Purpose: To enhance the scientific richness of the outcome database for heart failure in the Women's Health Initiative (WHI) cohort. The National Institutes of Health (NIH) established the Women's Health Initiative (WHI) in 1991 to address the most common causes of death, disability and impaired quality of life in postmenopausal women. The WHI addressed cardiovascular disease, cancer, and osteoporosis. The WHI was a 15 year multi-million dollar endeavor, and one of the largest U.S. prevention studies of its kind. The Collaborative Studies Coordinating Center at the University of North Carolina at Chapel Hill\’s Department of Biostatistics is abstracting information from hospital records already obtained under direction of the WHI project and validating whether identified cases contained heart failure events. For events that are classified as being heart failure, the sub-type of heart failure (systolic or diastolic) will be determined.

Participants: 5000 women who enrolled in the 2010-2015 WHI extension and who reported hospitalizations suspected to be for heart failure. Medical records for these hospitalizations will be reviewed and classified in this study.

Procedures (methods): For this study we will receive medical records obtained by the WHI. Specially trained nurses will then abstract diagnostic and treatment information from the records using standardized forms. Information from this abstraction will be condensed for a panel of physicians. The panel of physicians will then classify each event as to whether or not the event was a hospitalization for a diagnosis of heart failure. Investigators in this project will provide expertise in the validation and classification of heart failure events. Methods include detailed abstraction of medical records, processing of clinical diagnostic information to determine a computerized classification and managing of materials through a clinical review process by physicians with expertise in heart failure diagnosis and classification for each event. All records received and viewed by this application are de-identified and all identifers are masked. Investigators in this application are blinded to personal identifiers that resided with the parent study (WHI).