Data collection for care pathways in the Cleveland Clinic Health System.

Abstract

115 Background: Care pathways are established methods of reducing healthcare costs and disparities in oncology care. To demonstrate their impact, health systems must measure and report data on care pathway adherence and outcomes in near real-time. Automating data abstraction across a health system for oncology is difficult due to the amount and detail of data required. Manual abstraction of data is considered slow and costly. Many consider Electronic Medical Record (EMR) integration of care pathways essential in order to successfully implement and assess. Methods: 7 medical oncology care pathways and 45 medical oncologists across the health system were selected for a pilot study to assess the feasibility of implementing care pathways throughout the enterprise. The pilot study also allowed for testing of data collection capabilities. Patients eligible for the care pathways were prospectively identified by manual review of physician calendars. A small number of data points were manually abstracted from the patient EMR at the time of identification. Endpoints of interest, such as hospitalization rates, chemotherapy administered, time to treatment, and costs of care were reconciled through pre-existing databases within pharmacy, research, and finance. Tumor registry data identified a retrospective cohort. Results: Over 1,000 patients were prospectively identified for the care pathway pilot between 1/1/2014 and 12/31/2014. The tumor registry identified 700 additional retrospective patients. The rapid analyses possible as a result of these efforts demonstrated physician adherence, improved patient outcomes, and significant cost savings. In one example, a care pathway for metastatic non-small cell lung cancer reduced charges by more than $98,000/patient by recommending patients receive one standardized chemotherapeutic regimen. Conclusions: Timely data collection for oncology care pathways is feasible and cost effective without EMR integration. Manual identification of patients combined with pre-existing data sources allowed for near-real time analysis of care pathways and provided valuable information about care pathway impact. Institutions can implement and assess care pathways with resources already available to them.