Public outcry, confusion stops an Oregon study
Scrutiny of one study impacts the other
Investigators and officials with Oregon Health and Science University in Portland have learned a difficult lesson about research evaluating controversial social policies.
A 1999 National Institutes of Health-sponsored project to evaluate the results of a drug-testing program in Oregon public high schools was ultimately suspended after the federal Office for Human Research Protections (OHRP) decided that the goals of the research protocol appeared to have become intertwined with the drug-testing policy it was intended to study.
Critics claim that the design of the Student Athletic Testing Using Random Notification (SATURN) study coerced student athletes to participate in school-sponsored random drug testing and failed to provide appropriate informed consent procedures and confidentiality protections.
Researchers and university officials contend that their study protocol simply consisted of surveying students who independently agreed to participate in the drug-testing program and that critics both inside Oregon and nationwide have confused elements of the SATURN study with the testing program it was designed to evaluate.
According to a report published in the winter 2004 edition of the American Journal of Bioethics (AJOB),1 SATURN encouraged schools to adopt random drug-testing policies of students participating in extracurricular activities. In some schools, consent for random drug testing was a condition of the students’ being allowed to participate in extracurricular athletics. Positive drug tests resulted in the notification of school administration, the student’s parents, and the loss of athletic participation privileges.
The informed consent documents signed by participants in the SATURN study do not adequately inform participants of the risk of loss of confidentiality if they test positive for drug use and do not adequately explain the randomization methods used to place schools in the experimental and control groups of the study, argues Adil Shamoo, PhD, a professor at the University of Maryland School of Medicine, editor-in-chief of the publication Accountability in Research, and the co-author of the report in the bioethics journal.
The study also coerced participation because it allowed the school administration to decide whether to participate — by implementing a testing program — and then allowed the disclosure of drug testing results, which could result in disciplinary action.
The IRB at Oregon Health and Science University failed in its mission to appropriately oversee and monitor the study protocol, Shamoo states.
"I believe that research involving mandatory public health initiatives [like the drug-testing program] can be conducted, and one can design a study of the effects of a mandatory public health requirement that is totally de-linked from the mandatory requirement itself," he tells IRB Advisor.
Doing so is one of the key challenges of research into compulsory public health interventions, Shamoo notes. The subjects of the health interventions are inherently vulnerable populations, and efforts to conduct research studies on the populations that are concurrent with the intervention are difficult.
"This [the suspension of the SATURN study] has important implications in terms of mandatory public health requirements linked to a research protocol," he adds.
Researchers must be extremely careful that the research is clearly de-linked from the public health intervention and that participants are aware of the distinction. In the Oregon case, Shamoo contends, participants were unable to distinguish between participation in the testing program and participation in the research protocol.
Critics confused the facts
The SATURN project did adequately separate its intervention from the drug-testing initiative, and Shamoo’s article misstates several key facts, argues Gary Chiodo, DMD, director of the Office of Research Integrity at Oregon Health and Science University. Chiodo, with two of the researchers who worked on SATURN, published a response to the AJOB article in the same issue of the publication.2
Relying largely on inaccurate accounts in the news media, the article by Shamoo and Moreno comingles elements of the drug-testing program with features of the study design, Chiodo charges. The protocol, in addition to being approved by the local IRB, also was vetted by a scientific review panel at the National Institutes of Health, he says.
First, Chiodo says, schools independently decided whether to implement a policy of random testing of student athletes. The schools also independently — without input from researchers — decided whether students would be prohibited from participating in athletics if they refused to participate in the program.
SATURN’s purpose was to study the impact of the drug testing interventions — whether their use influenced student attitudes and reports of student drug use.
The research intervention consisted solely of the administration of written surveys to students who had already agreed to participate in the program, Chiodo says. The study was divided into two groups: an experimental group consisting of students at schools with random drug testing in place, and a control group of students at schools that did not use random drug testing.
In order for a school to be considered a research site and its student athletes eligible for recruitment, the school had to have already decided to implement a policy of random testing of student athletes. Schools randomized to the control group agreed to defer implementation of the program until after the SATURN surveys were completed. As a condition of participation, all schools agreed to defer implementation of the program if the site was randomized to the control group.
"The student populations for experimental and control schools had to be comparable for a legitimate randomization," Chiodo explains. "Students in schools that had not decided to implement drug testing could have an entirely different belief or mindset regarding drug and alcohol use than their peers in drug-testing schools."
In addition, none of the schools included in the SATURN study used punitive policies — policies that prescribed punitive measures if a student tested positive for drug use. Researchers excluded the schools with punitive policies in favor of only schools that had developed policies mandated counseling and referral for students testing positive.
"Punitive policies were those that prescribed a punishment for a positive test," Chiodo says. "This could include being suspended from athletic participation or other school activity, detention, involvement of law enforcement, or similar measures. Nonpunitive policies were those that prescribed an educational or counseling intervention. These students were not excluded from extracurricular activities."
Consent form made distinction
So, no student athletes in the SATURN study schools, neither those in the experimental nor those in the control schools, were at risk of having the results of their drug tests disclosed to their parents, authorities, or being excluded from athletic participation. Punitive policies were in place at some of the schools implementing random drug testing, but not at schools included in SATURN, he points out.
The consent forms used by SATURN researchers also clearly identified the research project as separate and distinct from the consent process to participate in the drug-testing program, Chiodo says. And researchers have evidence that the students and the parents of students (who received the consent forms) understood the difference.
"Because the schools’ intervention [random drug testing of athletes] is very distinct from the study intervention [completion of questionnaires] the chance for confusing the two was already reduced," he explains. "However, high schools students may not always perceive this distinction. The distinction was emphasized and the chance for confusing the two diminished by the informed consent process and documents and by involving the parents in the consent process. The SATURN informed consent documents addressed both parents and the students and made it clear that the schools’ policies and procedures were separate from the study’s interventions and goals. In addition, live meetings with parents, informational letters, and [frequently asked questions] sheets were used throughout the project to help reduce any confusion about school vs. study objectives and procedures. All of these emphasized that students did not have to participate in the study; and if they declined to do so, they would not lose any rights, privileges, or benefits to which they were otherwise entitled. Students received multiple messages to this effect."
Data support researchers
Data obtained during the study indicate that the students clearly grasped the difference. During the period of the SATURN study, student athletic participation increased in those schools that were participating in the study and, 30% of the athletes in the drug-testing experimental schools chose to participate in sports but declined to participate in the study.
People in Oregon and beyond who objected to the drug-testing programs then confused the study protocol with the programs, which were the subject of the research itself.
"Oregon citizens and individuals from around the world confused the SATURN project with the drug testing intervention it was designed to study," Chiodo says. "The study did not require drug testing of any student. It appears that those who objected to the study objected to any policy of drug testing students. This is a civil liberties issue and far afield from the specific aims of the study. The study held out the prospect of proving with scientific accuracy the ultimate outcome of random drug testing of high school students. Drug testing may be beneficial, have no effect, or, in fact, be harmful. SATURN was designed to prove which of these outcomes result from this intervention. If it is harmful or neutral, society would be much better off devoting the resources spent on drug testing to other interventions, such as education, after-school programs, peer mentoring, and other methods."
Although OHRP did suspend the study, researchers have anonymized the data they did collect and intend to publish the results they have, he adds.
Studies that delve into sensitive social issues will always be targets for those impassioned about the issue, even if they misunderstand the study, he contends. Investigators, institutions, and IRBs must be prepared for this.
"Such studies are critically important and must go forward," Chiodo concludes. "President Bush recently announced in a State of the Union address that he was recommending $23 million be spent on drug testing in schools. If such testing is ineffective or counterproductive, the funds will have been wasted and an opportunity to provide a more meaningful intervention will have been lost. The current system of scientific review group scrutiny followed by strict IRB review, holding the study to the highest ethical standards, did, in fact, work for the SATURN study. The study design was sound, the ethical bar was set high, the investigator team was responsive and acted with integrity, and the subjects were clear on what enrollment in the study meant."
1. Shamoo AE, Moreno JD. Ethics of research involving mandatory drug testing of high school athletes in Oregon. Am J Bioethics 2004; 4:25-31.
2. Chiodo GT, Goldberg L, Moe EL. Orbiting SATURN: Countering politically charged misinformation with facts. Am J Bioethics 2004; 4:4348.