If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Increased Readmission Rates but No Difference in Complication Rates in Patients Undergoing Inpatient Versus Outpatient Hip Arthroscopy: A Large Matched-Cohort Insurance Database Analysis
To compare the early postoperative outcomes of patients undergoing inpatient versus outpatient hip arthroscopy and identify any characteristics that may serve as predictors of these complications
Methods
The PearlDiver Mariner insurance database was queried for all patients who underwent hip arthroscopy from 2010 to 2019. Patients were matched based on Charlson Comorbidity Index, age, and sex. Outcomes recorded included postoperative complications and return to care within 90 days. Predictors of complications were assessed via multivariable logistic regression controlling for age, sex, Charlson Comorbidity Index, comorbidities, surgical setting, and procedure type
Results
The final matched cohort included 832 inpatients and 1,356 matched patients. Fifty-eight patients (7.0%) who underwent inpatient surgery versus 25 patients (1.8%) who underwent outpatient surgery were readmitted (P < .01). Of the readmitted patients, 31 inpatients (3.7%) and 5 outpatients (0.4%) were readmitted for hip-related reasons (P < .01). No significant differences were observed in emergency department visits (67 inpatients [8.1%] vs 84 outpatients [6.2%], P = .11), intensive care unit admissions (3 [0.4%] vs 1 [0.1%], P = .31), or revision hip arthroscopies (43 [5.2%] vs 65 [4.8%], P = .77). A multivariable model of complications correcting for baseline differences in chronic obstructive pulmonary disease, coronary artery disease, diabetes, hypertension, ischemic heart disease, tobacco use, and inpatient status found that age (adjusted odds ratio [OR], 0.92; 95% confidence interval [CI], 0.85-0.99; P = .03), coronary artery disease (adjusted OR, 12.82; 95% CI, 1.18-140.02; P = .03), and inpatient setting (adjusted OR, 20.59; 95% CI, 3.48-401.65; P = .01) were significantly associated with complications. No procedure type was associated with complication rates
Conclusions
Compared with the outpatient setting, inpatient hip arthroscopy is associated with higher rates of readmission in a cohort matched for age, sex, and comorbidities. Complications after inpatient hip arthroscopy appear to be related to comorbidities rather than procedure-related factors. The decision to conduct an inpatient hip arthroscopy should prioritize consideration of patient comorbidities over the type of procedure
Level of Evidence
Level III, retrospective cohort study.
Hip arthroscopy is a relatively new, increasingly common procedure that grew by 3.5-fold to 25-fold in the early 2000s, according to representative epidemiologic studies.
This procedure is conducted for a variety of indications including femoroacetabular impingement (FAI), labral tears, hip synovitis, early osteoarthritis, or intra-articular loose bodies.
The decision to conduct a surgical procedure in the hospital versus in an outpatient surgical center depends on patient factors (e.g., comorbidities, anesthesia concerns, and body habitus) and surgical factors (e.g., surgical complexity, expected blood loss and possible transfusion requirement, and need for special equipment). These factors should be carefully considered to mitigate the risk of postoperative complications that may occur after hip arthroscopy, such as nerve injury,
Because hip arthroscopy is a relatively new procedure, research is still ongoing to identify evidence-based factors that would contribute to a surgeon’s decision to choose the appropriate surgical setting.
Although previous studies have identified factors associated with long-term patient-reported outcomes after hip arthroscopy,
there are limited data addressing the differences in the complication rates of outpatient versus inpatient hip arthroscopy in the immediate postoperative period. Furthermore, little is known about risk factors associated with these complications and how their effects may vary between inpatient and outpatient settings. The purpose of this study was to compare the early postoperative outcomes of patients undergoing inpatient versus outpatient hip arthroscopy and identify any characteristics that may serve as predictors of these complications. Our hypotheses were that the matched cohort of patients undergoing inpatient surgery would have similar complication rates to patients undergoing outpatient surgery and that patients with chronic disease would be at an increased risk of postoperative complications regardless of where their initial surgical procedure was conducted.
Methods
Data Collection
Data were extracted using the M91Ortho data set of PearlDiver Technologies (Colorado Springs, CO), which contains data from 91 million orthopaedic patients enrolled in the Mariner commercial insurance database and Medicare Standard Analytical Files from 2007 to 2020. The Mariner database overall includes 144 million patients insured by commercial, Medicare, Medicaid, government, and cash payers. The database was queried for all patients who underwent hip arthroscopy (Current Procedural Terminology [CPT] codes 29860, 29861, 29862, 29863, 29914, 29915, and 29916). Patients were matched based on age, sex, Charlson Comorbidity Index (CCI), comorbidities, and procedure type at a 1:2 ratio.
Variables collected included age, sex, region, cost, and comorbidities. Early postoperative complications were identified by assessing for concomitant surgical-site infection (SSI), acute kidney injury, myocardial infarction, cardiac arrest, deep venous thrombosis, wound disruption or seroma, hematoma, nerve injury, pneumonia, pulmonary embolism, anemia requiring transfusion, urinary tract infection, avascular necrosis of the femoral head, sepsis, joint infection, or death within 90 days. We also recorded any ED visits (in general and for hip-related issues), readmissions, intensive care unit (ICU) admissions, or reoperations within 90 days of discharge. Hip-related issues were defined by an admission or ED visit with an International Classification of Diseases, Ninth Revision (ICD-9) or International Classification of Diseases, Tenth Revision (ICD-10) code for a broad range of hip pathologies (Appendix Table 1).
Statistical Analysis
Categorical data were reported as counts and percentages. Continuous data were reported as medians and interquartile ranges. Analysis of hip arthroscopy over time was reported by normalizing the annual procedure counts to the number of persons in the M91 data set each year and conducting a Pearson correlation between hip arthroscopy utilization and year. The year 2020 was excluded from this time-series analysis to avoid the known effects of the COVID-19 (coronavirus disease 2019) pandemic on elective surgery rates.
Differences between groups were analyzed using the Fisher exact test for categorical data and the Wilcoxon rank sum test for continuous data.
Predictors of complication rates were assessed using a multivariable logistic regression model controlling for age, sex, CCI, inpatient versus outpatient setting, procedure type, and factors previously reported to be linked to measured complications in the literature: diabetes mellitus, chronic kidney disease, obesity, tobacco use, chronic obstructive pulmonary disease (COPD), and coronary artery disease (CAD).
A second predictive model was performed controlling for age, sex, CCI, and procedure type via CPT code. To better elucidate the reasons for ED visits, readmissions, or ICU admissions if significant differences were found, we used the “association track” code, which lists the frequency of other International Classification of Diseases (ICD) or CPT codes applied to the same visit, to identify the most common reasons for the return to hospital-level care. The type I error rate was set at .05. Our sample size of 2,188 hip arthroscopy procedures gave our study over 90% power to detect a 10% difference in complication rates between the inpatient and outpatient groups. Analyses were performed using Bellwether (version 2.0; PearlDiver Technologies).
Results
Baseline Demographic Characteristics
A total of 46,867 hip arthroscopy procedures were performed during the study period; 1,677 patients (0.4%) were excluded for loss to follow-up within 90 days. Of the remaining 45,190 hip arthroscopies, 1,401 (3.1%) were performed in the inpatient setting. After matching, the final cohort comprised 832 patients who underwent inpatient hip arthroscopy and 1,356 matched patients who underwent outpatient hip arthroscopy (Fig 1). Baseline demographic characteristics were similar between groups regarding age, sex, CCI, insurance plan, and hip arthroscopic procedures (Table 1). Despite matching, inpatients were more likely to have COPD (177 inpatients [21.3%] vs 241 outpatients [17.8%], P = .05), CAD (51 inpatients [6.1%] vs 53 outpatients [3.9%], P = .02), diabetes (110 inpatients [13.2%] vs 130 outpatients [9.6%], P = .01), hypertension (229 inpatients [27.5%] vs 300 outpatients [22.1%], P < .01), and ischemic heart disease (38 inpatients [4.6%] vs 39 outpatients [2.9%], P = .05) and to use tobacco (176 inpatients [21.2%] vs 239 outpatients [17.6%], P = .05). Notably, the length of stay was 0 days for all patients in both groups.
Fig 1STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) diagram of patients included in matched-cohort study of inpatient versus outpatient hip arthroscopy.
Inpatient history and outpatient history are shown in Figure 2. From 2010 to 2019, the prevalence of outpatient hip arthroscopy increased from 47 operations to 72 operations per 1 million person-years (r2 = 0.77; 95% confidence interval [CI], 0.28-0.94; P = .009) (Fig 1). By contrast, the prevalence of inpatient hip arthroscopy decreased from 119 operations to 43 operations per 1 million person-years (r2 = –0.69; 95% CI, –0.92 to –0.11; P = .03). Overall, outpatient procedures comprised most hip arthroscopies.
Fig 2Prevalence of patients undergoing hip arthroscopy as inpatient versus outpatient from 2010 to 2020 within large insurance database, reported as number of cases per 1 million person-years.
Complication Rates After Inpatient Versus Outpatient Hip Arthroscopy
Complication rates were similar within the first 90 days between the inpatient and outpatient hip arthroscopy groups. No significant differences were observed between groups regarding SSI, acute kidney injury, cardiac arrest, deep venous thrombosis, wound disruption or seroma, hematoma, nerve injury, pneumonia, pulmonary embolism, urinary tract infection, avascular necrosis of the femoral head, sepsis, joint infection, or death (Table 2). However, inpatients were more likely to have anemia requiring a transfusion (6 inpatients [0.7%] vs 1 outpatient [0.1%], P = .03).
Table 2Complications of Matched Patients Who Underwent Inpatient and Outpatient Hip Arthroscopy
ED Visits, Readmissions, ICU Admissions, and Reoperations
We observed no difference between groups pertaining to overall ED visits (P = .11), hip-related ED visits (P = .56), and ICU admissions (P = .31) within 90 days of discharge after surgery. There were 43 revision hip arthroscopies in the inpatient group (5.2%) and 65 revisions in the outpatient group (4.8%, P = .77). The groups differed, however, in the rates of overall readmission and readmission for a hip-related issue. Readmission to the hospital within 3 months of discharge occurred in 58 patients in the inpatient cohort (7.0%) versus 25 patients in the outpatient cohort (1.8%, P < .01). Of those patients who were readmitted for hip-related reasons, 31 (3.7%) were from the inpatient group whereas 5 (0.4%) were from the outpatient group (P < .01, Table 3).
Table 3ED Visits, Readmissions, ICU Admissions, Reoperations, Revisions to THA, and Cases of Avascular Necrosis for Inpatient Versus Outpatient Hip Arthroscopy
An exploratory analysis of ICD codes associated with readmissions found that the most common diagnosis codes applied to readmissions included other specific congenital deformities of the hip (91 instances), sprain in the hip region (87 instances), and joint pain (83 instances). No non–hip-related diagnoses exceeded the censoring threshold of 11 counts in either group. A similar analysis of ICD codes associated with readmissions for hip-related reasons found that the top diagnoses associated with readmissions for hip-related reasons were other specific congenital deformities of the hip (80 instances), sprain in the hip region (47 instances), and joint pain (33 instances).
Risk Factors for 90-Day Complications
Two multivariable models were used to identify risk factors for complications: one assessing patient-related risk factors and the other assessing procedure-related risk factors. A multivariable analysis of complications correcting for baseline differences in age, sex, CCI, COPD, CAD, diabetes, hypertension, ischemic heart disease, tobacco use, and inpatient status found that age (adjusted odds ratio [OR], 0.92; 95% CI, 0.85-0.99; P = .03), CAD (adjusted OR, 12.82; 95% CI, 1.18-140.02; P = .03), and inpatient setting (adjusted OR, 20.59; 95% CI, 3.48-401.65; P = .01) were significantly associated with complications (Table 4). Stepwise addition of congenital hip abnormalities as an independent risk factor found that inpatient status (adjusted OR, 20.51; 95% CI, 3.18-429.45; P = .008) and congenital hip abnormalities (adjusted OR, 8.38; 95% CI, 1.46-61.09; P = .02) were significant predictors of complications (Appendix Table 2). A multivariable analysis of complications correcting for procedure type, sex, age, inpatient status, and CCI found that no procedure type was significantly associated with a risk of a complication (Table 5).
Table 4Risk of Complications Within 90 Days, Corrected for Baseline Comorbidities
The principal findings of this study are as follows: (1) In a patient cohort matched for age, sex, and comorbidities, inpatient hip arthroscopy was associated with higher rates of readmission compared with outpatient hip arthroscopy, even when correcting for baseline differences in comorbidities. (2) There were no significant differences in ED visits, ICU admissions, or medical complications between the inpatient and outpatient groups. (3) Inpatient status was a significant predictor of medical complications in the 90-day postoperative period when correcting for baseline comorbidities. (4) No procedure type significantly affected the risk of complications.
Despite matching on age and comorbidities, patients who underwent inpatient hip arthroscopy were more likely than patients who underwent outpatient hip arthroscopy to be readmitted within 90 days. In the current literature, the most common risk factors for readmission are increasing age, obesity, hypertension, corticosteroid use, perioperative blood transfusions, SSIs, wound complications, and thromboembolic events.
The association of body mass index with same-day hospital admission, postoperative complications, and 30-day readmission following day-case eligible joint arthroscopy: A national registry analysis.
Inpatient hip arthroscopy was still a significant risk factor for readmission despite controlling for baseline differences in comorbidities, suggesting that local rather than systemic pathology may contribute to these differences in readmission rates.
An exploratory analysis found that the leading diagnosis codes associated with readmissions were congenital hip deformity, hip sprain, and hip pain, which suggests that inpatients are being readmitted at higher rates specifically for hip derangements. This finding is limited by the coding of diagnoses in the administrative database; for example, it is possible that FAI may have been coded as a congenital hip disorder. Still, the potential for poorer outcomes associated with congenital hip deformities is supported by the literature. Several studies have shown that patients with developmental dysplasia of the hip experience higher failure rates and poorer outcomes after FAI arthroscopy compared with patients without dysplasia.
Predictors of poor clinical outcome after arthroscopic labral preservation, capsular plication, and cam osteoplasty in the setting of borderline hip dysplasia.
There may be factors related to the hospital recovery setting partially driving the higher readmission rates of inpatient procedures. Because patients are less mobile while in the hospital, they may be significantly less active overall in the postoperative period, thus increasing the risk of fibrosis and pain. Recovery in the familiar home environment may play a psychological role in helping the patient recover compared with an unfamiliar hospital setting, surrounded by other patients who are sick or recovering.
Because there was no significant difference between the inpatient and outpatient cohorts in the rate of any of the most common medical complications driving readmission, it is possible that nonmedical reasons underlie the differences in readmission rates. One important feature to consider is the socioeconomic status of included patients. It is well established that higher socioeconomic status positively influences patient outcomes after other hip and arthroscopic procedures, both short term and long term.
Low socioeconomic status is associated with increased postoperative complication, long-term stiffness, and increased revision rates after arthroscopic rotator cuff repair in 132,420 patients of the Medicare population.
There were no differences between groups in the proportion of patients using public insurance, which is often used as a proxy for socioeconomic status in the orthopaedic literature. In addition, social support may play a role in the differences found in readmission rates for patients undergoing inpatient versus outpatient hip arthroscopy. Social support, which includes assistance with postoperative care, nutrition, and physical comfort, as well as emotional support, may improve patient outcomes in the recovery period.
and thus may contribute to the higher rate of readmission in the inpatient group. Patients with less social support at home will need to be readmitted sooner or for less serious presentations simply because they have fewer options for at-home care. Finally, patients may elect to undergo surgery in the highly monitored inpatient setting because of anxiety about undergoing surgery away from specialized resources and staff available at a hospital. A higher degree of anxiety or lower self-efficacy has been associated with worse outcomes after hip arthroscopy.
Jochimsen KN, Noehren B, Mattacola CG, Stasi S Di, Duncan ST, Jacobs CA. Low self-efficacy increases the odds of elevated post-operative pain following hip arthroscopy for femoroacetabular impingement syndrome [published online February 24, 2021]. J Athl Train. doi:10.4085/139-20.
There is a paucity of data examining the safety outcomes of inpatient versus outpatient hip arthroscopy. However, our results are consistent with similar studies in the literature comparing the complication rates of other arthroscopic and hip procedures in the inpatient versus outpatient setting. In a database study on anterior cruciate ligament reconstruction, Lu et al.
suggested that inpatient surgery is associated with a greater risk of SSI and readmission. By contrast, a systematic review and meta-analysis by Ferrari et al.
found no difference in complication rates after anterior cruciate ligament reconstruction conducted in the inpatient versus outpatient setting. In a matched-cohort analysis of patients undergoing inpatient versus outpatient rotator cuff repair, Khazi et al.
Inpatient arthroscopic rotator cuff repair is associated with higher postoperative complications compared with same-day discharge: A matched cohort analysis.
reported increased rates of complications for inpatient surgical procedures. Although few studies have compared minimally invasive hip procedures in inpatient versus outpatient settings, similar studies have been conducted in hip arthroplasty patients. A study by Rosinsky et al.
found that there were no differences reported in inpatient versus outpatient complication rates, ED visits, or ICU visits. However, their arthroplasty study did not find differences in readmission rates after hip replacement by surgical setting as we did in our arthroscopy study. Notably, compared with hip arthroscopy, hip arthroplasty is a procedure with much greater morbidity that is often conducted in patients who are older and more ill. Therefore, a surgeon’s criteria to identify patients appropriate for outpatient hip arthroplasty may be much stricter than his or her criteria for hip arthroscopy. Ultimately, this may lead to a selection bias that leads to fewer differences in complication rates after this procedure.
There are several strengths of this study. By matching the patients based on age, sex, and comorbidities, we have mitigated some of the established confounding effects of these variables on postsurgical outcomes.
Our study also has strong generalizability because we studied a large cohort of patients representing multiple practice settings, surgeons, and environments. Finally, by analyzing specific safety outcomes (i.e., various medical complications, ICU visits, ED visits, and readmissions) and identifying their associated diagnoses, we were able to more specifically address how the inpatient versus outpatient setting affects complications that may arise in the short-term postoperative period.
Limitations
This study has several limitations common to database studies. Despite matching on the CCI and comorbidities, there were baseline differences in the rates of COPD, CAD, diabetes, hypertension, and ischemic heart disease. However, in a multivariable analysis correcting for these baseline differences, inpatient arthroscopy was still a significant risk factor for readmission. There also may have been baseline differences between populations not well captured within the CCI that led to an imbalance in the risk of readmission between groups.
Matching on the CCI may have led to balanced rates of complications related to cardiovascular, pulmonary, and renal disease but would not capture pre-existing musculoskeletal or pain-related pathology that may help explain differences between these groups.
In addition, the only procedural factor we could analyze in our data set was procedure code, or type of arthroscopy. Within a given procedure type, the complexity of the procedure may vary. For example, arthroscopy for removal of loose bodies may vary in the procedure duration, depth of invasion, and complication rates based on the size of the loose bodies and their accessibility within the hip region.
Because we could not capture these case specifics in our analysis, this may have contributed to the wide CI of the OR for surgical complications within each procedure type. Assessing procedure-specific differences in complication rates was also difficult because of a lower sample size within each arm, limiting our statistical power.
Finally, as with all large insurance database studies, our study relies on appropriate application and specificity of administrative claims coding. Administrative codes are generally broad rather than specific, leading to a loss of granularity in the data assessed by database studies. The choice of a given administrative code is user dependent and may lead to variations in data collection. For example, a patient with right-sided FAI may be coded as having “other specific joint derangements of right hip, not otherwise specified” (ICD-10 code D-M24851), “osteophyte, right hip” (ICD-10 code D-M25751), “chondrolysis, right hip” (ICD-10 code D-M94351), or “right hip pain” (ICD-10 code D-M25551), among other options. Our study attempts to mitigate some of the variability in diagnostic coding by defining our inclusion criteria using CPT codes, which are less heterogeneous and more specific. Furthermore, the severity of a particular pathology cannot be determined by the administrative code alone, and retrospective studies incorporating radiographic and arthroscopic information are warranted to better understand the factors underlying complications after inpatient or outpatient hip arthroscopy. Although the possibility of miscoding cannot be excluded, PearlDiver Technologies has dedicated staff who conduct regular quality checks to ensure the accuracy of the data set to mitigate this risk.
Conclusions
Compared with the outpatient setting, inpatient hip arthroscopy is associated with higher rates of readmission in a cohort matched for age, sex, and comorbidities. Complications after inpatient hip arthroscopy appear to be related to comorbidities rather than procedure-related factors. The decision to conduct an inpatient hip arthroscopy should prioritize consideration of patient comorbidities over the type of procedure.
Appendix
Appendix Table 1Codes Used for Querying PearlDiver Database
NOTE. Colon means all consecutive ICD or CPT codes between the numeric suffixes are included.
AVN, avascular necrosis; COPD, chronic obstructive pulmonary disease; CPT, Current Procedural Terminology; ED, emergency department; ICD-9, International Classification of Diseases, Ninth Revision; ICD-10, International Classification of Diseases, Tenth Revision; ICU, intensive care unit; SSI, surgical-site infection; THA, total hip arthroplasty.
The association of body mass index with same-day hospital admission, postoperative complications, and 30-day readmission following day-case eligible joint arthroscopy: A national registry analysis.
Predictors of poor clinical outcome after arthroscopic labral preservation, capsular plication, and cam osteoplasty in the setting of borderline hip dysplasia.
Low socioeconomic status is associated with increased postoperative complication, long-term stiffness, and increased revision rates after arthroscopic rotator cuff repair in 132,420 patients of the Medicare population.
Jochimsen KN, Noehren B, Mattacola CG, Stasi S Di, Duncan ST, Jacobs CA. Low self-efficacy increases the odds of elevated post-operative pain following hip arthroscopy for femoroacetabular impingement syndrome [published online February 24, 2021]. J Athl Train. doi:10.4085/139-20.
Inpatient arthroscopic rotator cuff repair is associated with higher postoperative complications compared with same-day discharge: A matched cohort analysis.
The authors report the following potential conflicts of interest or sources of funding: E.J.B. owns stock or stock options in Amgen, Johnson & Johnson, and Pfizer. R.M. has a leadership or fiduciary role in American Orthopaedic Society for Sports Medicine (AOSSM), International Society of Arthroscopy, Knee Surgery and Orthopaedic Sports Medicine, and Journal of Bone and Joint Surgery (American edition) and owns stock or stock options in ROMTech. J.C. receives consulting fees from Arthrex, ConMed Linvatec, Ossur, and Smith & Nephew and has a leadership or fiduciary role in AOSSM, AANA, and International Society of Arthroscopy, Knee Surgery and Orthopaedic Sports Medicine. S.J.N. receives grants or contracts from Allosource, Arthrex, Athletico, DJ Orthopaedics, Linvatec, Miomed, Smith & Nephew, and Stryker; receives royalties or licenses from Ossur and Springer; receives consulting fees from Stryker; has a leadership or fiduciary role in American Orthopaedic Association and AOSSM; and receives equipment, materials, drugs, medical writing, gifts, or other services from Springer and Stryker. B.F. receives grants or contracts from Arthrex, Smith & Nephew, and Stryker; receives royalties or licenses from Elsevier; receives consulting fees from Smith & Nephew; has a leadership or fiduciary role in AOSSM and Video Journal of Sports Medicine; owns stock or stock options in I-BrainTech, Jace Medical, and Sparta Biopharma; and receives equipment, materials, drugs, medical writing, gifts, or other services from Elsevier. Full ICMJE author disclosure forms are available for this article online, as supplementary material.
00018-9&id=gr1.jpg){kind=link}
00018-9&id=gr2.jpg){kind=link}