P ¼ 0.004), access vessel repair (8.9% vs. 5.4%, P ¼ 0.011), lower extremity revascularization (6.6% vs. 3.3%, P ¼ 0.003), and hypogastric embolization (10.3% vs. 6.4%, 0.008). Overall 30-Day Complications At 30 days, the NHD cohort had more major complications (37.8% vs. 9.4%, P < 0.001) but had similar rates of minor complications (3.8% vs. 2.3%, P ¼ 0.119). Specific complications are presented in Table II. Notably, for NHD versus home discharge, death (1.4% vs. 0.3%, P ¼ 0.011), unplanned return to the operating room (13.4% vs. 2.9%, P < 0.001), and postoperative lower extremity revascularization (4.1% vs. 1.0%, P < 0.001) were more frequent in NHD. However, unplanned readmission did not significantly differ between the 2 groups (9.3% vs. 6.6%, P ¼ 0.070). Predischarge Complications During the index hospitalization, all complications assessed on unadjusted analyses were more frequent in the NHD cohort except for venous thromboembolism and rupture, Table II. These included more major complications (32.7% vs. 6.7%, P < 0.001), minor complications (2.8% vs. 0.4%, P < 0.001), wound complications (1.0% vs. 0.1%, P ¼ 0.001), pulmonary complications (5.8% vs. 0.5%, P < 0.001), cardiac complications (5.2% vs. 0.4%, P < 0.001), and renal complications (3.8% vs. 0.1%, P < 0.001). The NHD cohort suffered more strokes (1.7% vs. 0.1%, P < 0.001), were more likely to require blood transfusion (22.3% vs. 5.5%, P < 0.001), and develop a sepsis/ shock (2.4% vs. 0.1%, P < 0.001) before discharge. Urinary tract infections were more common in NHD before discharge (2.1% vs. 0.3%, P < 0.001) as was ischemic colitis (2.1% vs. 0.2%, P < 0.001). Patients requiring NHD more likely had an unplanned return to the operating room (0.7% vs. 0.1%, P < 0.001), required lower extremity revascularization postoperatively (3.1% vs. 0.6%, P < 0.001), and had a longer overall (5 ± 2 vs. 1 ± 1 days, P < 0.001) and intensive care unit (ICU) (0 ± 2 vs. 0 ± 1 days, P < 0.001) length of stay. Postdischarge Complications In contrast, postdischarge complications were overall similar between the 2 groups, Table II. Major complications were more frequent in the NHD cohort (5.2% vs. 3.0%, P ¼ 0.037); however, minor complications did
4 Boitano et al.
Annals of Vascular Surgery
Table I. Baseline demographics and comorbidities Covariates a
Age (years) Female White race ASA class 3 or higher BMI <18.5 >10% weight lossb Albumin <3.0 Dependent functional status Current steroid use Current smoker COPD HTN CHF exacerbationc Diabetes mellitus ESRD on dialysis Bleeding disorder Anemia (hct<36%) Disseminated cancer Open wound Aneurysm diameter (cm)a High operative time Bilateral open femoral access Access vesselsdconduit, repair Lower extremity revascularization Hypogastric revascularization Hypogastric embolization Convert to open procedure
Total (n ¼ 6,476)
Discharge home (n ¼ 6,185)
NHD (n ¼ 291)
P
73.8 ± 8.7 1,137 (17.6%) 5376 (92.8%) 6,064 (93.6%) 91 (1.4%) 42 (0.7%) 3801 (58.7%) 106 (1.6%) 277 (4.3%) 1,962 (30.3%) 1,124 (17.4%) 5187 (80.1%) 67 (1.0%) 1,048 (16.2%) 48 (0.7%) 732 (11.3%) 832 (12.9%) 31 (0.5%) 50 (0.8%) 5.6 ± 1.4 3,393 (53.4%) 3,574 (55.5%) 361 (5.8%) 207 (3.4%) 289 (4.5%) 424 (6.6%) 31 (0.5%)
73.4 ± 8.6 1,032 (16.7%) 5124 (92.9%) 5784 (93.5%) 82 (1.3%) 40 (0.7%) 3640 (58.9%) 79 (1.3%) 262 (4.2%) 1,890 (30.6%) 1,060 (17.1%) 4949 (80.0%) 59 (1.0%) 995 (16.1%) 48 (0.8%) 692 (11.2%) 755 (12.2%) 31 (0.5%) 44 (0.7%) 5.6 ± 1.4 3,173 (51.3%) 3,389 (55.1%) 335 (5.4%) 189 (3.3%) 273 (4.4%) 394 (6.4%) 30 (0.5%)
80.0 ± 7.7 105 (36.1%) 252 (91.0%) 280 (96.2%) 9 (3.1%) 3 (1.0%) 162 (55.6%) 27 (9.3%) 15 (5.2%) 72 (24.7%) 64 (22.0%) 238 (81.8%) 8 (2.8%) 53 (18.2%) 0 (0%) 40 (13.8%) 77 (26.5%) 0 (0%) 6 (2.1%) 5.8 ± 1.2 220 (75.6%) 185 (63.6%) 26 (8.9%) 18 (6.6%) 16 (5.5%) 30 (10.3%) 2 (0.7%)
<0.001 <0.001 0.223 0.065 0.020 0.442 0.281 <0.001 0.449 0.037 0.039 0.460 0.010 0.336 0.277 0.184 <0.001 0.401 0.024 0.015 <0.001 0.004 0.011 0.003 0.382 0.008 0.654
ASA, American Society of Anesthesiologists; BMI, body mass index; COPD, chronic obstructive pulmonary disease; HTN, hypertension; CHF, congestive heart failure; ESRD, end-stage renal disease; Hct, hematocrit. a Mean ± standard deviation. b Within last 6 months. c Within last 30 days.
not differ significantly (1.0% vs. 2.0%, P ¼ 0.380). Pulmonary complications (1.4% vs. 0.2%, P ¼ 0.002) and death (1.4% vs. 0.3%, P ¼ 0.011) were the only postdischarge complication seen more frequently in the NHD cohort. Multivariable Model for NHD Multivariable binary logistic regression demonstrated the independent predictors for discharge to NHD, Table III. Octogenarians(oddsratio[OR]: 5.7, 95%confidenceinterval [CI], 2.3e14.1]; P < 0.001) and nonagenarians (OR: 14.6, 95% CI: 5.4e39.2; P < 0.001) and dependent functional status (OR: 5.4, 95% CI: 3.3e8.8; P < 0.001) were the strongest predictors of NHD. Other independent predictors of NHD identified in the model included female sex (OR: 2.2, 95% CI: 1.7e2.8, P < 0.001), open wound at baseline (OR: 3.5, 95% CI: 1.4e8.9; P ¼ 0.006), high operative time (OR: 2.7, 95% CI: 2.0e3.6; P < 0.001), and hypogastric embolization (OR: 1.6, 95% CI: 1.1e2.1 P ¼ 0.022). The final model had a C-statistic of 0.780. We included only preoperative covariates in the final model; however,
on sensitivity analysis using inpatient covariates, major predischarge complications (OR: 4.1, 95% CI: 3.1e5.6; P < 0.001) and ICU stay (OR: 1.7, 95% CI: 1.3e2.2; P < 0.001) were independently associated with NHD. On multivariable analysis, NHD did not independently predict unplanned readmission (OR 1.0, 95% CI: 0.6e1.5, P ¼ 0.842) after adjusting for female sex (OR 1.6, 95% CI: 1.3e2.0; P < 0.001), anemia (OR: 1.5, 95% CI: 1.1e1.9; P ¼ 0.003), current steroid use (OR: 2.0, 95% CI: 1.4e3.0; P < 0.001), hypertension (OR: 1.4, 95% CI: 1.0e1.8; P ¼ 0.030), high operative time (OR: 1.4, 95% CI: 1.1e1.7; P ¼ 0.002), bilateral groin cutdown (OR: 1.2, 95% CI: 1.0e 1.5; P ¼ 0.049), admission to the ICU (OR: 1.5, 95% CI: 1.2e1.8; P < 0.001), and major inpatient complications (OR: 1.4, 95% CI: 1.0e1.9, P ¼ 0.027). Similarly, NHD did not independently predict major postdischarge complication (OR: 1.0 95% CI: 0.6e1.9; P ¼ 0.875) after adjusting for preoperative anemia (OR: 1.6, 95% CI: 1.1e2.4; P ¼ 0.008), preoperative open wound (OR: 3.5,
Volume
-, -
2019
Predictors of discharge destination after EVAR 5
Table II. Thirty-day, predischarge, and postdischarge outcomes Outcome
Wound complications 30 day Predischarge Postdischarge Pulmonary complications 30 day Predischarge Postdischarge Cardiac complications 30 day Predischarge Postdischarge Stroke 30 day Predischarge Postdischarge Renal complications 30 day Predischarge Postdischarge Urinary tract infection 30 day Predischarge Postdischarge Venous thromboembolism 30 day Predischarge Postdischarge Bleeding requiring transfusion 30 day Predischarge Postdischarge Sepsis/septic shock 30 day Predischarge Postdischarge Ischemic colitis 30 day Predischarge Postdischarge Postoperative lower extremity revascularization 30 day Predischarge Postdischarge Unplanned reoperationa 30 day Predischarge Postdischarge Rupture 30 day Predischarge Postdischarge
Total (n ¼ 6,476)
Discharge home (n ¼ 6,185)
NHD (n ¼ 291)
P
84 (1.3%) 7 (0.1%) 77 (1.2%)
79 (1.3%) 4 (0.1%) 75 (1.2%)
5 (1.7%) 3 (1.0%) 2 (0.7%)
0.428 0.001 0.584
56 (0.9%) 45 (0.7%) 13 (0.2%)
37 (0.6%) 28 (0.5%) 9 (0.2%)
19 (6.5%) 17 (5.8%) 4 (1.4%)
<0.001 <0.001 0.002
57 (0.9%) 39 (6.0%) 18 (0.3%)
41 (0.7%) 24 (0.4%) 17 (0.3%)
16 (5.5%) 15 (5.2%) 1 (0.3%)
<0.001 <0.001 0.563
15 (0.2%) 7 (0.1%) 8 (0.1%)
10 (0.2%) 2 (0.1%) 8 (0.1%)
5 (1.7%) 5 (1.7%) 0 (0%)
<0.001 <0.001 >0.999
29 (0.5%) 19 (0.3%) 10 (0.2%)
17 (0.5%) 8 (0.1%) 9 (0.2%)
12 (4%) 11 (3.8%) 1 (0.3%)
<0.001 <0.001 0.369
79 (1.2%) 24 (0.4%) 55 (0.9%)
72 (1.2%) 18 (0.3%) 54 (0.9%)
7 (2.4%) 6 (2.1%) 1 (0.3%)
0.089 <0.001 0.518
28 (0.4%) 6 (0.1%) 22 (0.3%)
24 (0.4%) 5 (0.1%) 19 (0.3%)
4 (1.4%) 1 (0.3%) 3 (1.0%)
0.035 0.241 0.074
405 (6.3%) 405 (6.3%) 0 (0%)
340 (5.5%) 340 (5.5%) 0 (0%)
65 (22.3%) 65 (22.3%) 0 (0%)
<0.001 <0.001 >0.999
43 (0.7%) 13 (0.2 %) 30 (0.5%)
35 (0.6%) 6 (0.1%) 29 (0.5%)
8 (2.8%) 7 (2.4%) 1 (0.3%)
0.001 <0.001 >0.999
24 (0.4%) 17 (0.3%) 7 (0.1%)
17 (0.3%) 10 (0.2%) 6 (0.1%)
7 (2.4%) 6 (2.1%) 1 (0.3%)
<0.001 <0.001 0.275
76 (1.2%) 44 (0.7%) 30 (0.5%)
64 (1.0%) 35 (0.6%) 28 (0.5%)
12 (4.1%) 9 (3.1%) 2 (0.7%)
<0.001 <0.001 0.393
219 (3.8%) 7 (0.1%) 123 (1.9%)
180 (2.9%) 5 (0.1%) 115 (1.9%)
39 (13.4%) 2 (0.7%) 8 (2.8%)
<0.001 0.036 0.268
2 (0.1%) 1 (0.1%) 1 (0.1%)
1 (0.1%) 1 (0.1%) 1 (0.1%)
0 (0%) 0 (0%) 0 (0%)
>0.999 >0.999 >0.999 (Continued)
6 Boitano et al.
Annals of Vascular Surgery
Table II. Continued Outcome
Total (n ¼ 6,476)
Discharge home (n ¼ 6,185)
NHD (n ¼ 291)
P
ICU length of stay (days)b Length of stay (days)b Unplanned readmission Death Minor complications 30 day Predischarge Postdischarge Major complications 30 day Predischarge Postdischarge
0±1 1±1 433 (6.7%) 20 (0.3%)
0±1 1±1 406 (6.6%) 15 (0.3%)
0±1 5±2 27 (9.3%) 5 (1.4%)
<0.001 <0.001 0.070 0.011
156 (2.4%) 30 (0.5%) 126 (2.0%)
145 (2.3%) 22 (0.4%) 123 (2.0%)
11 (3.8%) 8 (2.8%) 3 (1.0%)
0.119 <0.001 0.380
694 (10.7%) 509 (7.9%) 200 (3.1%)
584 (9.4%) 414 (6.7%) 185 (3.0%)
110 (37.8%) 95 (32.7%) 15 (5.2%)
<0.001 <0.001 0.037
ICU, intensive care unit. a Missing dates are only included in overall 30-day reoperation. b Median ± interquartile range.
Table III. Logistic regression for nonhome discharge Covariate
Odds ratio
95% confidence interval
P value
Female Age <60 Age 60e69 Age 70e79 Age 80e89 Age 90+ Dependent functional status Open wound High operative time Hypogastric embolization C-statistic
2.2 Reference 1.2 1.8 5.7 14.6 5.4 3.5 2.7 1.6
1.7e2.9
<0.001
0.4e3.1 0.7e4.6 2.3e14.1 5.4e39.2 3.3e8.8 1.4e8.9 2.0e3.6 1.1e2.1
0.754 0.193 <0.001 <0.001 <0.001 0.006 <0.001 0.022 0.780
95% CI: 1.3e9.2; P ¼ 0.012), index procedural lower extremity revascularization (OR: 2.0, 95% CI: 1.1e3.5; P ¼ 0.015), admission to the ICU (OR: 1.5, 95% CI: 1.1e2.0; P ¼ 0.013), and major inpatient complications (OR: 2.3, 95% CI: 1.5e3.4; P < 0.001).
DISCUSSION This is one of the first studies to evaluate preoperative predictors of discharge to rehab after EVAR with procedure-specific data.10,11,13 The present study found 4.6% of previously home-dwelling patients require NHD following EVAR, and preoperative covariates predicted NHD with good discriminatory power. Independent predictors included female sex, age 80 years, dependent functional status, baseline open wound, high operative time, and hypogastric embolization. Age was the strongest predictor of discharge destination, and dependent
functional status was an important predictor with 5 times the odds of discharge to NHD compared with independent functioning patients. Not surprisingly, NHD patients experienced significantly more predischarge complications, including major complications, and were more likely to be admitted to the ICU. Although it has previously been shown that discharge to NHD is associated with worse postdischarge outcomes, NHD was not an independent predictor of readmission or postdischarge 30-day complications in the present study.11,19e23 In the present study, advanced age was the strongest predictor of discharge destination. Nine percent of octogenarians and 19.6% of nonagenarians required NHD as compared with 1.5% of patients younger than 60 years. Age is a well-known risk factor for worse outcome following EVAR.4,8,19,24,25 Gupta et al.24 found increasing age was associated with postdischarge 30-day morbidity and mortality in patients undergoing
Volume
-, -
2019
EVAR. Eslami et al.25 evaluated all patients undergoing elective AAA repair, both OAR and EVAR, and again older age (>70 years) was an independent predictor of 30-day mortality. In addition, our findings are consistent with other reports that advanced age is associated with NHD following vascular operations. In a recent study by Schermerhorn et al. evaluating the Medicare population, 15% of patients 85 years required NHD as compared with 2% of patients aged 67e69 years.4 Furthermore, in a study evaluating predictors of NHD following elective vascular surgery, patients 85 years had a 5-fold increased odds of NHD as compared with patients aged 65 years and younger.11 Despite the known importance of age on outcomes, there are some experts who suggest, regardless of chronological age, frailty is the most important determinant of outcome.26e28 There is currently no gold standard for measuring frailty, but regardless of how it is measured, studies have described worse outcomes in frail patients.11,27e30 Most frailty indices include variables evaluating preoperative nutritional status, weight loss, functional status, and baseline comorbid conditions.11,28e30 In the present study, dependent functional status was an important predictor of NHD, demonstrating 5 times the odds of NHD compared with patients independent at baseline. However, after risk adjustment, other indictors of frailty were not independent predictors of NHD. Gender differences in outcomes of aneurysm repair been reported previously, and female sex has been associated with poor outcomes after both EVAR and OAR.11,24,31e37 Consistent with other reports, this study identified female sex as an independent predictor of NHD.10,11 Several studies have postulated explanations for such differences and suggest that women have more complex anatomy, smaller vessel diameter, and untreated comorbid conditions.10,11,36,37 In addition, studies suggest that women are often in the caregiver role and might have a sicker family that is unable to care for them. Several reports have stated that advanced age at the time of treatment also contributes to worse outcomes and discharge to a skilled facility as they are likely to have outlived their spouse or caregiver.24,31e35 In accordance with this, female patients in this study were on average older by 3 years as compared with men (76 ± 8.7 vs. 73 ± 8.5 years, P < 0.001). Without socioeconomic factors, there is limited understanding of how social support impacted discharge status. Even with the advent of iliac branched devices and revascularization options to address perfusion
Predictors of discharge destination after EVAR 7
to the internal iliac artery, endovascular treatment of internal iliac artery aneurysms is most commonly performed by coil embolization with extension of the graft limb into the external iliac artery.32 A recent study evaluating concomitant hypogastric artery embolization during EVAR found that overall patients had increased operative time, more likely to require blood transfusion, and had longer ICU and hospital stay as well as a higher likelihood of discharge to NHD (12% vs. 7.8%, P ¼ 0.008).32 Our findings are consistent with this study, showing hypogastric embolization was also an independent predictor of discharge to NHD and was associated with a 60% increased risk of NHD (OR: 1.6, CI: 1.1e2.1, P ¼ 0.022). The inclusion of hypogastric coiling demonstrates the importance of procedurespecific variables and evaluating each procedure individually. High operative time was an independent predictor of discharge to NHD (OR: 2.7, CI: 2.0e3.6, P < 0.001) and was associated with bilateral groin cutdown on bivariate analysis (59.9% vs. 43.0%, P < 0.001), attempted percutaneous access but converted to groin cutdown (52.0% vs. 84.3%, P < 0.001), iliac stent placement (61.9% vs. 51.6%, P < 0.001), hypogastric revascularization (68.5% vs. 51.6%, P < 0.001), hypogastric embolization (78.1 vs. 50.6, P < 0.001), and lower extremity revascularization (88.4% vs. 50.0%, P < 0.001). In a recent study evaluating outcomes after EVAR with or without hypogastric coiling, longer-than-average operative repair was associated with increased risk of 30-day mortality, ischemic colitis, and renal failure requiring dialysis.32 In addition, Arya et al.11 found for every increase in an hour of operative time, there was a 26% increased odds of NHD in patients undergoing nonspecific vascular procedures. This increase in operative time could have been more attributed to differences in the vascular procedures evaluated. The present study demonstrates that within the same procedure type, EVAR, this is an important variable to evaluate. Interestingly, open wound at baseline was associated with 3 times the odds of discharge to NHD. Crouch et al. evaluated risk factors associated with decline in independent living status after vascular surgery and found open wound at discharge, primarily related to the surgical procedure, independently predicted such decline.13 However, this study uniquely identifies the importance of preoperative wounds on need for NHD after EVAR. Finally, this study did not find an association between NHD and readmission or postdischarge complication. This finding is in contrast to other reports, across surgical subspecialties, which have
8 Boitano et al.
found that NHD is associated with increased morbidity, mortality, and readmission.11,19e23 Arya et al.11 reported a 6-fold increase in postdischarge mortality and 2-fold increase in readmission rate in vascular surgery patients requiring NHD. However, this study failed to perform an adjusted analysis to evaluate this association, which may contribute to the difference in our findings. Greenblatt et al. found that after EVAR and OAR, discharge to destination other than home predicted readmission, and it was the most powerful predictor of early readmission.19 The difference in our findings with those of Greenblatt et al. may be related to the inclusion of OAR in their study. The present study did find an association on unadjusted analysis between NHD and 30-day death; multivariable analysis could not be conducted to determine whether this association held after risk adjustment due to small outcome numbers. In light of trial data from the EVAR 2 study, careful consideration and patient counseling based on individual risk factors are warranted as EVAR may not provide adequate risk reduction in high-risk populations, including those at high risk for NHD.38
Study Limitations This study is limited in that it utilizes a large registry such that the analysis is limited to the variables which are captured by the NSQIP Vascular Targeted EVAR database. There is limited information regarding socioeconomic status and social support, which may be important in the prediction of discharge destination. Furthermore, there were a number of variables with a large amount of missingness and thus could not be included as potential predictors of NHD. These included impaired sensorium, preoperative myocardial infarction, and history of hemiplegia or paraplegia. In addition, the NSQIP database does not include hospital identification such that a mixed-effects model could not be constructed to account for individual hospital practices. Finally, the analysis is limited to 30 days so this study was unable to evaluate whether NHD has implications on long-term survival or functional status. This study identifies independent risk factors for NHD. However, these predictors have not been prospectively validated. Thus, future studies should focus on validation of these findings. Furthermore, future studies should seek to determine a risk prediction score such that patients can be counseled preoperatively on the risk for NHD. In addition, studies should look at the cost savings and hospital
Annals of Vascular Surgery
length of stay associated with early identification of patients who are likely to be discharged to NHD.
CONCLUSION This is one of the first studies evaluating preoperative predictors of NHD following elective EVAR. In this study, using 6 preoperative variables on multivariable analysis, we were able to have high predictability of discharge destination. Age greater than 80 years and dependent functional status were the most important predictors of discharge destination. Future studies should seek to validate these findings in a prospective manner. Patients discharged to rehab did not experience increased postdischarge perioperative major complications and are not at higher risk for readmission following EVAR. These data are important for understanding the risk of NHD following elective EVAR and can help to better counsel patients with such characteristics who are at higher risk for NHD. REFERENCES 1. Levin DC, Rao VM, Parker L, et al. Endovascular repair vs open surgical repair of abdominal aortic aneurysms: comparative utilization trends from 2001 to 2006. J Am Coll Radiol 2009;6:506e9. 2. Schwarze ML, Shen Y, Hemmerich J, et al. Age-related trends in utilization and outcome of open and endovascular repair for abdominal aortic aneurysm in the United States, 2001-2006. J Vasc Surg 2009;50:722e729.e2. 3. Brewster DC, Jones JE, Chung TK, et al. Long-term outcomes after endovascular abdominal aortic aneurysm repair: the first decade. Ann Surg 2006;244:426e38. 4. Schermerhorn ML, Buck DB, O’Malley AJ, et al. Long-term outcomes of abdominal aortic aneurysm in the Medicare population. N Engl J Med 2015;373:328e38. 5. Prinssen M, Verhoeven EL, Buth J, et al. A randomized trial comparing conventional and endovascular repair of abdominal aortic aneurysms. N Engl J Med 2004;351:1607e18. 6. Participants Et.. Endovascular aneurysm repair versus open repair in patients with abdominal aortic aneurysm (EVAR trial 1): randomised controlled trial. Lancet 2005;365: 2179e86. 7. Lederle FA, Freischlag JA, Kyriakides TC, et al. Outcomes following endovascular vs open repair of abdominal aortic aneurysm: a randomized trial. JAMA 2009;302:1535e42. 8. Lederle FA, Freischlag JA, Kyriakides TC, et al. Long-term comparison of endovascular and open repair of abdominal aortic aneurysm. N Engl J Med 2012;367:1988e97. 9. Faggioli G, Scalone L, Mantovani LG, et al., Group Ps. Preferences of patients, their family caregivers and vascular surgeons in the choice of abdominal aortic aneurysms treatment options: the PREFER study. Eur J Vasc Endovasc Surg 2011;42:26e34. 10. Vogel TR, Nackman GB, Crowley JG, et al. Factors impacting functional health and resource utilization following abdominal aortic aneurysm repair by open and endovascular techniques. Ann Vasc Surg 2005;19:641e7.
Volume
-, -
2019
11. Arya S, Long CA, Brahmbhatt R, et al. Preoperative frailty increases risk of Nonhome discharge after elective vascular surgery in home-dwelling patients. Ann Vasc Surg 2016;35:19e29. 12. Aljabri B, Al Wahaibi K, Abner D, et al. Patient-reported quality of life after abdominal aortic aneurysm surgery: a prospective comparison of endovascular and open repair. J Vasc Surg 2006;44:1182e7. 13. Crouch DS, McLafferty RB, Karch LA, et al. A prospective study of discharge disposition after vascular surgery. J Vasc Surg 2001;34:62e8. 14. User Guide for the 2015 ACS NSQIP Procedure Targeted Participant Use Data Gile (PUF). Available at: https://www. facs.org/w/media/files/quality%20programs/nsqip/pt._nsqip_ puf_user_guide_2015.ashx. Accessed May 1, 2018. 15. Iannuzzi JC, Chandra A, Kelly KN, et al. Risk score for unplanned vascular readmissions. J Vasc Surg 2014;59:1340e1347.e1. 16. Shields E, Iannuzzi JC, Thorsness R, et al. Postoperative morbidity by procedure and patient factors Influencing major complications within 30 Days following shoulder surgery. Orthop J Sports Med 2014;2:2325967114553164. 17. Iannuzzi JC, Chandra A, Rickles AS, et al. Resident involvement is associated with worse outcomes after major lower extremity amputation. J Vasc Surg 2013;58:827e831.e1. 18. Iannuzzi JC, Rickles AS, Deeb AP, et al. Outcomes associated with resident involvement in partial colectomy. Dis Colon Rectum 2013;56:212e8. 19. Greenblatt DY, Greenberg CC, Kind AJ, et al. Causes and implications of readmission after abdominal aortic aneurysm repair. Ann Surg 2012;256:595e605. 20. Williamson WK, Nicoloff AD, Taylor LM Jr, et al. Functional outcome after open repair of abdominal aortic aneurysm. J Vasc Surg 2001;33:913e20. 21. Legner VJ, Massarweh NN, Symons RG, et al. The significance of discharge to skilled care after abdominopelvic surgery in older adults. Ann Surg 2009;249:250e5. 22. Ottenbacher KJ, Karmarkar A, Graham JE, et al. Thirty-day hospital readmission following discharge from postacute rehabilitation in fee-for-service Medicare patients. JAMA 2014;311:604e14. 23. Palmieri TL, Molitor F, Chan G, et al. Long-term functional outcomes in the elderly after burn injury. J Burn Care Res 2012;33:497e503. 24. Gupta PK, Engelbert TL, Ramanan B, et al. Postdischarge outcomes after endovascular abdominal aortic aneurysm repair. J Vasc Surg 2014;59:903e8. 25. Eslami MH, Rybin DV, Doros G, et al. Description of a risk predictive model of 30-day postoperative mortality after elective abdominal aortic aneurysm repair. J Vasc Surg 2017;65:65e74.e2.
Predictors of discharge destination after EVAR 9
26. Rockwood K, Howlett SE, MacKnight C, et al. Prevalence, attributes, and outcomes of fitness and frailty in community-dwelling older adults: report from the Canadian study of health and aging. J Gerontol A Biol Sci Med Sci 2004;59:1310e7. 27. Lin HS, Watts JN, Peel NM, et al. Frailty and post-operative outcomes in older surgical patients: a systematic review. BMC Geriatr 2016;16:157. 28. Makary MA, Segev DL, Pronovost PJ, et al. Frailty as a predictor of surgical outcomes in older patients. J Am Coll Surg 2010;210:901e8. 29. Ambler GK, Brooks DE, Al Zuhir N, et al. Effect of frailty on short- and mid-term outcomes in vascular surgical patients. Br J Surg 2015;102:638e45. 30. Arya S, Kim SI, Duwayri Y, et al. Frailty increases the risk of 30-day mortality, morbidity, and failure to rescue after elective abdominal aortic aneurysm repair independent of age and comorbidities. J Vasc Surg 2015;61:324e31. 31. Zettervall SL, Schermerhorn ML, Soden PA, et al. The effect of surgeon and hospital volume on mortality after open and endovascular repair of abdominal aortic aneurysms. J Vasc Surg 2017;65:626e34. 32. Farivar BS, Kalsi R, Drucker CB, et al. Implications of concomitant hypogastric artery embolization with endovascular repair of infrarenal abdominal aortic aneurysms. J Vasc Surg 2017;66:95e101. 33. Dillavou ED, Muluk SC, Makaroun MS. A decade of change in abdominal aortic aneurysm repair in the United States: have we improved outcomes equally between men and women? J Vasc Surg 2006;43:230e8. discussion 8. 34. Egorova NN, Vouyouka AG, McKinsey JF, et al. Effect of gender on long-term survival after abdominal aortic aneurysm repair based on results from the Medicare national database. J Vasc Surg 2011;54:1e12.e6. discussion 1e2. 35. Lo RC, Bensley RP, Hamdan AD, et al. Gender differences in abdominal aortic aneurysm presentation, repair, and mortality in the Vascular Study Group of New England. J Vasc Surg 2013;57:1261e8.e1e5. 36. Deery SE, Soden PA, Zettervall SL, et al. Sex differences in mortality and morbidity following repair of intact abdominal aortic aneurysms. J Vasc Surg 2017;65:1006e13. 37. Deery SE, Shean KE, Wang GJ, et al. Female sex independently predicts mortality after thoracic endovascular aortic repair for intact descending thoracic aortic aneurysms. J Vasc Surg 2017;66:2e8. 38. participants Et.. Endovascular aneurysm repair and outcome in patients unfit for open repair of abdominal aortic aneurysm (EVAR trial 2): randomised controlled trial. Lancet 2005;365:2187e92.