Developing and testing a discharge planning decision support tool for hospitalized pediatric patients.

PURPOSE: To develop and test a decision support tool that identifies patients who would benefit from early consult with discharge planners. DESIGN AND METHODS: A predictive, correlational design was used with parents/guardians of children (1 month to 18 years; N = 197). Data were collected by interviews and record reviews. Expert consensus determined referral to discharge planning. RESULTS: Mean age was 8.7 years; mean length of stay was 7.5 days. Forty percent (n = 79) were identified for early referral. The variable "substantial post-acute care needs" had the strongest association with expert consensus (internally validated AUC = 0.79). PRACTICE IMPLICATIONS: Findings from this study provide preliminary evidence for a decision support tool to improve the discharge planning process by reducing individual decision-making variability through systematic matching of patient needs to service delivery.