Development and Validation of a Screen for Specialized Discharge Planning Services

Background: There is no rigorously developed and empirically validated screening tool to identify, early in the hospital stay, those adults who will use specialized hospital discharge planning services. Objectives: To develop and validate a screen using hospital admission clinical data that discriminates between adults who use and do not use specialized discharge planning services. Methods: Subjects consisted of prospectively sampled adult patients admitted to two hospitals located in a Midwestern United States city in 1998 (tool development sample, n = 991) and 2002 (validation sample, n = 303). Variables suggestive of being predictive of use of specialized hospital discharge planning services were identified from the literature and were obtained from direct participant interviews, record review, and administrative databases. The outcome was a documented referral for involvement of specialized discharge planning personnel with the patient's plan of care and was identified from review of hospital records. Results: Of 24 variables examined, only age, disability, living alone, and self-rated walking limitation were jointly predictive of use of specialized discharge planning services in the development sample. Standardized coefficients from the joint model were used to estimate a screening score. A cut-point was derived and had a sensitivity of 75% and specificity of 78% in the development sample. The screen performed equally well in the validation sample and the development sample. Conclusion: A screening tool consisting of a limited number of characteristics readily available early in the hospital stay that were shown to be highly predictive of the use of specialized discharge planning services was developed. The application of such a tool will hopefully assist providers to deploy services appropriately and in a timely fashion.