Abstract
Objective
To examine if an underlying diagnosis of rheumatoid arthritis (RA) or osteoarthritis (OA) impacts the 90-day readmission rates after total hip or knee arthroplasty (THA or TKA).
Methods
Prospectively collected data from an integrated healthcare system Total Joint Replacement Registry of adults with RA or OA undergoing unilateral primary THA or TKA during 2009-2011 were analyzed. Adjusted logistic regression models for 90-day readmission were fit. Odds ratios with 95% confidence intervals (CI) were calculated. Study year was an effect modifier for the outcome, therefore separate analyses were conducted for each of the three study years.
Results
Of the 34,311 patients, 496 had RA and 33,815 had OA. Comparing RA and OA, there were: 73% and 61% women; 45% and 70% Caucasians; and the mean age was lower, 61 vs. 67 years (p<0.001). Respective crude 90-day readmission rates were 8.5% and 6.7%. The adjusted odds of 90-day readmission increased from year to year for RA compared to OA patients, from 0.89 (95% CI, 0.46-1.71) in 2009 to 1.34 (95% CI, 0.69-2.61) in 2010 to 1.74 (95% CI, 1.16-2.60) in 2011. The two most common readmission reasons were: joint prosthesis infection (10.2%) and septicemia (10.2%) in RA; joint prosthesis infection (5.7%) and other postoperative infection (5.1%) in OA.
Conclusions
RA is a risk factor for 90-day readmission after primary TKA or THA. An increasing risk of readmissions noted in RA in 2011 is concerning and indicates further studies should examine the reasons for this increasing trend.
Keywords: Total hip replacement, total knee replacement, readmission, osteoarthritis, rheumatoid arthritis, arthroplasty, joint replacement, diagnosis, risk factor
Introduction
Total joint arthroplasty is an elective, efficacious, surgical option for patients with end-stage arthritis (1). Hospital readmission after arthroplasty is an undesired event, with 90-day readmission rates ranging from 7-8% after primary total hip or total knee arthroplasty (THA, TKA) (2-4). Post-arthroplasty complications, such as surgical site infections, procedure-related complications, sepsis, thromboembolic and cardiovascular complications are responsible for the majority of readmissions (2, 5, 6). Osteoarthritis (OA) and rheumatoid arthritis (RA) are the two most common reasons for patients to undergo TKA or THA. Recent studies have reported that perioperative complications after THA and TKA were higher in patients with an underlying diagnosis of RA, compared to OA (7-9). Biologics, a new class of medication, are now used commonly for the treatment of RA, which may increase the risk of skin and soft tissue infections and thus impact the readmission rates in RA (10, 11). Given the differences in disease pathophysiology and treatment, an important question is whether underlying diagnosis is a risk factor for post-arthroplasty readmissions.
To our knowledge, only one Taiwanese study, including patients who under went TKA between 2002 and 2004, examined if the underlying diagnosis for arthroplasty was a predictor of hospital readmission after index arthroplasty. In that study, 768 patients with RA undergoing TKA had higher 90-day readmission (14.2% vs.11.3%) compared to 3,840 age- and gender-matched patients with OA (12). There is a lack of contemporary data from U.S. or Europe in this area, which are needed, given the differences in rates of treatment with biologics, baseline infection rates and the risk of various infections between different populations (13-15). Our main study objectives were to use a representative U.S. sample to assess (1) whether RA was an independent risk factor for 90-day readmission rate after primary THA or TKA, (2) the causes for readmission and (3) whether the causes of readmission differed between RA and OA.
Methods
Study methods and results are described as recommended in the Strengthening of Reporting in Observational studies in Epidemiology (STROBE) statement (16).
Study Design, Data Source, and Patient Sample
A retrospective analysis of a cohort of patients that underwent THA or TKA between 01/01/2009 and 12/31/2011 was conducted. Patients were identified using an integrated healthcare system Total Joint Replacement Registry (TJRR). The healthcare system covers over 9 million members in 7 US geographical areas. Details on the TJRR data collection procedures, structure, and participation have been previously reported (17, 18). In brief, the TJRR collects information at the point of care by surgeons or other dedicated providers using both paper and electronic data capture tools. Additionally, TJRR data are collected from other databases (including the electronic medical records) using electronic screening algorithms. The data repository for the registry is a SQL database and information from different data sources of the registry is linked using the institution's unique patient identifiers. Quality control of the registry's data is conducted quarterly and participation reports are provided to all locations on a quarterly basis as well. All TJRR outcomes (i.e. readmissions, revision, infection, thromboembolism, mortality) are captured prospectively and additionally the outcomes of revision, infection and thromboembolic events are adjudicated via chart review by trained clinical research associates (17, 18). In a chart review of a random sample of RA and OA cases (n=60) performed by an abstractor blinded to the database diagnosis, we found the true positive rate of diagnoses to be 88%- it was 77% for RA and 100% for OA. All registered cases are tracked until outcome and/or end of their lives. The voluntary participation of the TJRR in 2011 was 95% for both hip and knee procedures (18).
This study included all unilateral primary THA and TKA procedures, in patients aged 18 years old or older, with surgery indication for OA and/or RA registered during the study period (N=34311). Revision, partial, or conversion procedures were not included in the study. Individuals who died or terminated membership within 90 days of the original operative date without a readmission during the same time frame were considered lost to follow-up (n=90, 0.3%). Cases from the geographical regions where the organization owns the hospitals (Southern California, Northern California, Hawaii), covering 90% of the TJRR registered cases, were included The final sample included cases from 31 hospitals and 204 surgeons.
Outcome Measure
The study outcome was 90-day readmission. Readmission was obtained from the TJRR, which monitors it using quarterly extracts from the institutional electronic medical records of all hospital readmissions post discharge from the original joint arthroplasty procedure.
Exposure of Interest
The underlying diagnosis for TKA or THA was the exposure of interest. This was obtained from the TJRR intra-operative forms completed by the surgeon at the time of surgery and does not rely on administrative data coding schemes. If both RA and OA were checked (n=109), cases were considered to have RA.
Covariates
Operative joint (THA or TKA), gender, age (categorized into <65 and ≥65 years old), and operative year of the procedure (2009 vs. 2010 vs. 2011) were evaluated as effect modifiers of the RA and 90-day readmission outcome association. Patient, procedure, surgeon, and hospital variables were evaluated as possible confounders. Patient variables included: age (continuous); gender; race (White, Black, Hispanic, Asian, Other); body mass index (BMI, categorized into <30 kg/m2, 30-34 kg/m2, and ≥35 kg/m2); American Society of Anesthesiologists (ASA) class (<3 vs. ≥3); and medical comorbidities using the Elixhauser co-morbidity measure (consists of 29 individual co-morbidities including congestive heart failure, deficiency anemia, etc.)(19); in hospital non-surgical complications (inclusive of pneumonia, acute myocardial infarction, and acute stroke) and in hospital surgical complications (inclusive of surgery and implant related complications). The Agency for Healthcare Research and Quality (AHRQ) Inpatient Quality Indicators Technical Specifications were used to identify the non-surgical complications (20). Procedure variables evaluated included: discharge disposition (home vs. other) and length of stay (LOS, categorized into ≤2, 3, and ≥4 days). Surgeon variables included whether they had a total joint arthroplasty fellowship and average yearly volume (categorized into low volume ≤20 cases/year, medium volume 20-49 cases/year, high volume ≥50 cases/year). Hospital volume was also evaluated (categorized into low volume <100 cases/year, 100-199 cases/year, and ≥200 cases/year). Average surgeon and hospital yearly volumes were calculated using both primary and revision procedures performed during the calendar year.
Statistical Analyses
Means, standard deviations (SD), frequencies, and proportions were used to describe the study sample by whether the primary diagnosis was RA or OA. Categorical variables compared between the two groups using 2-sided chi-square or Fisher's exact tests where appropriate. A logistic regression using a generalized linear model was used to model the relationship between RA (OA being the reference category) and 90 day readmission (yes vs. no) while accounting for the nesting of observations within the hospital variable. Year of operation was found to be an effect modifier of the relationship studied and analysis were therefore stratified by year of operation. Variables were included in the final adjusted model if they were found to be confounders (changed risk estimates by >10% and p<0.10) or deemed clinically important for the relationship studied. Missing data were minimal in the current dataset and cases with missing data (N=267) were excluded from the final model. Sensitivity analyses were conducted, using a model with 10 multiple imputations to address the missingness of the data. Odds ratios (OR), 95% confidence internals (CI), and p-values based on a Wald test are reported. Data were analyzed using SAS (Version 9.2, SAS Institute, Cary, NC, USA) and p-value <0.05 was considered statistical significant.
Results
Demographic and Clinical Characteristics of Study Sample
In this sample there are 34,311 procedures, 33,815 for OA and 496 for RA. In the combined, OA and RA samples, there were: 61%, 61% and 73% women; 69%, 70% and 45% White; 21%, 21% and 19% patients with BMI ≥35 kg/m2; 41%, 40% and 50% with ASA class 3 or higher (Table 1). The mean age of OA patients was higher than RA patients, 67 (SD=9.8) vs. 61 (SD=13.5) years old, p<0.001.
Table 1. Study Sample Description, Overall and by Primary Indication for Surgery, 2009-2011.
| Total (N=34311) | Osteoarthritis (N=33815) | Rheumatoid Arthritis (N=496) | P Value | |||||
|---|---|---|---|---|---|---|---|---|
|
| ||||||||
| Total N | N | % | N | % | N | % | ||
| TJA | THA | 11615 | 33.9 | 11448 | 33.9 | 167 | 33.7 | 0.931 |
| TKA | 22696 | 66.2 | 22367 | 66.2 | 329 | 66.3 | ||
| Operative year | 2009 | 10389 | 30.3 | 10233 | 30.3 | 156 | 31.5 | 0.316 |
| 2010 | 11751 | 34.3 | 11571 | 34.2 | 180 | 36.3 | ||
| 2011 | 12171 | 35.5 | 12011 | 35.5 | 160 | 32.3 | ||
| Age category, years | <65 | 13516 | 39.4 | 13237 | 39.2 | 279 | 56.3 | <0.001 |
| ≥65 | 20795 | 60.6 | 20578 | 60.9 | 217 | 43.8 | ||
| Female gender | 21057 | 61.4 | 20694 | 61.2 | 363 | 73.2 | <0.001 | |
| Race1 | White | 23765 | 69.3 | 23542 | 69.6 | 223 | 45.0 | <0.001 |
| Hispanic | 4601 | 13.4 | 4455 | 13.2 | 146 | 29.4 | ||
| Black | 3115 | 9.1 | 3049 | 9.0 | 66 | 13.3 | ||
| Asian | 2168 | 6.3 | 2114 | 6.3 | 54 | 10.9 | ||
| Other/Multi | 615 | 1.8 | 609 | 1.8 | 6 | 1.2 | ||
| BMI category1, kg/m2 | <30 | 17363 | 50.6 | 17071 | 50.5 | 292 | 58.9 | <0.001 |
| 30-34 | 9648 | 28.1 | 9537 | 28.2 | 111 | 22.4 | ||
| ≥35 | 7288 | 21.2 | 7195 | 21.3 | 93 | 18.8 | ||
| ASA category1 | 1-2 | 20163 | 58.8 | 19915 | 58.9 | 248 | 50.0 | <0.001 |
| ≥3 | 13920 | 40.6 | 13674 | 40.4 | 246 | 49.6 | ||
| Diabetes | 8220 | 24.0 | 8125 | 24.0 | 95 | 19.2 | 0.012 | |
| Deficiency anemias* | 4238 | 12.4 | 4115 | 12.2 | 123 | 24.8 | <0.001 | |
| Surgeon TJA fellowship | 13593 | 39.6 | 13362 | 39.5 | 231 | 46.6 | 0.001 | |
| Hospital yearly volume, cases/yr | <100 | 1830 | 5.3 | 1807 | 5.3 | 23 | 4.6 | 0.637 |
| 100-199 | 13588 | 39.6 | 13397 | 39.6 | 191 | 38.5 | ||
| ≥200 | 18893 | 55.1 | 18611 | 55.0 | 282 | 56.9 | ||
| Surgeon yearly volume, cases/yr | <20 | 3159 | 9.2 | 3114 | 9.2 | 45 | 9.1 | 0.634 |
| 20-49 | 14507 | 42.3 | 14287 | 42.3 | 220 | 44.4 | ||
| ≥50 | 16645 | 48.5 | 16414 | 48.5 | 231 | 46.6 | ||
| Discharge disposition1 | Home | 27020 | 78.8 | 26629 | 78.8 | 391 | 78.8 | 0.968 |
| Other | 7216 | 21.0 | 7112 | 21.0 | 104 | 21.0 | ||
| LOS categories, days1 | ≤2 | 13054 | 38.1 | 12888 | 38.1 | 166 | 33.5 | 0.216 |
| 3 | 15795 | 46.0 | 15548 | 46.0 | 247 | 49.8 | ||
| 4 | 3510 | 10.2 | 3457 | 10.2 | 53 | 10.7 | ||
| ≥5 | 1880 | 5.5 | 1852 | 5.5 | 28 | 5.7 | ||
| Intra-operative complications | Surgical | 1497 | 4.4 | 1476 | 4.4 | 21 | 4.2 | 0.887 |
| Non-surgical | 281 | 0.8 | 277 | 0.8 | 4 | 0.8 | 0.975 | |
Missing data: race (n=47, <0.1%), BMI (n=12, <0.1%), ASA (n=228, <0.7%), discharge disposition (n=71, 0.2%), LOS (n=72, 0.2%)
TJA=Total Joint Arthroplasty. THA=Total Hip Arthroplasty. TKA= Total Knee Arthroplasty. BMI=Body Mass Index. ASA=American Society of Anesthesiologists. LOS=Length of Stay.
Deficiency Anemia: one of the comorbidities from the Elixhauser's index, included here due to the level of significance in analyses. Elixhauser's index includes 29 comorbidities, as follows: Hypertension, chronic pulmonary disease, hypothyroidism, deficiency anemias, renal failure, psychoses, fluid and electrolyte disorders, depression, peripheral vascular disease, rheumatoid arthritis/collagen vascular disorders, valvular disease, other neurological disorders, chronic blood loss anemia, congestive heart failure, liver disease, coagulopathy, alcohol abuse, solid tumor w/out metastasis, pulmonary circulation disease, drug abuse, weight loss, paralysis, lymphoma, metastatic cancer, acquired immune deficiency syndrome, peptic ulcer disease with bleeding, diabetes* (without complications), diabetes* (with complications), and obesity*. *These 3 conditions were not used from the algorithm due to availability of more robust data sources for these measures: diabetes status is available from regional Kaiser Permanente diabetes registries and obesity from the patients' electronic medical records body mass index (BMI) assessments.
RA patients had higher rates of multiple co-morbidities than the OA patients (Appendix 1; p<0.001). The distribution of co-morbidities is shown in Appendix 1. Deficiency anemias (p<0.001) were more common in RA patients than OA patients.
Crude Readmission Rates
The crude 90-day readmission rates for patients with OA and RA diagnoses were 6.7% and 8.5% respectively (p=0.127), with an overall combined 90-day readmission rate of 6.8%. There was an increasing trend in the incidence of 90-day readmissions in the RA patients by year, 5.8%, 8.9% and 10.6%, for 2009, 2010, and 2011, respectively (Table 2). For OA, the 90-day readmission rates were similar by year with 6.7%, 6.7% and 6.8% for 2009, 2010 and 2011, respectively.
Table 2. 90-day Readmission Rate for Overall Sample, by Primary Indication for Surgery, and by Operative Year.
| Total (N=34311) | Osteoarthritis (N=33815) | Rheumatoid Arthritis (N=496) | P Value | |
|---|---|---|---|---|
|
| ||||
| N (%) | N (%) | N (%) | ||
| Total Operative year | 2319 (6.8) | 2277 (6.7) | 42 (8.5) | 0.127 |
| 2009 | 734 (7.1) | 687 (6.7) | 9 (5.8) | 0.640 |
| 2010 | 866 (7.4) | 777 (6.7) | 16 (8.9) | 0.249 |
| 2011 | 916 (7.5) | 813 (6.8) | 17 (10.6) | 0.055 |
Crude and Adjusted Odds of 90-day Readmission
Crude odds of 90-day readmission were 1.29 (95% CI, 0.93-1.79) times higher in patients with RA compared to those with OA in the model combining data from years 2009 to 2011 (Table 3). Compared to OA, the odds of 90-day post-arthroplasty readmission in RA were 0.85 (95% CI 0.45-1.59) in 2009, 1.37 (95% CI, 0.73-2.57) in 2010 and 1.63 (95% CI, 1.09-2.44) in 2011. In the age, gender, ASA, and deficiency anemia adjusted model, the risk of 90-day readmission was not significantly different for RA vs. OA for the overall sample (OR=1.33; 95% CI, 0.94-1.87).
Table 3.
Crude and Adjusted Generalized Estimating Equation Models for the Odds of Patients Undergoing Total Joint Arthroplasty due to Rheumatoid Arthritis being Readmitted within 90 days Compared to those with Osteoarthritis, Overall and by Operative Year. Odds Ratio (OR), 95% Confidence Intervals (CI)
| Models | Crude OR (95%CI) | Adjusted OR (95%CI)1 |
|---|---|---|
| All years | 1.29 (0.93-1.79) | 1.33 (0.94-1.87) |
| 2009 | 0.85 (0.45-1.59) | 0.89 (0.46-1.71) |
| 2010 | 1.37 (0.73-2.57) | 1.34 (0.69-2.61) |
| 2011 | 1.63 (1.09-2.44) | 1.74 (1.16-2.60) |
Adjusted for age, gender, deficiency anemia, and ASA category.
OR=Odds Ratio. CI= Confidence Interval.
Because of this difference in risk by the year (effect modification), we ran additional models to accommodate this year and readmission risk interaction. No other effect modification (by age, gender, and operative joint) was observed. The adjusted risk of 90-day readmission increased from year to year for RA patients compared to OA patients. Compared to OA, the risk for 90-day readmission in RA patients was 0.89 (95% CI, 0.46-1.71) in 2009, 1.34 (95% CI, 0.69-2.61) in 2010 and 1.74 (95% CI, 1.16-2.60) in 2011. Models with multiple imputations to address missing data (0.7% of all), yielded consistent estimations, with the ORs being 0.88 (95% CI, 0.44-1.74), 1.33 (95% CI, 0.78-2.25), and 1.75 (95% CI, 1.04-2.92), respectively.
We further examined various patient characteristics and whether they changed over time (Table 4). There were minimal changes in patient characteristics over the 3-year study period, except slight increase in BMI and age in RA patients over the years. Additional adjustment for BMI in the multivariable-adjusted models did not change the significance of association of RA with 90-day readmission and had minimal/negligible change in point estimates.
Table 4. Comparison of characteristics for rheumatoid arthritis and osteoarthritis patients by calendar year.
| Year | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
|
|
|
||||||||||||
| 2009 | 2010 | 2011 | |||||||||||
| OA | RA | OA | RA | OA | RA | ||||||||
| N | (%) | N | (%) | N | (%) | N | (%) | N | (%) | N | (%) | ||
| All | 10233 | 98.5 | 156 | 1.5 | 11571 | 98.5 | 180 | 1.5 | 12011 | 98.7 | 160 | 1.3 | |
| TJA | THA | 3561 | 34.8 | 47 | 30.1 | 3831 | 33.1 | 65 | 36.1 | 4056 | 33.8 | 55 | 34.4 |
| TKA | 6672 | 65.2 | 109 | 69.9 | 7740 | 66.9 | 115 | 63.9 | 7955 | 66.2 | 105 | 65.6 | |
| Age category, years | <65 | 4067 | 39.7 | 84 | 53.9 | 4435 | 38.3 | 104 | 57.8 | 4735 | 39.4 | 91 | 56.9 |
| >=65 | 6166 | 60.3 | 72 | 46.2 | 7136 | 61.7 | 76 | 42.2 | 7276 | 60.6 | 69 | 43.1 | |
| Female gender | Female | 6274 | 61.3 | 110 | 70.5 | 7084 | 61.2 | 135 | 75.0 | 7336 | 61.1 | 118 | 73.8 |
| Race1 | White | 7260 | 71.0 | 75 | 48.1 | 7997 | 69.1 | 78 | 43.3 | 8285 | 69.0 | 70 | 43.8 |
| Hispanic | 1266 | 12.4 | 39 | 25.0 | 1476 | 12.8 | 54 | 30.0 | 1713 | 14.3 | 53 | 33.1 | |
| Black | 894 | 8.7 | 22 | 14.1 | 1090 | 9.4 | 22 | 12.2 | 1065 | 8.9 | 22 | 13.8 | |
| Asian | 585 | 5.7 | 15 | 9.6 | 784 | 6.8 | 25 | 13.9 | 745 | 6.2 | 14 | 8.8 | |
| Other/Multi | 211 | 2.1 | 4 | 2.6 | 208 | 1.8 | 1 | 0.6 | 190 | 1.6 | 1 | 0.6 | |
| BMI category (kg/m2)1 | <30 | 5109 | 49.9 | 105 | 67.3 | 5818 | 50.3 | 97 | 53.9 | 6144 | 51.2 | 90 | 56.3 |
| 30-34 | 2909 | 28.4 | 23 | 14.7 | 3245 | 28.0 | 50 | 27.8 | 3383 | 28.2 | 38 | 23.8 | |
| ≥35 | 2208 | 21.6 | 28 | 18.0 | 2507 | 21.7 | 33 | 18.3 | 2480 | 20.7 | 32 | 20.0 | |
| ASA category | 1&2 | 6023 | 58.9 | 85 | 54.5 | 6825 | 59.0 | 82 | 45.6 | 7067 | 58.8 | 81 | 50.6 |
| >=3 | 4145 | 40.5 | 70 | 44.9 | 4655 | 40.2 | 97 | 53.9 | 4874 | 40.6 | 79 | 49.4 | |
| Diabetes | 2534 | 24.8 | 31 | 19.9 | 2784 | 24.1 | 36 | 20.0 | 2809 | 23.4 | 28 | 17.5 | |
| Deficiency Anemia | 1358 | 13.3 | 34 | 21.8 | 1366 | 11.8 | 47 | 26.1 | 1391 | 11.6 | 42 | 26.3 | |
| LOS | <=2 day | 3035 | 29.7 | 39 | 25.0 | 4397 | 38.0 | 56 | 31.1 | 5456 | 45.4 | 71 | 44.4 |
| 3 day | 5209 | 50.9 | 86 | 55.1 | 5339 | 46.1 | 89 | 49.4 | 5000 | 41.6 | 72 | 45.0 | |
| 4 day | 1276 | 12.5 | 19 | 12.2 | 1162 | 10.0 | 23 | 12.8 | 1019 | 8.5 | 11 | 6.9 | |
| >=5 day | 673 | 6.6 | 11 | 7.1 | 655 | 5.7 | 11 | 6.1 | 524 | 4.4 | 6 | 3.8 | |
| Surgeon TJA Fellowship | Fellowship | 3907 | 38.2 | 78 | 50.0 | 4609 | 39.8 | 76 | 42.2 | 4846 | 40.4 | 77 | 48.1 |
| Hospital yearly volume, cases/yr | <100 | 515 | 5.0 | 8 | 5.1 | 583 | 5.0 | 6 | 3.3 | 709 | 5.9 | 9 | 5.6 |
| 100-199 | 3928 | 38.4 | 53 | 34.0 | 4597 | 39.7 | 75 | 41.7 | 4872 | 40.6 | 63 | 39.4 | |
| >=200 | 5790 | 56.6 | 95 | 60.9 | 6391 | 55.2 | 99 | 55.0 | 6430 | 53.5 | 88 | 55.0 | |
| Surgeon yearly volume, cases/yr | <20 | 916 | 9.0 | 14 | 9.0 | 1061 | 9.2 | 16 | 8.9 | 1137 | 9.5 | 15 | 9.4 |
| 20-49 | 4372 | 42.7 | 61 | 39.1 | 4810 | 41.6 | 86 | 47.8 | 5105 | 42.5 | 73 | 45.6 | |
| >=50 | 4945 | 48.3 | 81 | 51.9 | 5700 | 49.3 | 78 | 43.3 | 5769 | 48.0 | 72 | 45.0 | |
| Discharge disposition | Home | 8074 | 78.9 | 121 | 77.6 | 9017 | 77.9 | 144 | 80.0 | 9539 | 79.4 | 126 | 78.8 |
| Other | 2125 | 20.8 | 34 | 21.8 | 2535 | 21.9 | 36 | 20.0 | 2455 | 20.4 | 34 | 21.3 | |
| Intra-operative complications | Surgical | 488 | 4.8 | 7 | 4.5 | 507 | 4.4 | 10 | 5.6 | 481 | 4.0 | 4 | 2.5 |
| Non Surgical | 85 | 0.8 | 3 | 1.9 | 100 | 0.9 | 1 | 0.6 | 92 | 0.8 | 0 | 0.0 | |
Missing data: race (n=47, <0.1%), BMI (n=12, <0.1%), ASA (n=228, 0.7%), discharge disposition (n=71, 0.2%), LOS (n=72, 0.2%).
OA=Osteoarthritis. RA=Rheumatoid Arthritis. TJA=Total Joint Arthroplasty. THA=Total Hip Arthroplasty. TKA= Total Knee Arthroplasty. BMI=Body Mass Index. ASA=American Society of Anesthesiologists. LOS=Length of Stay.
Reasons for Readmission
The 42 RA patients in our sample with readmissions were readmitted 49 times within the 90 days period. The top 5 principal discharge ICD 9 codes used for their readmissions were: 996.66 (Infection and inflammatory reaction due to internal joint prosthesis, 10.2%), 038.9 (Unspecified septicemia, 10.2%), 715.36 (osteoarthrosis, localized, not specified whether primary or secondary, lower leg, 6.1%), 996.77 (other complications due to internal joint prosthesis, 6.1%), and 998.59 (other postoperative infection, 6.1%) The 2,277 OA patients were readmitted 2,703 times within the 90 days period. The top 5 principal discharge ICD 9 codes used for their readmissions were: 996.66 (infection and inflammatory reaction due to internal joint prosthesis, 5.7%), 998.59 (other postoperative infection, 5.1%), 715.36 (osteoarthrosis, localized, not specified whether primary or secondary, lower leg, 3.6%), 038.9 (unspecified septicemia, 3.5%), and 415.19 (other pulmonary embolism and infarction, 2.7%). The re-admissions with ICD 9 codes 715.36 were usually for TKA for the contralateral knee. In a chart review of a random sample of 15 patients readmitted with 715.36, including 2 THA and 13 TKA patients, TKA for the treatment of knee OA on the contralateral limb was the reason in each case.
Discussion
In this study with a large representative U.S. cohort, we studied patients undergoing primary TKA or THA, the two most common arthroplasty procedures in the U.S. accounting for >1 million procedures annually (21).A 90-day readmission rate of 6.8% TKA/THA after translates to >70,000 admissions annually in the U.S, with at least some being preventable. This rate is consistent with 90-day readmission rates of 7.8-8% after primary TKA (2, 3) and 7% after primary THA in previous studies (4). Thus, readmission after an elective THA/TKA is a problem of significant public health proportions. Several observations are new that deserve further discussion.
We found that RA was a risk factor for 90-day readmission rate after primary TKA or THA in 2011, the most recent study year. Compared to OA, RA was associated with higher adjusted odds of 1.74 of 90-day readmission in 2011. We also noted that the 90-day readmission increased during the study period from 2009 and 2011. To our knowledge, there is only one Taiwanese study that reported that compared to OA, RA was associated with a 37% higher 90-day readmission rate (12). Our finding of increasing 90-day readmission over 3-year period 2009-11 in RA but not in OA is of concern. In a previous study using a cohort of Medicare patients, the 30-day all-cause readmission after primary THA increased from 5.9% to 8.5% from 1991 to 2008 (22). Patients discharged to SNFs had higher odds of hospital readmission within 90 days of surgery than those discharged home (23). These findings indicate an increasing readmission rate overall after THA/TKA over time, particularly in RA patients. We considered several patient, procedure, surgeon and hospital variables as important covariates, and adjusted for those that were significant (age, gender, ASA and iron-deficiency anemia) in our multivariable-adjusted model, indicating that the increasing readmission rate in RA patients is not explained by these variables. Due to limited availability of information on various RA and OA treatments (such as medication regimen) and pre and post-operative rehabilitation regimens, we were unable to analyze the impact of these factors on readmissions. These could be factors mediating the increase in readmissions in RA patients, which needs to be addressed in future studies designed to investigate this question. Whether some readmissions are preventable depends on the underlying reason for readmission. It remains to be examined if pre- or post-operative optimization of RA management can reduce post-THA/TKA readmissions. Slight increases in age and BMI may have contributed to the noted high 90-day re-admission rates, a hypothesis that needs to be tested in future studies with an adequate sample of RA patients. Given that preventive strategies are generally low cost and readmission is expensive, such strategies are likely to be cost-effective and may even be cost saving, a hypothesis we propose for testing in a future study.
There was a significant overlap in the reasons for 90-day readmissions. However the main two reasons were slightly different for RA compared to OA. Joint prosthesis infection and unspecified septicemia were seen in RA and joint prosthesis infection and other postoperative infection, in OA. These reasons for readmission are similar to those reported in the other studies and do not separate RA from OA, particularly (5).
Our study findings must be considered considering study strengths and limitations. Study strengths include a large and representative sample, prospectively collected data, complete capture of outcomes (readmissions) due to an integrated system, and availability of clinically significant confounders for our analyses. Our study has several limitations. Due to effect modification by the year of study, our inferences are specific to the calendar year studied. Additionally, residual confounding is possible, given the observational study design. A limited number of patients did not allow analyses of THA and TKA patients separately; however, we had specified a priori that we would combine these two populations for the main analysis.
In conclusion, we found that RA was associated with higher likelihood of 90-day readmission and that the association was stronger in the most recent year i.e. 2011, compared to 2009. Our study described the reasons for 90-day readmissions separately for RA and OA. Future studies should explore whether specific medications used for the treatment of RA or OA are associated with higher complication rate and readmission rate and whether interventions targeting perioperative management OA or RA may decrease 90-day readmissions.
Supplementary Material
Significance and Innovation.
90-day post-arthroplasty readmission risk after THA or TKA is higher in patients with RA compared to OA.
We noted a significant increase in 90-day readmission from 2009 to 2011 in RA patients who underwent THA or TKA indicating that we need further studies to understand the underlying causes and design interventions to mitigate this higher risk
Acknowledgments
IRB approval: The Kaiser Permanente Institutional Review Board (#5488) approved this study and all investigations were conducted in conformity with ethical principles of research.
Grant support: No grants were obtained for this study. JAS is supported by grants from the Agency for Health Quality and Research Center for Education and Research on Therapeutics (AHRQ CERTs) U19 HS021110, National Institute of Arthritis, Musculoskeletal and Skin Diseases (NIAMS) P50 AR060772 and U34 AR062891, National Institute of Aging (NIA) U01 AG018947, National Cancer Institute (NCI) U10 CA149950, the resources and the use of facilities at the VA Medical Center at Birmingham, Alabama and research contract CE-1304-6631 from the Patient Centered Outcomes Research Institute (PCORI). None of the funding agencies played any role in study design, analyses, write-up or the decision to submit it for publication.
Footnotes
Financial Conflict: There are no financial conflicts related directly to this study. J.A.S. has received research and travel grants from Takeda and Savient; and consultant fees from Savient, Takeda, Regeneron and Allergan. Other authors declare no conflicts.
Author Contributions: Drs. Singh, Inacio and Paxton conceived the study idea and discussed it. Dr. Singh developed the protocol that was reviewed and edited by Drs. Inacio, Paxton and Namba. Dr. Inacio performed data analyses. All authors reviewed and interpreted the data. Dr. Singh wrote the first draft of the paper, which was revised by all authors. All authors provided permission to submit the final version of the paper.
Declaration: Dr. Singh (the manuscript's guarantor) affirms that the manuscript is an honest, accurate, and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned have been explained
The views expressed in this article are those of the authors and do not necessarily reflect the position or policy of the Department of Veterans Affairs or the United States government.
Data sharing: no additional data available.
References
- 1.Ethgen O, Bruyere O, Richy F, Dardennes C, Reginster JY. Health-related quality of life in total hip and total knee arthroplasty. A qualitative and systematic review of the literature. J Bone Joint Surg Am. 2004;86-A:963–74. doi: 10.2106/00004623-200405000-00012. [DOI] [PubMed] [Google Scholar]
- 2.Saucedo JM, Marecek GS, Wanke TR, Lee J, Stulberg SD, Puri L. Understanding Readmission After Primary Total Hip and Knee Arthroplasty: Who's at Risk? J Arthroplasty. 2013 doi: 10.1016/j.arth.2013.06.003. [DOI] [PubMed] [Google Scholar]
- 3.Schairer WW, Vail TP, Bozic KJ. What Are the Rates and Causes of Hospital Readmission After Total Knee Arthroplasty? Clin Orthop Relat Res. 2013 doi: 10.1007/s11999-013-3030-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Schairer WW, Sing DC, Vail TP, Bozic KJ. Causes and Frequency of Unplanned Hospital Readmission After Total Hip Arthroplasty. Clin Orthop Relat Res. 2013 doi: 10.1007/s11999-013-3121-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Avram V, Petruccelli D, Winemaker M, de Beer J. Total Joint Arthroplasty Readmission Rates and Reasons for 30-Day Hospital Readmission. J Arthroplasty. 2013 doi: 10.1016/j.arth.2013.07.039. [DOI] [PubMed] [Google Scholar]
- 6.Pugely AJ, Callaghan JJ, Martin CT, Cram P, Gao Y. Incidence of and risk factors for 30-day readmission following elective primary total joint arthroplasty: analysis from the ACS-NSQIP. J Arthroplasty. 2013;28:1499–504. doi: 10.1016/j.arth.2013.06.032. [DOI] [PubMed] [Google Scholar]
- 7.Ravi B, Croxford R, Hollands S, Paterson MJ, Bogoch E, Kreder H, et al. Patients with rheumatoid arthritis are at increased risk for complications following total joint arthroplasty. Arthritis Rheum. 2013 doi: 10.1002/art.38231. [DOI] [PubMed] [Google Scholar]
- 8.Stundner O, Chiu YL, Sun X, Goodman SM, Russell LA, Calloway JJ, et al. Perioperative outcomes in patients with rheumatoid versus osteoarthritis for total hip arthroplasty: a population-based study. Clin Exp Rheumatol. 2013;31:889–95. [PubMed] [Google Scholar]
- 9.Stundner O, Danninger T, Chiu YL, Sun X, Goodman SM, Russell LA, et al. Rheumatoid Arthritis vs Osteoarthritis in Patients Receiving Total Knee Arthroplasty: Perioperative Outcomes. J Arthroplasty. 2013 doi: 10.1016/j.arth.2013.05.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Dixon WG, Watson K, Lunt M, Hyrich KL, Silman AJ, Symmons DP. Rates of serious infection, including site-specific and bacterial intracellular infection, in rheumatoid arthritis patients receiving anti-tumor necrosis factor therapy: results from the British Society for Rheumatology Biologics Register. Arthritis Rheum. 2006;54:2368–76. doi: 10.1002/art.21978. [DOI] [PubMed] [Google Scholar]
- 11.Wallis RS. Biologics and infections: lessons from tumor necrosis factor blocking agents. Infectious disease clinics of North America. 2011;25:895–910. doi: 10.1016/j.idc.2011.08.002. [DOI] [PubMed] [Google Scholar]
- 12.Kang JH, Hsieh MS, Lin HC. Comparison of treatment outcomes following total knee arthroplasty among patients with rheumatoid arthritis and osteoarthritis: a nationwide population-based study. Rheumatology (Oxford) 2010;49:1409–10. doi: 10.1093/rheumatology/keq054. [DOI] [PubMed] [Google Scholar]
- 13.Dixon WG, Hyrich KL, Watson KD, Lunt M, Galloway J, Ustianowski A, et al. Drug-specific risk of tuberculosis in patients with rheumatoid arthritis treated with anti-TNF therapy: results from the British Society for Rheumatology Biologics Register (BSRBR) Ann Rheum Dis. 2010;69:522–8. doi: 10.1136/ard.2009.118935. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Schiff MH, Burmester GR, Kent JD, Pangan AL, Kupper H, Fitzpatrick SB, et al. Safety analyses of adalimumab (HUMIRA) in global clinical trials and US postmarketing surveillance of patients with rheumatoid arthritis. Ann Rheum Dis. 2006;65:889–94. doi: 10.1136/ard.2005.043166. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Sokka T, Kautiainen H, Toloza S, Makinen H, Verstappen SM, Lund Hetland M, et al. QUEST-RA: quantitative clinical assessment of patients with rheumatoid arthritis seen in standard rheumatology care in 15 countries. Ann Rheum Dis. 2007;66:1491–6. doi: 10.1136/ard.2006.069252. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Bern, Germany: University of Bern; 2007. STROBE Statement. Strengthening the reporting of observational studies in epidemiology. Link: http://www.strobe-statement.org/fileadmin/Strobe/uploads/checklists/STROBE_checklist_v4_cohort.pdf. 09/06/2012. [DOI] [PubMed] [Google Scholar]
- 17.Paxton EW, Inacio MC, Khatod M, Yue EJ, Namba RS. Kaiser Permanente National Total Joint Replacement Registry: aligning operations with information technology. Clin Orthop Relat Res. 2010;468:2646–63. doi: 10.1007/s11999-010-1463-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Paxton EW, Kiley ML, Love R, Barber TC, Funahashi TT, Inacio MC. Kaiser Permanente implant registries benefit patient safety, quality improvement, cost-effectiveness. Joint Commission journal on quality and patient safety / Joint Commission Resources. 2013;39:246–52. doi: 10.1016/s1553-7250(13)39033-3. [DOI] [PubMed] [Google Scholar]
- 19.Elixhauser A, Steiner C, Harris DR, Coffey RM. Comorbidity measures for use with administrative data. Med Care. 1998;36:8–27. doi: 10.1097/00005650-199801000-00004. [DOI] [PubMed] [Google Scholar]
- 20.Agency for Healthcare Research and Quality. Patient Safety Indicators Technical Specifications. 2013 http://www.qualityindicators.ahrq.gov/Downloads/Modules/PSI/V45/TechSpecs/PSI12PerioperativePulmonaryEmbolismorDeepVeinThrombosisRate.pdf. 02/08/2014.
- 21.HCUP. Atlanta, GA: Centers for Disease Control and Prevention; 2013. Number of all-listed procedures for discharges from short-stay hospitals, by procedure category and age: United States, 2010. http://www.cdc.gov/nchs/data/nhds/4procedures/2010pro4_numberprocedureage.pdf. 10/16/2013. [Google Scholar]
- 22.Cram P, Lu X, Kaboli PJ, Vaughan-Sarrazin MS, Cai X, Wolf BR, et al. Clinical characteristics and outcomes of Medicare patients undergoing total hip arthroplasty, 1991-2008. JAMA. 2011;305:1560–7. doi: 10.1001/jama.2011.478. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Bini SA, Fithian DC, Paxton LW, Khatod MX, Inacio MC, Namba RS. Does discharge disposition after primary total joint arthroplasty affect readmission rates? J Arthroplasty. 2010;25:114–7. doi: 10.1016/j.arth.2008.11.007. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.