Building population data bases: lessons from a community project

Observational data bases that tract patient outcomes and quality of life over long periods of time offer an alternative to traditional randomized clinical trials in assessing treatment effectiveness. This article reviews our experience building an observational database in a community setting. Thirty-eight urologists, radiation therapists, oncologists, and pathologists working at the five major hospitals in Hartford County, Connecticut, established a community-based observational data base to quantify the risks and benefits associated with different treatment options for clinically localized prostate cancer. In addition to classic demographic data, the data base includes information concerning patient comorbidities, bowel and bladder function, sexual function, and quality of life captured both before and after medical intervention. Case finding is accomplished via a rapid case ascertainment system. As part of this system, a trained abstractor reviews all the pathology reports generated by each hospital on a weekly basis and forwards reports of prostate biopsies and surgical specimens to the study office within 1 week. Patients with positive prostate biopsies are contacted by a trained nurse registrar to gather additional information using a structured interview. Patient consent for this project is obtained at the time of biopsy in the participating physician's office. To date, biopsy reports have been obtained on 581 patients of whom 202 have been diagnosed with localized prostate cancer. Interviews have been conducted with 42 patients. Initial review of the data base shows that prostate biopsies in Hartford are being performed primarily on men aged 59 to 77 years, but ages range from age 45 to 93.(ABSTRACT TRUNCATED AT 250 WORDS)