Presenting Author:

Victoria Rodriguez, M.D.

Principal Investigator:

Victoria Rodriguez, M.D.

Department:

Pediatrics

Keywords:

Readmission, Preventability, Perspectives

Location:

Ryan Family Atrium, Robert H. Lurie Medical Research Center

C96 - Clinical

Varying Perspectives on Preventability of Pediatric Readmission within 3 Days

Background: Unplanned hospital readmission is a key quality indicator that may contribute to unnecessary health care spending. 20-30% of pediatric readmissions are potentially preventable and a disproportionate frequency may occur within 3 days of discharge, but there is no gold standard defining a preventable readmission. Objective: To compare the perspectives of various stakeholders on the preventability of readmission. Design/Methods: For patients readmitted within 72 hours, we conducted a retrospective chart review, parent interview, discharge provider interview, and readmission provider interview. We excluded patients with a planned readmission, not discharged home between visits (i.e. transfer to another facility), or age > or = to 18 years. Chart reviews and interviews assessed the cause of admission using a structured questionnaire inquiring as to potentially preventable issues in medical treatment, discharge process, social determinants of health, and included a final determination on if the readmission was preventable. Demographic and quantitative data were analyzed using descriptive statistics and kappa coefficients were used to compare discharge, readmission, and chart reviewer physician perspectives. Results: 88 cases met eligibility criteria from 7/20/16-11/22/16. There were no significant differences in patient variables, such as race, insurance status, and length of stay, between patients with preventable and non-preventable readmissions as determined by chart review. Physician chart reviewers found 45 (51%) readmissions preventable. Initial discharge physicians (n=66) identified 18 preventable readmissions (27%). Readmission physicians (n=69) identified 21 preventable readmissions (30%). Families (n=22) identified 12 preventable readmissions (54%). Agreement was fair on preventability between chart review physicians and discharge physician (kappa .263) and chart review and readmission physician (kappa .351). Agreement was good between discharge and readmitting physicians (kappa .598). Conclusions: Independent chart reviewers and families found readmissions to be preventable more frequently than treating (discharge and readmission) physicians. Discharging physicians assessed a readmission to be potentially preventable with the lowest frequency. Further study is needed to determine whether self-affirming bias by treating physicians under-reports preventability or whether preventability is over-ascribed by the independent reviewer once the outcome of readmission is realized.