Several studies have reported considerable variability in IRB decisions about research protocols (see, e.g., Burman et al., 2001, 2003; Silverman et al., 2001; Stair et al., 2001; Hirshon et al., 2002; and McWilliams et al., 2003). As additional examples of variability, Tables 8.1 and 8.2 report selected results of IRB reviews of two protocols for multicenter clinical trials conducted through the Cystic Fibrosis Therapeutics Development Network (Ramsey, 2003). Both trials included adults as well as children. For one protocol involving an early-phase study of the safety and tolerability of an oral pancreatic enzyme product, the time to IRB approval for nine IRBs ranged from 3 to 18 weeks with a median of 5 weeks. Six IRBs gave full approval for the enrollment of children as planned; three IRBs approved the study for adult participants only, although one gave approval to include children following appeal. For a second protocol for a randomized, double-blind, placebo-controlled study to investigate the safety, tolerability, and pharmacokinetics of an anti-inflammatory agent in adult and child patients, four of seven IRBs gave full approval whereas three approved the research for adult participants only. Of the IRBs giving full approval, three categorized the research as involving a minor increase over minimal risk and no direct benefit, but one categorized it as involving more than minimal risk with the prospect of direct benefit.
To cite another example, the REACH (Reaching for Excellence in Adolescent Care and Health) project of the Adolescent Medicine HIV/AIDS Research Network published its experience with IRB determination of risk among 11 investigational sites as reported by site investigators (Rogers et al., 1999). The study involved adolescent subjects between 12 and 18 years of age who were either HIV positive through either sexual activity or drug use or who were HIV negative but engaged in high-risk activity. The objective was to examine the progression of HIV while controlling for the comorbidity of other sexually transmitted diseases. The study procedures included face-to-face interviews, a computerized interview (site blinded), laboratory analysis of clinical samples (blood, urine, blinded drug screen), physical examinations (including gynecological and urogenital examinations), and a wrist radiograph to determine bone age. All REACH sites had a certificate of confidentiality (see Appendix C) from the federal government. The blood volume to be drawn was 100 milliliters (ml) at baseline and annually, 60 ml at 3 and 6 months, and 80 ml at 6 months for HIV-positive subjects and approximately 60 ml at baseline and annually and 50 ml at 6 months for HIV-negative subjects. Four IRBs considered the protocol to present no greater than minimal risk, while one considered it greater than minimal risk. One judged it to present greater than minimal risk for HIV-negative subjects only, and two additional IRBs required changes in order to consider the protocol to be minimal risk. One imposed a screen for anemia prior to blood draws, and the other required that the wrist x-rays