Abstract
HIV patients are at increased risk for depression. However, a comprehensive set of quality of care indicators for depression in the context of HIV does not exist. We report the results of a recent expert panel convened to develop a comprehensive set of HIV depression quality indicators. We systematically reviewed the depression and HIV depression literature to provide the evidence for the quality indicators. The HIV depression quality indicators were divided into six chapters: general, bereavement, substance abuse, viral hepatitis, cognitive impairment and drug interactions. A modified Delphi expert panel method was used to evaluate the validity and feasibility of the candidate quality indicators. The expert panel included HIV physicians, psychiatrists and a pharmacist, all with experience with HIV depression. The in-person eleven-member panel rated 80 candidate HIV-depression quality indicators. Indicators were evaluated using main and sensitivity analysis criteria for validity and feasibility. Seventy-four percent (59/80) of the quality indicators met a priori main analysis criteria for validity and feasibility and 61% met sensitivity analysis criteria. Approved indicators that are more specific to HIV depression include: depression screening frequency; depression screening and treatment in the context of hepatitis C; optimizing depression and HIV treatment in the context of cognitive impairment; and starting antidepressants at lower doses but advancing to doses typical for the general population. Most adverse medication interaction indicators did not meet main analysis validity criteria. The quality indicators identified in this study provide a useful tool for measuring and informing the quality of HIV-depression care.
Keywords:
Acknowledgements
The authors wish to thank the following expert panel members for their participation in this project: Paul Volberding, Matthew Goetz, Chris Mathews, Robert Schooley, David Looney, Todd Korthuis, Igor Grant, Sid Zisook, Seth Himelhock, Ian Everall and Pam Belperio. This project was supported by a US Department of Veteran Affairs HSR&D Quality Enhancement Research Initiative (VA-QUERI) Locally Initiated Project Grant.