Memory Overload: The Impossibility of Informed Consent

Truly informed consent is impossible. Michael Sharp's 2004 oncology consent form study found that consent forms typically included 14 topics and 2700 words in 11 pages—not including HIPAA. Some were over 6000 words long. Add another 500–1000 words for HIPAA, and the typical oncology consent form averages 3500 words. That's too much information for anyone's brain to remember and process for an informed decision.

Mark Hochhauser, PhD

Memory research finds that working memory—how much information is available at a given time—is only one to four items. This psychological limit makes it impossible for prospective subjects to remember all the information they are expected to remember in the consent process plus consent form. Because consent forms are required to include so much detailed information, they create their own failure to inform. More information equals too much information, which creates memory overload, not more understanding.

Readability doesn't improve memory

Federal regulatory agencies, IRBs, and bioethicists have not yet linked memory limits to the consent process. Hard-to-understand (remember?) consent forms are to be written at a 6th–8th-grade reading level to improve understanding. But writing consent forms at lower grade levels will have little effect on the number of topics, words, or pages in a typical consent form. The number of consent form topics (and HIPAA) cannot be reduced; they are required by federal regulations. How can prospective subjects remember 14 topics, some of which are explained in excruciating detail, when their working memory can only remember four? Writing consent forms that follow plain English recommendations might make consent forms more readable and more attractive, but will probably have little impact on reader memory.

Despite requirements for consent forms at lower grade levels, evaluation studies find little or no improvement in understanding. Even modest improvements are questionable because of serious flaws in consent form comprehension studies, such as the inability to operationally define comprehension, lack of validity and reliability of the comprehension measures, use of comprehension measures (true–false, multiple-choice) susceptible to guessing, etc. Nevertheless, consent form writers are still encouraged to write consent forms at a 6th–8th-grade reading level, although in my 11 years' experience on an IRB I have never read a consent form at that level.

Memory-based consent forms

Federal regulators and IRBs have it backwards. They assume consent forms will be more understandable if they are more "readable"—even if there is no evidence to support that assumption. They start with the 14 topic consent form and assume that it can be understood if only consent form writers use shorter sentences and smaller words. But because memory is not part of the informed consent discussion, that assumption is little more than magical thinking.

The consent process and consent form regulations and recommendations should be based on, or at least influenced by, research on memory and information processing strategies. Some investigators address comprehension by asking subjects to describe the study back to them during the consent process. Even if subjects can do that reasonably well, how long does their memory for that information last? Would subjects have the same memory and understanding if questioned a day later? A week later? A month later? The success of short-term memory does not guarantee that the same information will be converted to long-term memory; these are two different psychological and biological processes. Because federal consent requirements contradict the brain's ability to remember and process those requirements, consent forms cannot be remembered or understood by prospective subjects.

Mark Hochhauser, PhD, IRB member, North Memorial Health Care