What Are the 30-day Readmission Rates Across Orthopaedic Subspecialties?

The Centers for Medicare & Medicaid Services (CMS) now include hip and knee replacements in the Hospital Readmission Reduction Program. The 30-day readmission rate is an important quality metric; however, the incidence has not yet been defined across the numerous orthopaedic subspecialties. Elucidating the readmission rate for each subspecialty may indicate that certain services are being disincentivized by the CMS reimbursement program. Furthermore, the "planned" and "unplanned" definitions of readmission have not been well examined to determine their clinical relevance and representation of safe patient care. Therefore, reducing the 30-day readmission rate has become a top priority in orthopaedic quality assurance.

(1) What are the 30-day readmission rates for the different orthopaedic subspecialties? (2) What are the risk factors associated with readmission within 30 days? (3) What are the causes of 30-day readmissions? (4) What is the interrater agreement among CMS, hospital, and clinician definitions of planned and unplanned readmissions?

We retrospectively examined one tertiary care academic hospital's quality improvement database and identified 4792 discharges from the department of orthopaedics during a continuous 24-month period. Discharges were divided and analyzed according to the subspecialty of orthopaedic care. Demographics and comorbidities were extracted from the database and subjected to univariate and multivariate analysis to determine risk factors for 30-day readmission. Further chart review was conducted on all cases of 30-day readmission to identify causes. The authors' determination of planned versus unplanned was compared with two other definitions (hospital and CMS) and analyzed for agreement by using Fleiss' kappa for multiple rater.

The all-cause 30-day readmission rate was 4% (95% confidence interval [CI], 3.8-4.8). The unplanned readmission rate was 3% (95% CI, 2.8-3.8). After controlling for relevant confounding variables, we found that length of stay (odds ratio [OR], 1.10 per day; p < 0.001), American Society of Anesthesiologists score (OR, 1.89 per point; p < 0.001), and care under trauma (OR, 2.55; p < 0.001) or "other" (OR, 1.65; p = 0.009) as compared with joint subspecialty were associated with increased risk of readmission. Of the 160 unplanned readmissions, 93 (58%) were surgical and 67 (42%) were medical. The most common surgical cause was surgical site infection (38% of surgical readmissions) and the most common medical causes were gastrointestinal bleed, pulmonary embolus, and unrelated trauma (each 9% of medical readmissions). There was poor agreement (Fleiss' kappa = 0.120) among the three definitions of planned readmission.

There are important differences in the risk of readmission by subspecialty across orthopaedics and the CMS-driven disincentives may be applied unequally across these subspecialties. This could result in hospitals deemphasizing those service lines and could potentially limit access to care for the patients most in need. Avenues of readmission reduction should be further studied including telephone followup programs and outpatient management of threatened wounds. Clinical, hospital, and CMS definitions of planned readmission have poor agreement, suggesting that hospitals are being unnecessarily penalized. The CMS should develop a more clinically relevant definition of 30-day readmission to more accurately evaluate the rate of readmissions.

Level III, therapeutic study.