CHOICE: Coalition of Hospices Organized to Investigate Comparative Effectiveness

More than 5,000 hospices provide care for over 1.5 million patients every year. Hospice patients receive care in their own homes, nursing homes, hospitals, and dedicated hospice units. Comparative effectiveness research may help hospices provide higher quality care; however, research in this setting is difficult to conduct because patients enroll in hospice late in the course of illness.

This project studied the feasibility of abstracting quality measures from the electronic medical record (EMR) at 12 hospices in the Coalition of Hospices Organized to Investigate Comparative Effectiveness (CHOICE) network. CHOICE is a collaborative of hospices that have agreed to share their data for research purposes. Participating hospices ranged in size from 400 to 1,700 patients, were located in 11 States, and were not-for-profit organizations.

The specific aims of this project were as follows:

• Evaluate the feasibility and validity of abstracting quality measures from EMRs among a network of hospice members. 
• Assess whether the measures that are abstracted are associated with families' perceptions of care. 
• Identify which quality measures hospice stakeholders identify as the highest priority for use in performance improvement and clinical trials. 

The project established that extracting EMR data from hospice on a large scale was feasible. Key data elements included demographic data, indicators of illness severity, health outcomes, and indicators of quality. The data were used to track preferences regarding site of death, compare underrepresented hospice subpopulations such as children and older adults in assisted living, and develop an acuity index for use in future studies. Analyses identified a significant association between families’ perception of quality of care and hospice staff visits on the last day of life. The CHOICE steering committee identified the following indicators for future data collection and benchmarking: 1) pain and dyspnea screening on admission, 2) management of opioid induced constipation, 3) preferences regarding resuscitation and hospitalization, and 4) visits by hospice staff on the patient’s last day of life.

This project laid the foundation for using EMR data for comparative effectiveness research and an initial proof of concept for a national quality improvement research network for hospice care.