Patients as subjects for research: ethical dilemmas for the primary care clinician-investigator.

BACKGROUND: Past studies suggested an association between human papillomavirus (HPV) and cervical intraepithelial neoplasia (CIN). In 1987, University of North Carolina (UNC) Hospitals Family Practice Center clinicians were approached for a control population to study this association. METHODS: One hundred five patients attending the UNC Hospitals Neoplasia Clinic with biopsy-proven CIN 2 or 3 and 268 control patients attending the UNC Family Practice Center for a routine Papanicolaou smear were enrolled in this case-control study. Case and control patients consented to having an additional cervical specimen taken and to being interviewed. The cervical specimens were classified by the Southern blot and polymerase chain reaction techniques for HPV. RESULTS: Early results suggested the control patients who had HPV were at high risk of developing CIN. Interventions were made to inform these patients of this risk and need for closer follow-up, causing a wide range of patient reactions. The final results showed no association of HPV with CIN, indicating the early interventions were premature. CONCLUSIONS: Physicians engaged in research need to be prepared to deal with the discovery of health risks in the otherwise "normal" control patient. They bear the ethical responsibility of scrutinizing study design and methods and planning communication with patients from the inception of a study.