BACKGROUND: Post-marketing safety surveillance studies often require non-traditional methods for data collection. In the case of rare outcomes and outcomes with possible long latency, traditional large databases may not be sufficient. Some studies require collection of data directly from patients. However, gaining access to patients for interview or to their medical records for abstraction is becoming increasingly more difficult in the new post-HIPAA era.
OBJECTIVES: Describe logistical challenges encountered in an ongoing 10-year post-marketing safety surveillance study in which patients are identified through cancer registries and possible prior exposure is ascertained by telephone interview.
METHODS: We conducted descriptive analyses of site approval requirements and time-to-approval for a postmarketing safety surveillance study of adult osteosarcoma patients in the United States (U.S.). To identify a cohort of at least 40% of diagnosed cases of this rare tumor each year, oncology referral centers (ORCs) with the greatest volume of adult osteosarcoma cases were targeted for recruitment. ORCs include comprehensive cancer centers and regional SEER or state central cancer registries. We describe time to final approval, time to first data delivery, number of different consent protocols, and mean number of submissions for ORCs initiated for this study.
RESULTS: Between December 2002 and December 2004, the IRB application process was initiated at 15 ORCs. As of January 2005, approvals have been granted for seven ORCs. Of these, five required revision of original procedures due to data privacy concerns. Average time from first contact to final approval was 5.2 months (range 1–12 months). Average time from final approval to data delivery ranged from 1 to 8 months and is still pending for two ORCs. Mean number of formal submissions was 3.1, ranging from 1 to 8. Of seven ORCs, six different protocols were required to obtain patient consent. Mean time to approval and number of submissions will increase once ORCs applications still under review are approved.
CONCLUSIONS: The route to ultimate approval requires persistence and the ability to adapt the design to meet institutional requirements. Two key challenges encountered were: (1) institutions’ lack of understanding of case-finding surveillance study methods (in contrast to clinical trials or casecontrol studies) and (2) privacy concerns post-HIPAA. Researchers should be aware of and plan for the increased resources and time required to launch post-marketing safety studies.
