A clinical trial was conducted in a low income country. The trial involved two schools. At the first school (the control school), children would receive a one-off drug treatment for a common infection (such “mass drug administration” is the norm for treating and breaking the transmission cycle of this infection). At the second school, children would receive the same one-off drug treatment plus a long term oral nutritional supplement. The hypothesis being tested was that the supplement might boost the effectiveness of the anti-infective drug. The researchers also wanted to assess whether this supplement might have any side effects when administered in this way. The risk of side effects was low, but nevertheless there was a risk.
In the trial protocol, the researchers stated that children would be asked to invite their parent/guardian to the school on the day that the trial was due to begin, and the parent/guardian would be asked to sign an informed consent form. The consent form stated that the parent/guardian would be free to withdraw the child from the study at any time, and that the child would not be disadvantaged in any way from such withdrawal. The protocol also stated that if the parent/guardian was absent, a representative from the Parent Teachers Association (PTA) would sign the informed consent. An IRB approved the study.
However, in the clinical trial report, the researchers stated that they did not seek parental consent. Instead, they simply asked one PTA representative at each school to sign consent on behalf of all the children in the school. In the trial report submitted to the journal, the researchers justified bypassing the parents as follows: "some children could be excluded from participating in the study and hence disadvantaged from receiving treatment if their parents/guardians did not give consent." The editors feel that the parents themselves would have been best placed to assess the relative risks and benefits of allowing their child to participate in the trial.
We have been unable to find out from the authors whether they told the IRB that they deviated from the protocol and whether the IRB approved the deviation.
One of the researchers commented that he felt the type of consent used was appropriate for this particular country’s setting. Another of the researchers commented that the PTA representatives who gave “mass consent” were themselves parents.
However, the journal editors are concerned that (a) there was inadequate protection of a vulnerable group; and (b) the researchers appeared to use a consent process that made participation more likely rather than a process that fully protected the children. The paper has been rejected due to methodological weaknesses and the editors’ concerns about the informed consent process, but do the editors need to take further action?
The advice from the Forum was to contact the local IRB and determine whether they are aware that there was a discrepancy between the proposed study and what was carried out. Were the IRB aware that the investigators deviated from the initial protocol? If it transpires that the IRB did not approve the study that was eventually carried out, the advice was to contact the authors’ institutions informing them of the situation and asking them to investigate.
The editor wrote to the IRB’s secretary and chairman. Unfortunately, as is often the case with IRBs in low income settings, there was only a mailing address (no email, no fax, no telephone number). The editor sent his letter in the post by priority mail, and so assumes that it was received. He awaits a response.
However, in the mean time, the editor received a letter from the senior author of the paper. The letter stated that he has been in contact with the IRB and the IRB has now granted retrospective approval for the consent that was used in the trial (ie, asking one PTA representative at each school to consent on behalf of all the parents). The senior author wrote to the journal asking if we would now reconsider the paper. The editors gave a firm no, saying that (a) such retrospective approval from the IRB does not allay the initial concerns about mass consent by two individuals and (b) the paper had too many scientific weaknesses to merit publication.