Background/Purpose: Systemic lupus erythematosus (SLE) is a chronic illness that has been associated with high hospital readmission rates. Based on a recently published (Arth Rheumatol (2014) 66:2828) electronic database analysis of Medicare/Medicaid hospital discharges, 16.5% of patients with the ICD-9-CM code 710.0 for SLE were readmitted to any hospital within thirty days. We propose a retrospective, observational study to assess the rate of and predictive factors influencing a thirty-day hospital readmission for adult patients with SLE.

Methods: All index admissions, defined as hospital admissions from adult patients (≥ 18 years of age) who had a primary or secondary ICD-9-CM diagnosis of 710.0 from October 1, 2012 to December 30, 2014, were discovered through data mining. Individual electronic patient charts from these admissions were manually reviewed to determine if the patient fulfilled the 1997 ACR revised criteria for SLE. Hospitalizations were excluded if they were scheduled admissions, resulted in a discharge to acute rehab, maternity related or resulted in death. The remaining hospitalizations were analyzed to determine if they resulted in a thirty-day readmission, using the same methods of determining thirty-day readmission counts performed in the benchmark study of readmission rates in Medicare’s fee-for-service program.

Results: 1003 index admissions (obtained from 433 unique patients) met our inclusion criteria. Only 196 of the 433 patients (45.3%) fulfilled criteria for SLE. The other 237 patients (54.7%) did not meet criteria: 185 patients had an inappropriate diagnosis of SLE (i.e., chart review contained sufficient data to support an alternate diagnosis), 34 patients had limited cutaneous lupus, and 18 patients had insufficient data to either confirm or refute a diagnosis of SLE. After application of exclusion criteria, 88 patients remained, accounting for 293 admissions, broken down into 221 index admissions (75.4%) and 72 thirty-day readmissions (24.6%). Of the 88 patients, 33 patients (37.5%) had ≥1 thirty-day readmission. Demographic data including age (p=0.92), gender (p=0.75), race/ethnicity (p=0.82), median household income by zip code (p=0.078) and primary expected payer (p=0.322) was compared between the groups. Of those with ≥1 thirty-day readmission, 11 were deceased (33.3%) compared to 3 deceased in those without a thirty-day readmission (5.5%) (p=0.0005).

Conclusion: We discovered a 37.5% rate of thirty-day readmission in patients meeting the 1997 revised criteria for the diagnosis of SLE. Among patients with ≥1 thirty-day readmission, a lower mean household income per zip code (p=0.078) was observed, approaching significance. All cause mortality was significantly increased in patients with SLE and ≥1 thirty-day readmission (p=0.0005). Additionally, our study revealed a 54.7% rate of miscoding of hospital records for adult patients with a primary or secondary diagnosis of SLE designated by the ICD-9-CM code 710.0, questioning the validity of using ICD-9-CM codes to identify patients with SLE.

« Back to 2016 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/thirty-day-readmission-rates-for-patients-diagnosed-with-systemic-lupus-erythematosus-analysis-at-a-tertiary-care-center/