Risk Factors for Unexpected Admission Following Outpatient Rotator Cuff Repair: A National Database Study

Introduction Rotator cuff repair (RCR) procedures are some of the most common orthopaedic surgeries performed in the United States. Compared to other orthopaedic procedures, RCRs are of relatively low morbidity. However, complications may arise that result in readmission to an inpatient healthcare facility. The purpose of this study is to identify the demographics and risk factors associated with unplanned 30-day readmission after RCR. Methods The American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) database was used to identify patients that underwent elective RCR from 2015-2019. Univariate and multivariate analyses were utilized to assess patient demographics, comorbidities, and peri-operative variables predicting unplanned 30-day readmission. Results Of the identified 45,548 patients that underwent RCR, 597 (1.3%) required readmission within 30 days of the procedure. Multivariate analysis identified male sex (OR 1.36, 95% CI: 1.10, 1.67), hypertension (OR 1.29, 95% CI:1.03, 1.62), chronic obstructive pulmonary disease (COPD) (OR 2.07, 95% CI: 1.46, 2.93), American Society of Anesthesiologists (ASA) Class III (OR 1.85, 95% CI: 1.07, 3.18), ASA Class IV (OR 5.38, 95% CI: 2.70, 10.72), and total operative time (OR 1.002, 95% CI: 1.000, 1.004) as independent risk factors for unplanned readmission. Conclusion Unplanned 30-day readmission after RCR is infrequent. However, certain patients may be at increased risk for unplanned 30-day admission to an inpatient facility. This study confirmed male sex, COPD, hypertension, ASA Class III, ASA Class IV, and total operative time to be independent risk factors for readmission following outpatient RCR.


Introduction
Rotator cuff repair (RCR) procedures are some of the most common orthopaedic surgeries performed in the United States, with approximately 250,000 RCRs performed each year [1]. The volume of RCRs has increased substantially over the last few decades, with a major shift away from open to arthroscopic technique and routine performance in the outpatient ambulatory setting [2][3][4][5]. In general, compared to other orthopaedic procedures, RCRs have a relatively low complication rate of up to 10.6% [6][7][8]. Common complications reported after RCR are postoperative stiffness, hardware-related complications, and failure of repair [7,9,10].
Occasionally, complications may arise that result in unplanned 30-day readmission to an inpatient facility, with rates reported up to 1.6% [7,8,[11][12][13][14][15][16][17][18]. Reported complications requiring unplanned 30-day readmission include surgical site infections, postoperative pain, and medical complications such as deep venous thrombosis and pulmonary embolism [11,13]. Previous literature has been inconclusive on risk factors for complications and readmissions after RCR. The purpose of this study is to identify the demographics and risk factors that are associated with admission to an inpatient facility within 30 days of RCR.

Database
This study utilized the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) database. Data were collected by trained clinical reviewers from over 700 participating hospitals and included patient demographics, comorbidities, surgery in Current Procedural Terminology (CPT) codes, diagnoses in International Classification of Disease 9th and 10th (ICD-9, ICD-10, respectively) revision codes, and 30-day post-operative surgical outcomes.

Patient population
The ACS-NSQIP was queried for patients who underwent RCR from 2015 to 2019 using CPT codes 23410, 23412, 23420, and 29827, yielding 60,008 cases. Patients with incomplete data were excluded (n = 14,460), resulting in a study population of 45,548 patients. CPT codes 23410, 23412, and 23420 were identified as open RCR procedures, and 29827 signified arthroscopic RCR procedures.

Variables collected
The following demographic, lifestyle, and comorbidity variables were recorded: age, sex, body mass index (BMI), American Society of Anesthesiologists (ASA) classification, race, Hispanic ethnicity, bleeding disorders, chronic obstructive pulmonary disease (COPD), diabetes mellitus, hypertension, congestive heart failure, tobacco use, and chronic steroid use. Perioperative variables that were collected included anesthesia type (general versus regional), total operative time, and procedure type (open versus arthroscopic). The primary outcome of 30-day readmission was defined as unplanned hospital readmission likely related to the principal procedure.

Statistical analyses
All data were analyzed using the Statistical Package for the Social Sciences (SPSS) version 23.0 (IBM Corp., Armonk, NY). The criterion for statistical significance was set at α = 0.05. Independent-sample student's ttests, chi-square test, and, where appropriate, Fisher's exact tests were used in univariate analyses to identify demographic, lifestyle, and peri-operative variables related to 30-day readmission following RCR. Multivariate logistic regression modeling was subsequently performed. Odds ratios (ORs) with 95% confidence intervals (CIs) were calculated and reported.

Results
Of the 45,548 patients included in our sample, 597 were readmitted within the 30-day postoperative period, corresponding to a readmission rate of  The results of the univariate analysis revealed statistically significant relationships between readmission status and the tested patient variables are itemized in Table 2.  Patient race and anesthesia type were not significantly associated with readmission.
Multivariate logistic regression modeling confirmed that the following patient variables were associated with statistically significantly increased odds of readmission (see Table 3 [16]. Hill et al. found an overall readmission rate of 0.98%, with the most common reason being pulmonary embolism [13]. The authors found operative time >1.5 hours, age >40 years, ASA Class III or IV, and chronic steroid use as risk factors for readmission following arthroscopic shoulder surgery [13].
Reports of gender as a risk factor for readmission have been variable. While some reports, including the current study, demonstrate male sex is associated with an increased risk of complications and readmission, others have identified female sex to be associated with a higher rate of complications [7,15,16,19]. Gil et al. found female sex to be a significant predictor of admission after outpatient surgery [20]. On the other hand, while Kosinski et al. did not find an increased risk of readmission for male sex, they determined female sex to be associated with lower odds of admission [17]. In contrast, our findings suggested that male sex had higher odds of readmission after RCR.
Increased BMI and the presence of medical comorbidities, including COPD, hypertension, chronic steroid use, dialysis, and metastatic cancer, have also been reported to be associated with an increased rate of complications and unplanned readmission after RCR [12,13,15,16,[19][20][21]. In addition, several authors have demonstrated higher ASA Class > II to be a significant risk factor for complication and readmission [12,13,15,19,20]. The current study found patients with comorbidities, such as COPD and hypertension, to have an increased risk of readmission within 30 days of RCR. The current study also found ASA Class III and IV to be risk factors for readmission within 30 days of RCR. However, we did not find an association between CHF, diabetes, bleeding disorder, or steroid use with an increased risk of admission within 30 days of surgery.
Smoking has long been reported as a risk factor for complications following the repair of rotator cuff tears [14,16,18]. Best et al. used the ACS-NSQIP database from 2011-2016 to evaluate 5,157 patients undergoing open rotator cuff repair [22]. The authors determined that smokers are at increased risk of short-term complications, including venous thromboembolism and pulmonary embolism [22]. In a study by Kashanchi et al., the authors investigated the association between smoking status and postoperative complications within 30 days of arthroscopic RCR [23]. They demonstrated smoking to be a significant predictor of surgical complications, return to the operating room, readmission, and sepsis or septic shock [23]. Smoking was not identified as an independent risk factor for readmission after RCR in our multivariate logistic regression model.
Longer operative times have been previously recognized as a risk factor for complications after various orthopaedic procedures, including shoulder arthroscopy, anterior cruciate ligament reconstruction, and total joint arthroplasty, among others [13,18,19,[24][25][26]. Day et al. and Hill et al. found an increased risk of complications and 30-day readmission, respectively, with operative times > 90 minutes [13,18]. Boddapati et al. specifically examined shoulder arthroscopy procedure time and its effect on the rates of short-term postoperative complications, readmissions, and overnight hospital stays [19]. The authors identified an increased risk of superficial surgical site infections and overnight hospital stay for procedures lasting between 45 minutes and 90 minutes, and for procedures lasting greater than 90 minutes, when compared with procedures that were less than 45 minutes [19]. Similarly, our study demonstrated increased odds of readmission within 30 days of RCR with higher total operative times.  [14]. The authors demonstrated the arthroscopic group had a significantly lower risk of complications, a lower rate of superficial infection, a lower incidence of return to the operating room within 30 days, and a lower risk of hospital readmission [14]. Moreover, several of the previously mentioned studies have also demonstrated increased rates of complications and readmission with the open technique [16,18,20].
The authors included 45,548 patients in this study. Despite the large number of cases available for review, there are some limitations to the current study. Inherent limitations to the NSQIP database are potential coding errors and misclassifications, which may lead to incomplete patient capture. However, inter-rater reliability disagreement within the database has been shown to be low, at less than 1.8% [27]. Furthermore, the ACS-NSQIP database is reliant on reports of participating hospitals. Therefore, there is potential that the data may only reflect the population of the participating institutions, rather than a much larger population. However, as the number of participating institutions increases, the data extrapolated may be better applied to a larger population.

Conclusions
The current study analyzed a large sample of patients that underwent RCR and evaluated for risk factors associated with unplanned 30-day admission. Our analysis revealed statistically significant relationships between readmission status and the following patient variables: patient age, sex, BMI, ASA classification, Hispanic ethnicity, bleeding disorder, COPD, diabetes, hypertension, steroid use, current smoker, total operative time, and procedure type. Multivariate logistic regression modeling confirmed male sex, COPD, hypertension, ASA Class III and IV, and total operative time to be independent risk factors for readmission following outpatient RCR. Such findings can allow the orthopaedic surgeon to identify patients at higher risk for readmission preoperatively, discuss expectations with their patients, and reduce costs by preventing avoidable complications.

Additional Information Disclosures
Human subjects: All authors have confirmed that this study did not involve human participants or tissue. Animal subjects: All authors have confirmed that this study did not involve animal subjects or tissue.

Conflicts of interest:
In compliance with the ICMJE uniform disclosure form, all authors declare the following: Payment/services info: All authors have declared that no financial support was received from any organization for the submitted work. Financial relationships: All authors have declared that they have no financial relationships at present or within the previous three years with any organizations that might have an interest in the submitted work. Other relationships: All authors have declared that there are no other relationships or activities that could appear to have influenced the submitted work.