Japan Association for Medical Informatics

[AP2-E2-3-01] Meal Completion as a Substitutional Event for Clinical Pathway Completion: Extension of Comparison from Single Hospital Database to That of Multiple Hospitals

*Hiroki Furuhata1,2, Kenji Araki1, Taisuke Ogawa1 (1. Department of Hospital Institutional Research, University of Miyazaki Hospital, Japan, 2. Graduate School of Medicine and Veterinary Medicine, University of Miyazaki, Japan)

Clinical Pathway, Meal, Propensity Score Matching

We demonstrated that clinical pathway completion can improve quality indicators such as length of hospital stay or mortality rate. Moreover, we extended the completion rate of clinical pathway from binary to percentage using electronic medical records that can identify whether each treatment is registered in clinical pathway or not. Since few databases can record the level of detail of treatment as our case, this study explored events that can substitute for clinical pathway completion. The prospective event was named “meal completion” decided by whether additional fee of meal was claimed every day or not. Approximately 8259 patients were selected for data analysis. With application of a propensity score matching, two groups (1997 patients each) were created for comparison of clinical pathway completion rate between patients with meal completion and those without. Before applying propensity score matching, the difference of rate of meal completion was about 9% (p < 0.001, chi-square test) between patient with clinical pathway completion (25.2%) and those who without clinical pathway completion (34.3%). After applying propensity score matching, the difference of rate of clinical pathway completion was about 5% (p = 0.007, chi-square test) between patients with meal completion (32.5%) and those who without meal completion (27.6%). Although the difference of the rate was small, meal completion is important due to its direct relationship to the patient’s condition. Therefore, meal completion should be validated use this event as a study hypothesis.