The first documented shoulder arthroplasty dates back to 1893,1,2 with major surgical and technological improvements having occurred since then. These advances have increased during the past several decades with the advent of modern anatomic implants such as the hemiarthroplasty and the total shoulder arthroplasty (TSA). In addition, in November 2003, the Food and Drug Administration approved a nonanatomic reverse total shoulder arthroplasty (RTSA) for use.1 Among the expanding list of indications for shoulder arthroplasty, primary glenohumeral osteoarthritis and cuff tear arthropathy are expected to increase in prevalence in the older adult population. In the United States, the elderly population is expected to grow due to the aging of the “baby boomer” generation.3 More specifically, the US Census Bureau anticipates that the population 65 years and older will more than double to 88.5 million by 2050.3 The number of total joint arthroplasties performed each year in the United States has been increasing during the past several decades and is projected to continue to increase based on the growing and aging population.4 These increases are expected to place clinical and financial burdens on the health care system.
Given the increasing volume of TSAs performed in the United States, it is important to employ postoperative discharge practices that can both maintain acceptable clinical outcomes and minimize the financial burden on the health care system. Although there are undoubtedly patients who benefit from postdischarge inpatient services to improve functional outcomes, many patients may not require this level of care for clinical optimization. Hence, it is reasonable to investigate the implications of discharge to inpatient facilities vs home regarding short-term complications. The purpose of this investigation was to use the National Surgical Quality Improvement Program (NSQIP) database to analyze the rates of morbidity following shoulder arthroplasty. The authors hypothesized that, independent of predischarge patient factors, discharge to inpatient facilities is independently associated with increased short-term morbidity.
Materials and Methods
This retrospective cohort study used the NSQIP database. This national registry was established in 2005. Additional data have been added to it annually, with data available for more than 4 million unique surgical encounters as of 2015.5 Data collected include patient demographics (eg, age, sex), intraoperative variables (eg, operative time, blood transfusions), and information about postoperative complications and readmissions up to 30 days after the operative event. These data are prospectively entered into the registry by specially trained surgical clinical reviewers. The NSQIP database is audited by the American College of Surgeons to ensure accuracy and reliability of the data. As a result, there is extremely high interrater reliability, with a disagreement rate of only 1.56% as of 2015.6
Baseline Patient Characteristics and Perioperative Complications
This study used NSQIP data from 2005 to 2015. Primary anatomic and RTSA patients were selected by identifying cases from the registry with primary Current Procedural Terminology code 23472. Patients were then divided into 2 cohorts based on discharge destination: home vs non-home. Non-home discharge destinations included “rehab,” “separate acute care,” “skilled care, not home,” and “unskilled facility not home.” Patients with dirty wound classifications, surgical procedures that were considered emergent, missing discharge destination, or missing baseline patient and operative characteristics were excluded from this study. Baseline patient and operative characteristics that were assessed in this study included age, sex, body mass index, American Society of Anesthesiologists (ASA) class, functional status, and comorbidities, including hypertension, diabetes mellitus, chronic obstructive pulmonary disease, and smoking history.
Clinical outcomes that were assessed included 30-day readmissions and medical complications, including cardiac complications (myocardial infarction or cardiac arrest), respiratory complications (pneumonia, failure to wean from the ventilator, unplanned reintubation), deep venous thrombosis or pulmonary embolism, urinary tract infection, septic complications (septic shock or sepsis), wound complications (superficial, deep, or organ space surgical site infection or wound dehiscence), and death.
Unadjusted baseline patient characteristics were compared using Pearson's chi-square test, and a propensity score–adjusted comparison was also performed. Propensity scoring is a statistical analysis used in observational data that attempts to estimate the effect of an exposure while accounting for covariates that may predict exposure. For this investigation, propensity scoring was used to estimate the effect of discharge to non-home facilities on patient complications while accounting for covariates such as age and comorbidity status, which may influence the rates of discharge to non-home facilities. Multiple methods for using the propensity score, including matching, stratification, inverse probability, and covariate adjustment, have been described.7
In this study, propensity scores were calculated for each patient based on the conditional probability of discharge to a skilled care facility, given each patient's age, sex, body mass index, medical comorbidities, functional status, and ASA class and the presence of any predischarge complications. Covariate adjustment employing the propensity score was used to reduce the effects of confounding in the observational study design. This is a useful statistical technique to emulate the process of randomization that is often used in prospective trials within the orthopedic literature.8–11
All unadjusted comparisons of 30-day outcomes were performed using Pearson's chi-square test, and propensity score–adjusted comparisons were performed using bivariate regressions that included each patient's propensity score as a covariate. Complications that were statistically significant in the unadjusted model were carried forward to the propensity score–adjusted comparison. Bonferroni corrections were applied to account for multiple comparisons. Independent predictors of having a non-home discharge were identified using multivariate logistic regressions with non-home discharge as the outcome while adjusting for baseline patient characteristics. SPSS version 23 software (IBM Corporation, Armonk, New York) was used for all statistical analyses.
A total of 9058 patients were identified from the NSQIP database for the period 2005 to 2015 using the inclusion and exclusion criteria described above. Of the included patients, 56.2% were female, 47.3% were 71 years or older at the time of surgery, and 10.3% were classified as obese class III (morbidly obese, body mass index ≥40 kg/m2) (Table 1). The patient population in this study had the following baseline comorbidities: hypertension (66.9%), diabetes mellitus (17.1%), chronic obstructive pulmonary disease (6.5%), and smoking history (10.5%). The study cohort also included 2.7% of patients with dependent functional status, 53.3% of patients with ASA class III or higher, and 0.9% of patients with predischarge complications.
Baseline Characteristics of Patients Undergoing Total Shoulder Arthroplasty
Of the included patients, 7996 (88.3%) were discharged to home and 1062 (11.7%) were discharged to an inpatient facility (Table 1). Patients discharged to non-home facilities were more likely to be older at the time of the surgery (un-adjusted P<.001), be female (unadjusted P<.001), be obese (unadjusted P=.001), have comorbidities (hypertension unadjusted P<.001; diabetes mellitus unadjusted P<.001; chronic obstructive pulmonary disease unadjusted P<.001; or smoking history unadjusted P=.002), be functionally nonindependent (unadjusted P<.001), and have higher ASA class (unadjusted P<.001). After propensity score analysis, none of these values were statistically significant.
When evaluating predictors of discharge to non-home facilities using multivariate logistic regression, several predictors were identified as being statistically significantly associated with discharge to a non-home facility. These predictors included female sex (odds ratio [OR], 2.87; 95% confidence interval [CI], 2.44–3.38; P<.001), age 71 to 80 years at the time of surgery (OR, 3.36; 95% CI, 2.49–4.52; P<.001), age older than 80 years at the time of surgery (OR, 10.44; 95% CI, 7.61–14.31; P<.001), body mass index greater than or equal to 40 kg/m2 (OR, 1.54; 95% CI, 1.22–1.94; P<.001), diabetes mellitus (OR, 1.35; 95% CI, 1.13–1.60; P=.001), dependent functional status (OR, 3.75; 95% CI, 2.81–5.00; P<.001), and ASA class II (OR, 2.00; 95% CI, 1.70–2.35; P<.001) or class III/IV (OR, 3.33; 95% CI, 2.32–4.77; P<.001) (Table 2).
Multivariate Logistic Regression for Predictors of Being Discharged to a Non-Home Facility
Overall, 4.49% of patients had post-discharge complications (Table 3). Of these complications, the most common were readmission (2.85%), urinary tract infection (0.73%), respiratory complications (0.68%), and deep venous thrombosis/pulmonary embolism (0.63%). Using unadjusted comparison of postdischarge complications by discharge destination, the following complications were found to be statistically significantly associated with non-home facility discharge: cardiac, respiratory, deep venous thrombosis/pulmonary embolism, urinary tract infection, septic shock or sepsis, death, and readmission. The single complication group that was not found to be statistically significant was wound complications.
Unadjusted Comparison of Postdischarge Complications by Discharge Destination
On propensity-adjusted analysis of the postdischarge complications by discharge destination, complications determined to be statistically significantly associated with non-home discharge included cardiac (OR, 4.19; 95% CI, 1.75–10.04; P=.001), respiratory (OR, 2.63; 95% CI, 1.47–4.70; P=.001), urinary tract infection (OR, 2.66; 95% CI, 1.52–4.67; P=.001), and death (OR, 7.51; 95% CI, 2.42–23.27; P<.001) (Table 4). There was also increased risk associated with development of deep venous thrombosis/pulmonary embolism (OR, 1.91; 95% CI, 1.00–3.65; P=.049), septic shock or sepsis (OR, 2.66; 95% CI, 0.98–7.22; P=.054), and readmission (OR, 1.54; 95% CI, 1.11–2.14; P=.010). However, these risks did not reach statistical significance. Overall, complications were statistically significantly associated with non-home discharges compared with home discharges (OR, 2.05; 95% CI, 1.59–2.64; propensity-adjusted P<.001).
Propensity-Adjusted Comparison of Postdischarge Complications by Discharge Destination
The number of shoulder arthroplasties performed in the United States is increasing. When considering the increase in volume, it is critical to consider the financial implications of these procedures for the health care system. A significant portion of the financial cost of these procedures has been attributed to the immediate postoperative period, more specifically the postdischarge destination. Using the NSQIP database, the current study found an association between discharge to non-home facilities and increased rates of postoperative morbidity.
In the hip and knee arthroplasty literature, various studies have projected massively increased demands for primary and revision procedures through 2030.4 The need for optimization of economic resources, operative efficiency and capacity, and implant longevity to keep pace with these demands has been noted.4 Although studies on TSA lag behind the hip and knee literature, there are already several studies pointing to increasing demands for shoulder replacements.1,12 Concurrently, there has been a push from the more traditional fee-for-service reimbursement to alternative payment options such as a bundled payment system. In the fee-for-service plan, the health care provider is paid for the volume of services provided without incorporation of patient outcomes. In comparison, bundled payment systems have been established with the goal of improving health care value by correlating financial reimbursement with clinical outcome. Within the bundled payment time frame (typically 30 to 90 days), the posthospitalization costs of home vs inpatient facility discharge have important financial implications. In the hip and knee literature, a single study of 250 Medicare beneficiaries found that 36% of episode of care payments were attributed to post-acute care facilities, and additional studies have reported no association between these facilities and improved clinical outcomes.10,13–17 Furthermore, multiple studies have reported that discharge to these non-home facilities following hip and knee arthroplasty is associated with increased rates of morbidity.10,11
In 2017, Rosas et al18 analyzed the national discharge disposition costs for 50,979 patients during the first 90 days of care following TSA procedures. The mean cost was reported as $12,691±$471.9 for discharge home, $15,093±$622.4 for discharge to a skilled nursing facility, $15,473±$2366 for discharge to an intermediate care facility, and $24,731±$1922 for discharge to inpatient rehabilitation. More specifically, they reported that discharge to an inpatient rehabilitation center increased the 90-day cost by 194% as compared with discharge home. The ideal discharge management of these patients would optimize clinical outcomes while simultaneously minimizing the financial burden. As long as clinical and functional outcomes following discharge home vs in-patient rehabilitation are similar, it would be financially prudent to use the most cost-minimizing option when medically appropriate.
In this study, those who were ultimately discharged to non-home facilities were at greater odds of having several complications, including cardiac, respiratory, urinary tract infection, and death. Intuitively, it seems that the sicker patients would go to rehabilitation and therefore end up with more complications because they have more issues at baseline. However, the current analysis used propensity score adjustment to reduce or eliminate cofounding data and still found that patients who were discharged to non-home facilities instead of home had a much higher rate of complications in the early postoperative period. This is an important finding, as many patients believe spending time in a rehabilitation center is better than going directly home. The greatest preoperative predictors for discharge to non-home facilities included older age at the time of surgery, female sex, and obesity. In addition, patients with other medical comorbidities, dependent functional status, and higher ASA class stratifications were also found to have higher rates of discharge to non-home facilities. That risk factors such as age, sex, smoking status, and medical comorbidities are associated with postdischarge complications and re-admission has also been noted in several other studies.18–23 The results of this study parallel those of previous studies, as similar variables were predictors of discharge to places other than home.
Although there are certainly instances when discharge to non-home facilities is warranted for optimization of functional status, this investigation highlighted the importance of appropriate preoperative evaluation and management based on risk. The cost of discharge to non-home facilities is high, and this study evaluated the clinical consequences related to postoperative complications based on discharge destination. As the use of shoulder arthroplasty continues to grow, ongoing discussions have occurred regarding performing these procedures in ambulatory surgery centers with same-day discharge to home. The current study found specific patient factors that are predictors of discharge to rehabilitation following shoulder arthroplasty. Surgeons can use these factors as screening tools to select patients who may benefit from same-day discharge following shoulder arthroplasty. Furthermore, the results of this study can help generate preoperative expectations for postdischarge destination as well as aid discussions and interventions with multi-disciplinary teams to clinically and financially optimize the postdischarge course.
There were several limitations to this study. The first was human error associated with collecting and recording data in the system. However, the NSQIP maintains strict collection and reporting protocols that are performed by trained clinical reviewers. A 2010 audit reviewed the NSQIP data collection protocols and interrater reliability, revealing overall disagreement rates between data collector and auditor as low as 1.56% in 2008.6 The audit concluded that the program's training and audit procedures are highly effective and indicated that data have been reliable since the program's inception with annual improvement. An additional limitation was based on the large sample and the inclusion of various hospital systems across the country. Geographic and hospital-specific factors may impact clinical outcomes and rates of complications in ways that were not evaluated in this study. Although these hospital-specific factors may alter clinical outcomes, it is also important to consider patient-specific factors that were not evaluated in the current study. These factors include, but are not limited to, insurance status, mobility status preoperatively, socioeconomic status, and preoperative patient expectations, which may influence postdischarge destinations. As expected, patients discharged to non-home facilities were older and had more comorbidities at baseline, so one may expect their complication rate to be higher following surgery. However, on propensity adjustment, patients discharged to non-home facilities still had a higher rate of complications regardless of preoperative comorbidities and risk factors. Finally, the Current Procedural Terminology code used in this study (23472, defined as “total arthroplasty of glenohumeral joint with glenoid and proximal humerus replacement”) does not distinguish between anatomic total shoulder replacement and reverse total shoulder replacement. Reverse total shoulder arthroplasty is commonly used in broader clinical settings, including failed hemiarthroplasty or TSA and in patients with rotator cuff insufficiency, proximal humerus fractures, and rheumatoid arthritis.24 This is important to consider because there are data indicating that patients undergoing RTSA for these conditions are at increased risk for discharge to non-home facilities, experience higher rates of in-hospital death, have higher rates of perioperative complications, are older, and have higher rates of comorbidities as compared with those undergoing TSA procedures.24–28 Because this study did not evaluate the differences between patients undergoing TSA vs RTSA, the authors acknowledge that further investigations are necessary to evaluate risks specific to these procedures.
This study revealed an association between postdischarge placement in non-home facilities and an increase in short-term morbidity, regardless of preoperative patient comorbidities. These findings have both clinical and financial implications for the health care system as the aging baby boomer population is expected to have an increased demand for shoulder arthroplasty procedures.
- Kim SH, Wise BL, Zhang Y, Szabo RM. Increasing incidence of shoulder arthroplasty in the United States. J Bone Joint Surg Am. 2011;93(24):2249–2254. doi:10.2106/JBJS.J.01994 [CrossRef]
- Bankes MJ, Emery RJ. Pioneers of shoulder replacement: Themistocles Gluck and Jules Emile Pean. J Shoulder Elbow Surg. 1995;4(4):259–262. doi:10.1016/S1058-2746(05)80018-7 [CrossRef]
- Pallin DJ, Espinola JA, Camargo CA Jr, . US population aging and demand for inpatient services. J Hosp Med. 2014;9(3):193–196. doi:10.1002/jhm.2145 [CrossRef]
- Kurtz S, Ong K, Lau E, Mowat F, Halpern M. Projections of primary and revision hip and knee arthroplasty in the United States from 2005 to 2030. J Bone Joint Surg Am. 2007;89(4):780–785.
- American College of Surgeons. ACS National Surgical Quality Improvement Program (ACS NSQIP). https://www.facs.org/quality-programs/acs-nsqip. Accessed November 2017.
- Shiloach M, Frencher SK Jr, Steeger JE, et al. Toward robust information: data quality and inter-rater reliability in the American College of Surgeons National Surgical Quality Improvement Program. J Am Coll Surg. 2010;210(1):6–16. doi:10.1016/j.jamcollsurg.2009.09.031 [CrossRef]
- Austin PC. An introduction to propensity score methods for reducing the effects of confounding in observational studies. Multivariate Behav Res. 2011;46(3):399–424. doi:10.1080/00273171.2011.568786 [CrossRef]
- Gholson JJ, Shah AS, Gao Y, Noiseux NO. Morbid obesity and congestive heart failure increase operative time and room time in total hip arthroplasty. J Arthroplasty. 2016;31(4):771–775. doi:10.1016/j.arth.2015.10.032 [CrossRef]
- Duchman KR, Gao Y, Pugely AJ, Martin CT, Callaghan JJ. Differences in short-term complications between unicompartmental and total knee arthroplasty: a propensity score matched analysis. J Bone Joint Surg Am. 2014;96(16):1387–1394. doi:10.2106/JBJS.M.01048 [CrossRef]
- McLawhorn AS, Fu MC, Schairer WW, Sculco PK, MacLean CH, Padgett DE. Continued inpatient care after primary total knee arthroplasty increases 30-day post-discharge complications: a propensity score-adjusted analysis. J Arthroplasty. 2017;32(suppl 9):113S–118S. doi:10.1016/j.arth.2017.01.039 [CrossRef]
- Fu MC, Samuel AM, Sculco PK, MacLean CH, Padgett DE, McLawhorn AS. Discharge to inpatient facilities after total hip arthroplasty is associated with increased postdischarge morbidity. J Arthroplasty. 2017;32(9S):144S–149S. doi:10.1016/j.arth.2017.03.044 [CrossRef]
- Day JS, Lau E, Ong KL, Williams GR, Ramsey ML, Kurtz SM. Prevalence and projections of total shoulder and elbow arthroplasty in the United States to 2015. J Shoulder Elbow Surg. 2010;19(8):1115–1120. doi:10.1016/j.jse.2010.02.009 [CrossRef]
- 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(1):188–193. doi:10.1007/s11999-013-3034-3 [CrossRef]
- Chimenti CE, Ingersoll G. Comparison of home health care physical therapy outcomes following total knee replacement with and without subacute rehabilitation. J Geriatr Phys Ther. 2007;30(3):102–108. doi:10.1519/00139143-200712000-00004 [CrossRef]
- Kathrins B, Kathrins R, Marsico R, et al. Comparison of day rehabilitation to skilled nursing facility for the rehabilitation for total knee arthroplasty. Am J Phys Med Rehabil. 2013;92(1):61–67. doi:10.1097/PHM.0b013e3182643fd5 [CrossRef]
- Mallinson TR, Bateman J, Tseng HY, et al. A comparison of discharge functional status after rehabilitation in skilled nursing, home health, and medical rehabilitation settings for patients after lower-extremity joint replacement surgery. Arch Phys Med Rehabil. 2011;92(5):712–720. doi:10.1016/j.apmr.2010.12.007 [CrossRef]
- Mahomed NN, Davis AM, Hawker G, et al. Inpatient compared with home-based rehabilitation following primary unilateral total hip or knee replacement: a randomized controlled trial. J Bone Joint Surg Am. 2008;90(8):1673–1680. doi:10.2106/JBJS.G.01108 [CrossRef]
- Rosas S, Kurowicki J, Triplet JJ, Berglund DD, Horn B, Levy JC. Discharge disposition costs following total shoulder arthroplasty: a comprehensive national analysis. J Shoulder Elbow Surg. 2017;1:1–6.
- Minhas SV, Kester BS, Lovecchio FC, Bosco JA. Nationwide 30-day readmissions after elective orthopedic surgery: reasons and implications. J Healthc Qual. 2017;39(1):34–42. doi:10.1097/JHQ.0000000000000045 [CrossRef]
- Belmont PJ Jr, Kusnezov NA, Dunn JC, Bader JO, Kilcoyne K, Waterman BR. Predictors of hospital readmission after total shoulder arthroplasty. Orthopedics. 2017;40(1):E1–E10. doi:10.3928/01477447-20160915-06 [CrossRef]
- Weinreb JH, Cote MP, O'Sullivan MB, Mazzocca AD. Shoulder arthroplasty: disposition and perioperative outcomes in patients with and without rheumatoid arthritis. Am J Orthop. 2016;45(4):E204–E210.
- McCormick F, Nwachukwu BU, Kiriakopoulos EB, Schairer WW, Provencher MT, Levy J. In-hospital mortality risk for total shoulder arthroplasty: a comprehensive review of the Medicare database from 2005 to 2011. Int J Shoulder Surg. 2015;9(4):110–113. doi:10.4103/0973-6042.167938 [CrossRef]
- Xu S, Baker DK, Woods JC, Brabston EW III, Ponce BA. Risk factors for early readmission after anatomical or reverse total shoulder arthroplasty. Am J Orthop. 2016;45(6):E386–E392.
- Ponce BA, Oladeji LO, Rogers ME, Menendez ME. Comparative analysis of anatomic and reverse total shoulder arthroplasty: inhospital outcomes and costs. J Shoulder Elbow Surg. 2015;24(3):460–467. doi:10.1016/j.jse.2014.08.016 [CrossRef]
- Jones KJ, Dines DM, Gulotta L, Dines JS. Management of proximal humerus fractures utilizing reverse total shoulder arthroplasty. Curr Rev Musculoskelet Med. 2013;6(1):63–70. doi:10.1007/s12178-013-9155-1 [CrossRef]
- Ramirez MA, Ramirez J, Murthi AM. Reverse total shoulder arthroplasty for irreparable rotator cuff tears and cuff tear arthropathy. Clin Sports Med. 2012;31(4):749–759. doi:10.1016/j.csm.2012.07.009 [CrossRef]
- Sivasundaram L, Heckmann N, Pannell WC, Alluri RK, Omid R, Hatch GF III, . Preoperative risk factors for discharge to a postacute care facility after shoulder arthroplasty. J Shoulder Elbow Surg. 2016;25(2):201–206. doi:10.1016/j.jse.2015.07.028 [CrossRef]
- Schairer WW, Nwachukwu BU, Lyman S, Craig EV, Gulotta LV. Reverse shoulder arthroplasty versus hemiarthroplasty for treatment of proximal humerus fractures. J Shoulder Elbow Surg. 2015;24(10):1560–1566. doi:10.1016/j.jse.2015.03.018 [CrossRef]
Baseline Characteristics of Patients Undergoing Total Shoulder Arthroplastya
|Characteristic||All Patients (N=9058)||Unadjusted Cohort|
|Discharge to Home (n=7996)||Discharge to Inpatient Facility (n=1062)||Unadjusted P||Propensity-Adjusted P|
| ≤60 y||17.4%||18.9%||5.6%|
| 61 to 70 y||35.3%||37.4%||19.2%|
| 71 to 80 y||34.8%||34.1%||40.4%|
| >80 y||12.5%||9.6%||34.7%|
|Body mass index||.001b||.942|
| <30 kg/m2 (non-obese)||49.9%||50.0%||49.5%|
| 30–34.9 kg/m2 (obese class I)||26.4%||26.7%||24.2%|
| 35.0–39.9 kg/m2 (obese class II)||13.4%||13.5%||12.5%|
| >40 kg/m2 (obese class III)||10.3%||9.8%||13.7%|
| Diabetes mellitus||17.1%||16.1%||24.4%||<.001b||.822|
| Chronic obstructive pulmonary disease||6.5%||5.8%||11.2%||<.001b||.823|
| Smoking history||10.5%||10.8%||7.8%||.002b||.921|
|Dependent functional status||2.7%||1.8%||1.3%||<.001b||.579|
|American Society of Anesthesiologists class||<.001b||.551|
| I or II||46.8%||49.7%||24.8%|
|Any predischarge complication||0.9%||0.9%||0.9%||.796||.993|
Multivariate Logistic Regression for Predictors of Being Discharged to a Non-Home Facility
|Predictor||Odds Ratio||95% Confidence Interval||P|
| ≤60 y||Reference||-||-|
| 61 to 70 y||1.55||1.14–2.10||.005|
| 71 to 80 y||3.36||2.49–4.52||<.001a|
| >80 y||10.44||7.61–14.31||<.001a|
|Body mass index|
| <30 kg/m2 (non-obese)||Reference||-||-|
| 30–34.9 kg/m2 (obese class I)||1.07||0.90–1.27||.457|
| 35.0–39.9 kg/m2 (obese class II)||1.09||0.87–1.36||.458|
| >40 kg/m2 (obese class III)||1.54||1.22–1.94||<.001a|
| Diabetes mellitus||1.35||1.13–1.60||.001a|
| Chronic obstructive pulmonary disease||1.40||1.10–1.77||.006|
| Smoking history||1.26||0.97–1.64||.086|
|Dependent functional status||3.75||2.81–5.00||<.001a|
|American Society of Anesthesiologists class|
| III or IV||3.33||2.32–4.77||<.001a|
|Any predischarge complication||1.28||0.63–2.61||.499|
Unadjusted Comparison of Postdischarge Complications by Discharge Destination
|Complication||All Patients (N=9058)||Discharge to Home (n=7996)||Discharge to Inpatient Facility (n=1062)||Unadjusted P|
|Any complication (major or minor)||4.49%||3.74%||10.17%||<.001a|
|Deep venous thrombosis/pulmonary embolism||0.63%||0.53%||1.41%||.001a|
|Urinary tract infection||0.73%||0.53%||2.26%||<.001a|
Propensity-Adjusted Comparison of Postdischarge Complications by Discharge Destination
|Complication||Odds Ratio||95% Confidence Interval||Propensity-Adjusted P|
|Any complication (major or minor)||2.05||1.59–2.64||<.001a|
|Deep venous thrombosis/pulmonary embolism||1.91||1.00–3.65||.049|
|Urinary tract infection||2.66||1.52–4.67||.001a|