Utility of Mayo Clinic's early screen for discharge planning tool for predicting patient length of stay, discharge destination, and readmission risk in an inpatient oncology cohort.

Abstract

To examine the feasibility of using the Mayo Clinic's Early Screen for Discharge Planning (ESDP) tool in determining its predictive ability in an inpatient oncology hospital setting for variables including length of stay (LOS), discharge destination, and readmission risk. A prospective observational study was conducted at a metropolitan tertiary cancer centre in Melbourne, Australia. The ESDP score, along with patient outcomes and characteristics, were collected to examine the relationships between positive and negative ESDP scores and patient outcomes. A total of 136 participants met inclusion criteria for this study. The proportion with positive ESDP scores was greater in those with unplanned hospital admissions compared with planned admissions (χ2(1, n = 136) = 3.94, p = 0.047). The ESDP status was not a significant predictor of oncology hospital LOS (rpb = 0.116, p = 0.178); however, the ESDP scores did predict discharge destination (χ2(2, n = 136) = 20.22, p < .001). Those re-admitted within 14days were more likely to have negative ESDP scores than those not readmitted within this time period (χ2(1, n = 136) = 5.22, p = 0.022). Those with positive ESDP scores received a greater number of hospital services whilst admitted than those with negative scores (rpb = 0.388, p < .001) and were more likely to receive particular types of services. The findings from this study suggest that the ESDP tool could be useful in an adult inpatient oncology population in a hospital with defined specialised hospital discharge planning services (SHDCPS). The ESDP may be beneficial for early identification of service types likely to be required in care and likely discharge destination, both of which can assist discharge planning (DP); however, the ESDP was not useful for predicting LOS or readmission risk in the adult inpatient oncology population without a SHDCPS model in place.