NOTES FROM THE FIELD
End-of-Life and Hospice Care for Ovarian Cancer

The project "Evaluation of End-of-Life Care for Ovarian Cancer" aims to understand the occurrence and management of pain and other major complications at the end of life and to examine referral and use of hospice among ovarian cancer patients in HMO environments.

This is a CDC-funded project (Steve Coughlin) administered through the Association for Community Health Plans (Cia Byrnes). It involves Group Health (Kari Bohlke), HealthPartners (Chrii Rolnick), Henry Ford (Chris Neslund-Dudas), KPNC (Lisa Herrinton, PI), and KPNW (Mark Hornbrook). It was built upon the successful work of Chris and Kari to examine prostate cancer care at the end-of-life. Some of you may remember that the ovarian study proposal was put together in one weekevidence of how this group of researchers benefitted from the cohesiveness and communication structures established by the CRN.

Data collection will be through medical record review and data linkage. During Year 1, we developed and piloted the abstraction instrument and learned a great deal about our electronic hospice data.

We spent the first part of the year discussing and refining the specific aims. Numerous data elements were wanted, but some could not be reliably recorded due to data quality and/or availability issues. We also made a major decision to record the medical record data using an encounter approach, so that every encounter would be noted. This would have the benefit of requiring minimal interpretation from the abstractors.

When we moved into the pilot study, we found that we had been too ambitious in specifying the level of detail for capturing ovarian cancer complications. In addition, we found that the encounter approach was unwieldly and that greater interpretation could be demanded from the abstractors.

Ultimately, we decided to obtain details for the major complications only: pain, nutritional management, effusion, and obstruction. In addition, we plan to train the abstractors to define episodes of care, so that one line of data may represent a twomonth course of management for an episode of pain, for example.

The study has just completed Year 1 of its three-year plan. Regrettably, Year 2 budget constraints were such that data collection will occur at only three of the five sites, although all investigators will stay involved. We are seeking additional funding from NCI to bring the two outstanding sites back into data collection. Moving into Year 2, we expect to conduct a bit more piloting before starting data collection in earnest. We look forward to better understanding the death experience in ovarian cancer and hope that this work ultimately will improve systems of care and increase the options available to dying women and their families.

- Lisa Herrinton, KPNC

<< Page 2 | Page 4 >>