Abstract
Dear Practical Ethicist,
After discussion of a research protocol, our institutional review board (IRB) sometimes spends a great deal of time editing the consent form. While the edits are sometimes necessary to fix inconsistencies with the protocol, the changes suggested by members are often more editorial; reformatting the way a list of risks is presented, changing the grammar of sentences, removing medical terms (even when they are defined) and replacing them with lay language. Some members have started to complain about the amount of time spent “wordsmithing” the consent document, while others insist that the changes are necessary to make the consent document understandable to the potential participants. How do we find middle ground?
Sincerely,
Finding Understandability Necessary
Dear FUN,
Informed consent requires that research participants to be given sufficient information about the research, that they understand that information, and that they freely agree to take part in research (U.S. Department of Health, Education, and Welfare, 1978; World Medical Association, 2013). Many IRBs/RECs (Research Ethics Committees) spend most of their time editing the informed consent document. Sometimes, this editing is to correct errors, or to ensure that required elements of consent are included. Often, members justify the changes as necessary to ensure that information provided to the participant or the participant’s representative be in understandable language (International Council for Harmonisation, 2015; Protection of Human Subjects, 2009). This raises the questions of what understandable language is, and how an IRB/REC decides whether information is understandable.
Ethics review must, inter alia, ensure that the information given in the consent process is accurate and includes sufficient content that potential participants will need to make an informed decision whether to take part in research. What does “information that is given to the participant or the participant’s representative” mean? Does this refer to each separate item of information, of which the consent document is one, or does it refer to the consent process overall? If the consent document must be in language understandable to the participant or the participant’s representative, this is a difficult standard to meet for every possible participant. If 10% of participants have questions about what they read after reviewing the consent document, the language was not understandable to 10% of participants. If even one participant has one question that needs to be answered to understand the information, the consent document was not completely understandable to every participant. We must conclude that it is impossible to write a consent document where in all cases the information will be understandable to every participant or participant’s representative.
It is useful to review the US 45 CFR 46.111 criteria for approval, which state, In order to approve research the IRB/REC shall determine that all of the following requirements are satisfied: . . . Informed consent will be sought from each prospective participant or the participant’s legally authorized representative, in accordance with, and to the extent required. (Protection of Human Subjects, 2009)
As written, this criterion applies to the consent process and not the written documentation of consent. The only criterion for the content of the consent document is that it has to embody the required elements of disclosure. There is no specific requirement that the informed consent document be written in language understandable to the participant or representative, which as discussed previously, is an impossible standard to meet. Therefore, “information that is given to the participant or representative” must refer to the sum total of the information, including the discussion with the research team. Any effort to modify consent documents for understandability must take into account that no consent document will be completely understandable to every participant and that investigators following their ethical obligations will ensure that participants understand the information regardless of the complexity of the consent document.
Notably, the criterion for approval refers to individual participants rather than participants in general. The requirement is for language understandable to the participant rather than language understandable to participants. Can we say that if the majority of participants are given understandable information that the criterion is met? We do not do this for any other criterion. For example, we do not approve research where the risk–benefit relationship is reasonable for most participants but not all. Therefore, to approve research, the consent process requires each participant or representative to understand the information presented during the consent process.
Commonly, IRB/RECs assess understandability by checking the “readability” level of the consent document, using functions built into word processing software. There are many difficulties with defining understandable language in terms of readability level. Lower reading level scores mean that the document has shorter words, and shorter, less complex sentences. Revising a consent document to lower the readability level can be achieved by removing the long names of drugs, procedures, anatomical structures, and physiologic processes, resulting in a consent document that excludes terms that need to be explained so that participants understand them and make an informed decision. Aiming for a specific reading level based on the mean literacy level of a general population means that half of all potential participants are given written information that is not understandable. Even if we could lower the reading level to the fourth grade in the United States, 25% of the general population would be unable to understand the document (Clear Language Group, 2018). In addition, since the software calculates reading level based only on word and sentence structures, the document could be written backwards—clearly not understandable—but would still have the same reading score. Therefore, reading level as determined by software cannot be a sufficient threshold to determine that information is understandable.
We must conclude that understandable information is an end goal of the consent process and not a requirement of the consent document. Current ethical codes and guidelines support this assertion. The Declaration of Helsinki states, “After ensuring that the potential subject has understood the information, the physician or another appropriately qualified individual must then seek the potential subject’s freely-given informed consent” (World Medical Association, 2013). The Belmont Report states, “Because the subject’s ability to understand is a function of intelligence, rationality, maturity and language, it is necessary to adapt the presentation of the information to the subject’s capacities” (U.S. Department of Health, Education, and Welfare, 1978). International Ethical Guidelines for Health-related Research Involving Humans states “An oral presentation of information or the use of appropriate audiovisual aids, including pictographs and summary tables, are important to supplement written information documents to aid understanding” (Council for International Organizations of Medical Sciences, 2016).
IRB/REC members often worry that they cannot approve (especially in advance of enrollment) the consent process, so they must write the consent document as if it is the entire consent discussion. This concern can be alleviated if the investigator describes their consent process in the submitted materials and agrees to adhere to that process, similar to the way that application materials often have investigators attest to adhere to the protocol. A consent process might, for example, include evidence of engagement with a Community Advisory Board or other rigorous community engagement process, especially where study risks are greater than minimal. In other cases, a consent process might include a test of understanding (e.g., Lindegger et al., 2006). Where there is credible evidence that the investigator will not follow the consent process described in the protocol, the research should be disapproved. Similarly, where there is credible evidence that the investigator will not ensure that the information provided is complete or will not ascertain that the participant has comprehended the information, the IRB/REC should not approve the project until they are confident about the overall consent process, rather than reducing their concerns by trying to perfect the consent document.
In summary, to approve research, the IRB/REC must determine that information provided to the participant or the participant’s representative is understandable to the participant or the participant’s representative. However, “information” must refer to the totality of information, including oral communication with the participant, because no written document will be understandable to every participant. “Understandability” is the endpoint of the consent process after the information in the consent document has been communicated and explained, and after all questions are answered. Ensuring understandability requires the investigator to ascertain that the participant or representative comprehends the information. If there is concern that the investigator will not do this, then the IRB/REC should disapprove the research in the same way they should disapprove the research if there is concern that the investigator will not follow the protocol.