P3.I170: Identifying risk factors for readmission in children with complex medical conditions

Background: Children with medical complexities (CMC) represent 1% of the pediatric population but account for over a third of pediatric health care costs. Our clinic is a complex care centered medical home with 580 patients aged 0-21. We noticed a high readmission rate within our patient population with an average of 40% of our patients readmitted within 30 days of hospital discharge. We aim to identify patients most at risk of readmission and target future interventions in the hope of reducing the number of readmissions. We hypothesized that certain factors like length of hospital admission, having a new diagnosis or having a preferred language other than English may lead to an increase in hospital readmissions.

Methods: We conducted a retrospective chart review of all patients from January to December 2019 and July 2021 to June 2022 who came to hospital follow-up appointments. We collected data on patient demographics and preferred language, length of stay (LOS), technology dependence, primary service of admission, new diagnoses during the stay, whether the child received intensive care, 30-day same-cause readmissions and emergency department (ED) revisits, and whether the child died on a subsequent admission. We dichotomized LOS into < = 3 days and > 3 days. We also analyzed factors associated with LOS > 3 days. We used Fisher’s exact test for all analyses with a p-value of 0.05 considered significant. Mann-Whitney U was used for all continuous variables.

Results: We analyzed 288 patient charts. We found that patients with technology (p = .021) and race identified as other (p =.011) were more likely to be readmitted, while there was no association with dichotomized LOS, receiving intensive care, new diagnosis, gender, preferred language, or subsequent death (Table 1). Patients with technology (p = 0.031), race identified as other (p = 0.046), and ethnicity identified as Hispanic (p = 0.013) had more ED revisits. Patients with technology (p = .001), who received intensive care (p = .000), and who had a new diagnosis (p = .009) were more likely to have LOS>3 days. We noted that patients admitted were young overall (figure 1) and younger patients were more likely to return to the ED.

Conclusion: Technology dependence was associated with increased risks of same-cause readmission as well as ED visits. Younger patients were more likely to be admitted in general, we hypothesize that this is likely due to a combination of frequent respiratory infections and the parental learning curve. Understanding risk factors associated with revisits and readmissions will aid clinic staff in triaging calls and giving anticipatory guidance. We are exploring ways to decrease the risk of readmission in our technology-dependent patients by partnering with our hospital through consults, increasing time on home settings before discharge and modified in-home education.