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
Maimonides Medical Center, Department of Orthopaedic Surgery, Brooklyn, NY, USAGeorge Washington University School of Medicine and Health Sciences, Department of Orthopaedic Surgery, Washington, D.C., USA
Risk factors associated with primary THA readmissions have not yet been thoroughly analyzed when stratified by underlying indication. Given that a majority of THAs are done electively in the context of osteoarthritis (OA), it remains to be explored whether or not THAs performed non-electively in the trauma setting have different readmission patterns. Therefore, the aims of this study were to identify: 1) causes of readmissions; 2) patient-related risk-factors for readmissions; and 3) costs associated with the reasons for readmissions.
Materials and methods
Patients who sustained a femoral neck fracture and underwent primary THA from 2005 to 2014 were identified. Those subsequently readmitted within 90-days following the procedure comprised the study cohort whereas those not readmitted served as the comparison cohort. Primary outcomes included identifying causes of readmissions, identifying patient-related risk-factors associated with readmissions and determining healthcare expenditures associated with the different readmission etiologies. A regression analysis was used to calculate the odds (OR) for readmissions. A p-value less than 0.01 was considered to be statistically significant.
Results
The regression model demonstrated the greatest patient-related risk factors included: electrolyte and fluid disorders (OR: 1.80, p < 0.0001), morbid obesity (OR: 1.60, p < 0.0001), pathologic weight loss (OR: 1.58, p < 0.0001), congestive heart failure (OR: 1.41, p < 0.0001), were the leading risk factors for readmissions. Pulmonary-related causes ($42,357.71) of readmission were the leading driver of costs of care.
Conclusion
Orthopaedic surgeons should identify and optimize pre-operative management of patient-related risk factors that increase readmissions following primary THA for femoral neck fractures. Additionally, pulmonary-related causes of readmission lead to the highest costs of care.
Total hip arthroplasty (THA) is one of the most commonly performed surgical procedures in the United States (U.S.), most often used for end-stage hip osteoarthritis (OA), but also for inflammatory arthritic conditions and within the setting of traumatic injuries.
Given that the number of primary THAs is expected to increase 71% to 635,000 procedures annually by 2030, the amount of spending in billions is likely to increase significantly as well.
Such large healthcare expenditures associated with THA has garnered significant government and private insurance attention, leading to the implementation of methods on reducing unnecessary healthcare expenditures, such as reducing the incidence of hospital readmissions.
As such, identifying risk factors for readmission could assist in reducing costs and alleviating this expensive financial burden on the U.S. healthcare system.
Kurtz et al. recently showed that the causes for 30-day readmissions following primary THA were the same as those for 90-day readmissions, which include dislocations, deep infections, wound infections, periprosthetic fractures, and hematomas.
Most prior studies have only examined risk factors and causes for THA readmissions at single institutions with fewer multi-center studies or as an elective procedure.
Rheumatoid arthritis is associated with higher ninety-day hospital readmission rates compared to osteoarthritis after hip or knee arthroplasty: a cohort study.
Although a majority of THAs are done electively in the context of OA, it remains to be explored whether or not THAs performed non-electively in the traumatic setting have different readmission costs, patient-associated risk factors, and causes. Such understanding is critical for optimizing pre-operative patient care to reduce healthcare costs associated with primary THA as well as for developing risk-stratification guidelines for alternative payment models including the Comprehensive Care for Joint Replacement model. Though prior studies have looked at the rates and causes of 90-day readmission for those undergoing non-elective partial or THA for femoral neck fractures, the authors did not examine the patient associated risk-factors or costs associated with different causes of readmissions.
The association between hospital length of stay and 90-day readmission risk for femoral neck fracture patients: within a total joint arthroplasty bundled payment initiative.
Therefore, the purpose of this study was to analyze a nationwide administrative database to identify: 1) causes of readmissions; 2) patient-related risk-factors for readmissions; and 3) costs associated with the reasons for readmissions.
2. Materials and Methods
2.1 Data source
A retrospective query from January 1st, 2005 to the first quarter of 2014 was performed of the 100% Medicare Standard Analytical Files (SAF) from the PearlDiver (PearlDiver Technologies, Fort Wayne, Indiana) database. The syntax and subscription-based database holds the records of nearly 200 million patients from a private payor insurance claims or the Medicare insurance claims, and has been used extensively for orthopaedic trauma research. Using syntax coding language, data from PearlDiver is extracted via reimbursement codes in the form of International Classification of Disease (ICD) and Current Procedural Terminology (CPT). As such, investigators have access to a myriad of research variables at their disposal such as complications, diagnoses, discharge dispositions, in-hospital lengths of stay, and economic data. Information is subsequently downloaded as a comma separated value (.csv) Microsoft (Microsoft Corporation, Redmond, Washington) Excel spreadsheet. Since the.csv spreadsheet is devoid of any protected health information, the current study was deemed exempt from our Institutional Review Board approvals.
2.2 Patient cohorts
The database was initially queried for all patients who underwent primary THA using CPT and ICD-9 procedural codes 27130 and 81.51, respectively. From this cohort, patients who sustained a femoral neck fracture were identified using ICD-9 diagnoses codes 820.00 to 820.09. To minimize heterogeneity and minimize any confounding within the investigation, the study only assessed intracapsular femoral neck fractures, and did not include other extracapsular fractures which patients may have sustained. Using Boolean command operations, the inclusion criteria for the study group consisted of all patients who underwent primary THA for the treatment of femoral neck fractures within the same day of the injury and were subsequently readmitted within 90-days following their index procedure. Patients undergoing primary THA for femoral neck fractures who were not readmitted within 90-days following the index procedure served as the comparison cohort within the investigation (Fig. 1). Using the “DIAGONLY” syntax command of the PearlDiver database provided a comprehensive list of the different causes of readmissions following the procedures categorized by their respective ICD-9-D diagnostic codes. Since patients may be readmitted for multiple chief complaints, the primary cause of readmission was used as the reference point. This was achieved by using the “FIELDNUMBER” command of the database. Each ICD-9-D code was then grouped and categorized into one of the following categories: cardiac causes, pulmonary-related causes, infectious, thromboembolic, genitourinary (GU)/renal causes, neoplastic, miscellaneous, gastrointestinal (GI), and musculoskeletal etiologies. These categories were chosen as they have been used in previously published studies.
Primary endpoints of this study were to determine the causes of readmission within the episode of care time period, determine patient-related risk-factors associated with readmissions following primary THA for the treatment of femoral neck fractures, and determine the associated healthcare expenditure associated with the different readmission etiologies. Patient-related risk entered into the regression model to determine the association with readmissions following the procedure included: age, gender, acquired immunodeficiency syndrome (AIDS), alcohol use disorder, arrhythmias, body mass index (BMI), congestive heart failure, coagulopathies, diabetes mellitus, depressive disorders, electrolyte and fluid disorders, hypertension, hypothyroidism, iron deficiency anemia, liver failure, lymphoma, metastatic cancer, neurological disorders, opioid use disorder, peptic ulcer disease, peripheral vascular disease, renal failure, rheumatoid arthritis, sleep apnea, tobacco use, and pathologic weight loss. These conditions were chosen as they comprise aspects of both the Charlson- (CCI) and Elixhauser-Comorbidity Indices (ECI), which have been widely used within orthopaedic studies. Additionally, these comorbid conditions were chosen at the time of admission as they are essential determinant in future complications. Economic data was determined by calculating the episode of care costs for each of the various causes of readmissions following primary THA for femoral neck fractures. Reimbursements were used as a proxy for costs of care, as done in previously published studies using the same database.
2.4 Data analyses
Statistical analyses were performed using the open programming language R (R, Foundation for Computational Statistics, Vienna, Austria). Baseline demographics between the study and control cohorts were compared using Pearson's chi-square analyses. Welch's t-test was used to compare mean ECI scores between the two groups. Fischer's exact test was utilized in cohorts with a sample size of less than 11. It should be of note that the age is provided as a categorical variable within the database with five-year increments, and as such, Pearson's chi-square analyses were used to test for significance between the two cohorts. To determine the association of specific patient-related risk factors on the risk of readmissions a multivariate binomial logistic regression model was constructed to calculate the odds (OR) and respective 95% confidence intervals (95%CI) of various comorbidities on the dependent variable of interest. For age, patients less than the age of 64 was used as a reference point; whereas for sex, female patients were considered the reference group. Additionally, analysis of variance (ANOVA) was performed to assess for statistical differences among the costs of care for the different etiologies which led to readmissions within 90-days following the index procedure. Tukey's post-hoc analysis was performed to determine which variables were found to be statistically significant which led to significance in the ANOVA analysis. Due to the ease of finding statistical significance with large administrative registries, a Bonferroni-correction was performed to reduce the probability of a type I error. As such, an alpha value less than 0.001 was considered as the threshold to define statistical significance.
3. Results
After the inclusion and exclusion criteria, a total of 57,191 patients were included in the investigation between the study (n = 10,330) and control (n = 46,681) cohorts. When assessing baseline demographic profiles (Table 1), patients older than the age of 85 represented the greatest cohort (29.12%) of being readmitted following the index procedure, whereas patients less than the age of 64 had the least representation of being readmitted (9.24%). Female patients had a higher percentage of being readmitted compared to their male counterparts (69.25 vs. 30.75%). When comparing the prevalence of comorbid conditions, patients readmitted within the episode of care period following primary THA for the treatment of femoral neck fractures had significantly higher prevalence of certain comorbid conditions. Specifically, study group patients had higher prevalence of alcohol use disorder (8.98 vs. 6.34%, p < 0.0001), morbid obesity – defined as a BMI greater than 40 kg/m2 (2.65 vs. 1.17%, p < 0.0001), congestive heart failure (54.91 vs. 36.34%, p < 0.0001), diabetes mellitus (43.41 vs. 33.18%, p < 0.0001), sleep apnea (12.00 vs. 8.66%, p < 0.0001) in addition to other comorbid conditions. Readmitted patients also had a significantly higher mean ECI scores (11 vs. 8, p < 0.0001).
Table 1Comparison of patient demographics among patients readmitted within ninety-days following primary total hip arthroplasty for femoral neck fractures. AIDS = acquired immunodeficiency syndrome; ECI = Elixhauser-Comorbidity index.
When assessing the leading causes of readmissions within 90-days following primary THA for the treatment of femoral neck fractures, musculoskeletal causes (24.5%) was the leading reason, followed by infectious causes (18.6%), and cardiac etiologies (12.5%), whereas neoplastic (1.2%) and endocrine-related causes (1.0%) were associated with the lowest reasons for readmissions within the study cohort. A comprehensive list of the causes of readmissions for each of the three leading etiologies can be found in Supplementary Tables 1–3. The regression model demonstrated that the greatest patient-related risk factors associated with readmissions within 90-days following primary THA for the treatment of femoral neck fractures included: electrolyte and fluid disorders (OR: 1.80, 95%CI: 1.62–1.92, p < 0.0001), morbid obesity (OR: 1.60, 95%CI: 1.36–1.88, p < 0.0001), pathologic weight loss (OR: 1.58, 95%CI: 1.50–1.66, p < 0.0001), congestive heart failure (CHF) (OR: 1.41, 95%CI: 1.34–1.48, p < 0.0001), opioid use disorders (OR: 1.31, 95%CI: 1.15–1.49, p < 0.0001), depressive disorders (OR: 1.25, 95%CI: 1.19–1.31, p < 0.0001) and other conditions (Table 2). Obesity trended towards statistical significance (OR: 1.12, 95%CI: 1.00–1.24, p = 0.034), but failed to reach statistical significance with the threshold used for this investigation (Table 2). When analyzing healthcare expenditures between the different etiologies of readmissions, pulmonary-related causes ($42,357.71) were the leading driver of costs of care followed by infectious ($40,359.01) and musculoskeletal causes ($38,983.95). The three drivers associated with the lowest costs included: psychiatric ($33,511.02), gastrointestinal ($32,548.05), and thromboembolic causes. ($32,159.87).
Table 2Multivariate binomial logistic regression analysis determining the association of patient-related risk factors for readmissions within ninety-days following primary total hip arthroplasty for femoral neck fractures. OR = odds-ratio; 95%CI = 95% confidence interval; BMI = body mass index; Bolded values signify statistical significance.
ANOVA testing demonstrated statistically significant differences across the costs of care for the different causes of readmission within the episode of care time interval following primary THA for intracapsular femoral neck fractures (p < 0.0001). Post-hoc analysis demonstrated there was significant differences in the costs of care between pulmonary-related complications and the other etiologies which led to readmissions with this investigation (p < 0.0001).
4. Discussion
As the rates of THA and its associated costs continue to increase within the U.S., identifying patient risk factors associated with readmission rates is necessary for effective pre-operative medical management that reduces healthcare costs.
Past studies have looked at patient-associated risk factors and costs associated with 30- and 90-day readmissions for primary THA within the elective setting; however, the literature is limited within the traumatic setting.
Rheumatoid arthritis is associated with higher ninety-day hospital readmission rates compared to osteoarthritis after hip or knee arthroplasty: a cohort study.
As such, the purpose of this study was to characterize causes of readmissions, identify patient-associated risk factors and costs for 90-day readmissions in patients who underwent non-elective primary THA following femoral neck fractures. The investigation found musculoskeletal, infectious, and cardiac related manifestations as being the main reason for readmissions following surgical fixation of femoral neck fractures. Additionally, a myriad of patient-related risk factors associated with readmissions were observed such as: cardiac arrythmias, morbid obesity, CHF, and others. Additionally, pulmonary-related causes of readmission were the leading driver of costs of care while thromboembolic causes of readmission were the lowest driver of costs.
The results of this retrospective investigation are consistent with the findings from prior literature exploring readmissions following THA. In regard to comorbidities, several studies have shown certain patient-associated risk factors to be independently associated with readmission following THA. Schairer et al. found that cardiac valve disease (hazard ratio (HR): 2.52, p = 0.009), substance abuse (HR: 2.05, p = 0.094), and diabetes with complications (HR: 11.55, p < 0.001), are independent risk-factors for unplanned readmission following primary or revision THA.
Moreover, Saucedo et al. showed that coronary artery disease, age less than 50, and BMI < 18.5 kg/m2 are independent risk factors for readmission 90-days after THA.
This finding of BMI < 18.5 kg/m2 being an independent risk factor is in contrast to our study in which BMI < 19 only trended towards significance and instead morbid obesity was found to be statistically significant.
This contrast is likely due to the study by Saucedo et al. being limited to a single institution and not stratifying results by elective versus non-elective indications for THA.
Additionally, the finding of depressive disorders being an independent risk factor for 90-day readmission is consistent with a prior study by Lavernia et al., where patients with mental health conditions were found to be readmitted twice as frequently following hip arthroplasty compared to patients without psychiatric diseases.
Furthermore, in a study by Pugely et al., obesity, steroid use, bleeding disorders, dependent functional status, and high American Society of Anesthesiologists (ASA) class predicted 30-day readmission after elective primary THA.
This is similar to our study's findings in which we found both a BMI between 40 and 70 and coagulopathies to be independent risk factors for readmission. Contrary to our study in which many comorbidities were found to be risk factors for readmission, a study by Zmistowski et al. used the Charlson Comorbidity Index (CCI) to quantify the general health of patients but found that CCI was not a significant independent risk factor for unplanned readmissions following total joint arthroplasty.
Moreover, in contrast to our finding of CHF's association with 90-day readmissions following primary THA for femoral neck fractures, a study by Paxton et al. found that CHF was not an independent risk factor for 30-day readmission following primary THA, though they did not stratify by underlying indication for THA.
Additionally, that study by Paxton et al. was especially limited due to geographic factors, including only patients from California and Hawaii, which limits the generalizability of the study to the entire U.S. population.
Healthcare policy makers and insurance companies can use these findings for improving alternative payment models through risk adjustment for patients at high risk of readmission. Such adjustments may help prevent penalizing orthopaedic surgeons that are caring for high-risk patients.
To our knowledge, this is the first study to quantify healthcare expenditures associated with readmissions following primary THA for femoral neck fractures. This study found that pulmonary related causes of readmission following primary THA for femoral neck fractures are the leading driver of costs while the lowest costs were associated with thromboembolic causes of readmission. Pulmonary-related causes of readmissions include, but are not limited to, COPD exacerbations (most common), pneumonia, and acute respiratory distress syndrome (ARDS).
Acute COPD exacerbations have been shown to lead to an estimated $13.2 billion in spending annually while pneumonia-based readmissions have an estimated cost of $10 billion annually.
More specifically, a prior study analyzing readmissions following THA showed that the average total cost for pneumonia-based readmissions was $18,636 dollars.
Additionally, the average cost per readmission for respiratory failure has been shown to be nearly $30,000 and near $14,000 for pneumonia following head and neck cancer reconstruction surgery.
Moreover, pulmonary complications have been shown to lead to the highest costs ($9,429, p < 0.001) added to readmission costs following radical cystectomy.
Collectively, these prior studies highlight that pulmonary-causes of readmission can lead to a significant financial burden on individual patients and healthcare systems. Future studies should aim to understand why pulmonary-causes of readmission are the leading driver of costs in patients undergoing non-elective THA so that early interventions can be implemented to prevent these complications and reduce costs.
As with other administrative database studies, our study is not without limitations. Though we had a large patient population size with adequate statistical power, our usage of the Medicare database for analysis has several inherent limitations. Any database requires accurate coding and may have minor inaccuracies due to miscoding or not coding diagnoses. For example, a prior study showed that the prevalence of obesity and alcohol abuse based on codes in a large administrative database was significantly lower than that from a direct questioning survey.
Differences in the prevalence of obesity, smoking and alcohol in the United States nationwide inpatient sample and the behavioral risk factor surveillance system.
Additionally, as Medicare is a federal health insurance program primarily for those above or equal to age 65 with only a select population under age 65 (e.g., those with end-stage renal disease), it is possible that our database is not a true cross-sectional representation of the patient population of femoral neck fractures. Furthermore, large datasets such as ours for Medicare patients may find statistically significant associations that are not necessarily clinically significant or translatable.
Additionally, due to restrictions within the PearlDiver database, analyzing time intervals for each of the different etiologies of readmission was not attainable, but can serve as the basis for future prospective studies. Despite these limitations, to the best of our knowledge, this study is the first to analyze a nationwide population and assess a large number of patient-related risk factors, healthcare costs, and causes of 90-day readmissions in patients undergoing primary THA for femoral neck fractures within the Medicare population.
5. Conclusion
The results of this study demonstrate specific patient risk factors and healthcare expenditures associated with 90-day readmissions following primary THA for femoral neck fractures. Orthopaedic surgeons need to be aware of patient comorbidities such as electrolyte and fluid disorders, pathologic weight loss, and CHF that increase patient risk for 90-day readmission. Future studies should continue to delineate patient-specific risk factors and healthcare costs associated with THA for different indications to improve future risk-stratification and patient education. Additionally, future studies should evaluate long-term outcomes for surgeons managing these high-risk patients. More studies focused on ways to minimize readmission risk and optimize care in these high-risk patient populations to reduce expenditures for the more costly causes of readmissions such as pulmonary causes is needed.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Appendix A. Supplementary data
The following is the Supplementary data to this article:
Rheumatoid arthritis is associated with higher ninety-day hospital readmission rates compared to osteoarthritis after hip or knee arthroplasty: a cohort study.
The association between hospital length of stay and 90-day readmission risk for femoral neck fracture patients: within a total joint arthroplasty bundled payment initiative.
Differences in the prevalence of obesity, smoking and alcohol in the United States nationwide inpatient sample and the behavioral risk factor surveillance system.
00449-5&id=gr1.jpg){kind=link}