Risk Factors for 30-Day Re-Admission in Adults with Sickle Cell Disease

Background: Re-admission to the hospital within 30 days is a measure for quality care and a point of emphasis for reducing health care costs in individuals with chronic disease. Potentially modifiable risk factors for 30 day re-admission in children with sickle cell disease (SCD) includes not being seen in the SCD clinic within 30-days of discharge (OR 7.7, 95% CI 2.4–24.4), 3 or more admissions in the previous 12 months (OR 7.3, 95% CI 2.8–18.9) and co-morbid asthma (OR 2.9, 95% CI 1.2–7.3) (Pediatr Blood Cancer. 2009 Apr;52(4):481-5). Limited data exists regarding potentially modifiable risk factors for 30-day re-admission in adults with SCD. The primary objective of this study was to define modifiable risk factors for 30-day re-admission in adults with SCD, leading to a prospective intervention study to decrease re-admission rates.

Procedure: At a tertiary care center, we performed a retrospective analysis of the medical records, from 2010 to 2013, to determine risk factors for 30-day re-admission in adults with SCD. Initial admission was defined as the first admission associated with uncomplicated vaso-occlusive pain episode in each focus year (2011- 2013). To decrease bias associated with multiple admissions from the same individual, the first admission for vaso-occlusive pain in each year was evaluated as the index admission for each record. Cases and controls were defined as adults with SCD initially admitted for pain and subsequently re-admitted to the hospital within 30 days of the initial admission. A multi-variable logistic regression analysis was performed on seven postulated risk factors. All data was collected and double checked by a single reviewer, and at least 10% of the chart was checked by a medical student for further assurance of accuracy.

Results: A total of 158 first admissions and 49 re-admissions (31%) were evaluated. The mean age of the cohort was 30.38 (IQR 13.55 years). The median time to re-admission was 10 days (IQR 19 days). Approximately 50% of the cohort was not evaluated in the outpatient setting by the hematology team within 30 days post-discharge. Upon discharge patients either were not given a follow up appointment (35%) or were given an appointment beyond 30 days of discharge (13%). Only two predictors were significantly associated with re-admission within 30 days: not having a primary care provider listed in the electronic medical record (odds ratio 0.35, 95% CI 0.146-0.858; p = 0.022) and the number of hospital admissions due to vaso-occlusive pain in the prior year (odds ratio 1.28, 95% confidence interval 1.15-1.42; p < 0.001), table and figure below. Five covariates were not significantly associated with re-admission within 30 days: age (odds ratio 0.982, 95% CI 0.94-1.02; p = 0.369), sex (odds ratio 0.715, 95% CI 0.28-1.81, p =0.481), hemoglobin phenotype (odds ratio 0.50, 95% CI 0.19-1.287; p = 0.15), median lifetime oxygen saturation (odds ratio 0.892, 95% CI 0.75-1.05; p = .186), and presence of government insurance (odds ratio 1.90, 95% CI 0.67-5.37; p =0.222).

Conclusions: Not having a primary care provider listed in the electronic medical record and multiple admissions in the prior year are potentially modifiable risk factors for re-admission within 30 days in adults with SCD. In addition, discharge planning with a hematology visit scheduled within a week of discharge may also impact the 30-day re-admission rate. We recently introduced a strategy focused on improved discharge planning, ensuring a primary care provider for every adult patient with SCD and targeted therapeutic intervention for those with high admissions.

Figure.

View largeDownload slide

A graph depicting the predicted probablity of a re-admission within 30 days in indivdiuals with and without hospitalization versus the number of hospitalizations in the prior years.

Figure.

View largeDownload slide

A graph depicting the predicted probablity of a re-admission within 30 days in indivdiuals with and without hospitalization versus the number of hospitalizations in the prior years.

Close modal