Unplanned hospital readmissions in the first month after childhood cancer diagnosis.

Abstract

e18770 Background: Overall, unplanned 30-day readmission rates at children’s hospitals are 6.5% ( JAMA, 2013;309:4). However, the rate of unplanned readmissions within 30 days following a childhood cancer diagnosis is unknown. We aimed to: i) Determine the rate of unplanned readmissions in pediatric oncology patients during the first 30 days following discharge from the hospital after diagnosis; and ii) Identify risk factors associated with these unplanned readmissions. Methods: Newly diagnosed patients enrolled in a study at Children’s of Alabama were followed for 30 days after their initial discharge and data on all readmissions collected. Unplanned readmission rates were calculated using the Agency for Healthcare Research & Quality All-Condition Readmission Measure. Intensity of Treatment Rating (ITR), a validated categorization of the intensity of pediatric cancer treatment (1 = least to 4 = most intense) was assigned based on each patient’s treatment plan. Demographic and clinical variables were summarized using descriptive statistics. Bivariate logistic regression analyses were performed. Patients with less than ten days at risk of readmission were excluded from the regression model. Covariates with a significant association (P < 0.1) with unplanned readmission were included in the multivariable regression models. Results: There were 105 patients included in the study (mean age, 7.7±5.4 yrs; 62% male, 66% non-Hispanic white; 50% with leukemia; 50% with ITR-rating of 3; discharged with a median of 8 [range, 2-14] home medications). The unplanned readmission rate was 48.6% (51 of 105 patients had at least one unplanned readmission during the 30-day post-discharge period). Of the 3135 patient days of follow-up, 887 were inpatient (486 planned and 401 unplanned). The unplanned inpatient days were for treatment of fever (81%), gastrointestinal issues (11%), pain (2%), neurologic issues (5%) and bleeding (1%). The proportion of patients with any unplanned readmission by Intensity of Treatment Rating (ITR) were ITR-2: 39%, ITR-3: 59%, ITR-4: 71%. Multivariable regression (adjusted for sociodemographic and clinical variables) demonstrated the following associations with unplanned readmission: ITR (OR, 1.24 per level; 95% CI, 1.01-1.50); number of discharge medications (OR, 1.05 per medication; 95% CI, 1.01-1.09). Conclusions: The 30-day unplanned readmission rate in newly diagnosed pediatric oncology patients is substantially higher (48.6%) than for children’s hospitals overall (6.5%). The odds of unplanned readmission increased independently with the number of discharge medications prescribed (5% increased odds per medication) and with the Intensity of Treatment Rating (24% increased odds per rating level). These factors may be useful in tailoring anticipatory guidance for families regarding the likelihood of readmission during the first 30 days following a pediatric cancer diagnosis.