NOTE: This blog post is a highlighted summary of the longer paper from Tim Pratt. To read the full published version go here.
Recently, the FDA released a document entitled “Informed Consent Information Sheet: Guidance for IRBs, Clinical Investigators, and Sponsors (Draft Guidance, July 2014)” which can be viewed here. This is the first such guidance FDA has proposed on Informed Consent since 1998, and represents a substantial expansion and elaboration of the Agency’s thinking on the topic; in 1998 the document issued was a mere 7 pages, this time, the guidance covers nearly 42 pages. The Agency has called for public comment on the guidance by September 15, 2014, and interested parties can go here to do so.
There are multiple new topics covered, as well as expansion on old ones, and specific highlighted focal areas including coercion, undue influence, treatment of children and others. What is of especial concern to sponsors and CRO/MROs is the focus FDA is placing on ensuring that informed consent documents are provided to potential research subjects in “understandable” language, which they define on page 5 of the Draft Guidance as “information is presented to potential subjects in a language and at a level the subjects can comprehend (including an explanation of scientific and medical terms)”. FDA cites concerns from research showing basic or below-basic health literacy as a problem that is widespread in the United States, and contends that it “affects adults in all racial or ethnic groups” and that this affects more than 1/3 of all adults. The basic and below-basic classification levels of health literacy are characterized by the ability to locate easily identifiable information in short commonplace prose texts, following instructions in charts or forms, and locate easily identifiable quantitative information for simple operations (at most, one-step, and primarily addition) that is very concrete and familiar or where the arithmetic operation is specified or easily inferred.
There are some problems with what FDA is contending here though. Firstly, the National Assessment of Adult Literacy (NAAL) research FDA cites is the only research of its type the agency references in its Draft Guidance. This may be because, as I discovered on a literature search, broad-based rigorous studies assessing literacy of any type, but most especially health, are few and far between. It is, as any good scientist knows, quite dangerous to base a course of action however off a single study or trial. Additionally, the NAAL research was conducted in 2003. The data contained therein is thus 11 years old. If a sponsor of CRO/MRO were to approach FDA seeking approval of a therapy based on a single trial conducted over a decade ago, and especially where that therapy had the potential to bring about sweeping changes, it is highly unlikely that the Agency would grant approval for that therapy. Further, the Agency’s contention that the problem affects over a third of all adults in the United States is not borne out by the research itself. What the NAAL findings actually say is that 66% of adult Hispanics have a basic to below-basic understanding, compared to Whites at 28%. Other ethnic and racial groups fall into a broad spectrum between those figures. It seems clear that the problem (not that 28% is in anyway “ok”) is markedly greater in some groups than others and that perhaps addressing the biggest problems first should be a consideration.
Assuming, however, despite the limitation of being a single trial over a decade old, that the NAAL research remains valid, how should we attempt to ensure that patients who are considering becoming subjects are making truly informed decisions about their healthcare? It is extremely difficult to convey, in writing, complex medical information and concepts. Medical professionals use terminology designed to precisely convey specific information. Attempting to convey the same concept in “lay” terms does two things – 1) it dramatically increases the length of the text required, and 2) it introduces a lack of precision. With regard to the former point, I am reminded of a conversation with my father whilst fixing a motorcycle together early in my medical career, when I was telling him about a jejunostomy I had witnessed being performed at Royal Adelaide Hospital. This caused him to launch into a diatribe about medical professionals using technical jargon to “lord it over mere mortals” and keep everyone else in the dark. My response was to ask him to hand me the nearby wooden handled 12 inch long tool with a 1.5 pound double ended hardened steel striking surface, one side of which was flat and circular in lateral aspect and the other of which was semi-circular and convex in shape in order that I might cause a transfer of kinetic energy through the rapid motion of the head of said tool against the 8 inch steel device I currently had attached by means of friction to a hexagonally shaped protuberance of a recalcitrant helically wound fixation device which had undergone an oxidative process such that it would not move without application of more energy than I could supply without the use of said wooden handled metal headed tool: silently, he handed me the hammer to loosen the rusty bolt.
When attempting to convey complex medical information, Informed Consents can easily become extraordinarily lengthy, in order to attempt to “drive down” the “reading level” to grade 6-8, as is common guidance given to investigators. It appears FDA is signaling that we may be faced with having to make the consents even easier to read, and hence longer still. Unfortunately, a dozens-of-pages-long “consent” can have the opposite of the intended effect and be dangerous when inattentiveness or inability to digest such lengthy documents precludes a true comprehension on the part of the patient. As care and treatments become ever more complex, and regulatory oversight increases, the situation becomes worse, with concerns expressed by some that consent documents are becoming just too long for comprehension.1, 2 Indeed there has been published evidence to suggest that if a consent form exceeds four pages in length, it is unlikely to be read.3 Others have placed the tipping point at a mere 1000 words.4 Recently however, in a cultural and medical environment not markedly dissimilar to the United States, it was found that consent forms had increased from a median of 7 pages to 11 pages from the period 2000-2005, largely as the result of increasing regulatory oversight aimed at increasing patient safety.5
With respect to the latter point regarding precision, an example of the lack of precision that can occur in the quest for non-professional understanding can be found in paper used by Texas Tech University Health Services to guide the writing of informed consents6—in this document the authors advise writers to substitute “under the skin” for “intradermal” and yet a mere two lines later offer the same advice for the term “subcutaneous”; as any clinician knows, these two medical terms are hardly the same thing, although it could possibly be argued that to a patient the two procedures (assuming injection) are so similar as to appear the same. Perhaps more concerning, and within the same section as the previous example is the advice to rename “malignancy” to “cancer”, which is a term almost guaranteed to cause fear and alarm on the part of the patient, who may not be aware that many neoplasms (a term that certainly should be avoided in informed consents) are in fact benign. A critic of the foregoing comments might observe that obviously even the “lower level” terms would need to be explained. That critical observation would be justified, but also support the argument under development here – the danger in attempting to convey complex and precise information at a lower level is the result that simple but verbose statements must necessarily replace precision (witness my anecdote above and imagine the space it would take to actually write out the words said instead of “hand me the hammer to hit this bolt” – not only is the summary bad enough, but words such as “protruberance” “convex” and “oxidative” would have to be changed into lengthy explanatory phrases of their own!).
If making the reading level of Informed Consents yet lower, as FDA seems to be indicating is needed, is not the answer due to expanding length and imprecision, what is? I contend that FDA, rather thanpermit the giving of oral consents and explanations in 21 CFR 27 (b) (2) and essentially, per the Draft Guidance only in instances of non-English speaking, blindness or absolute illiteracy, and then of the same “quantity and quality” as the “long form” consent (that is, in essence, the long form is read out to the patient) (pages 19-20 Draft Guidance), actually mandate a shift in practice to require the giving of oral based informed consents. There is no person better positioned to communicate effectively to the patient the true meaning and nuances of a clinical trial and its alternatives than a fully trained clinical investigator, or a Registered Nurse (RN), if that person is equally trained on the study in question and has expertise in the given therapy area (this latter is not necessarily true of all research coordinators (RC), hence the differentiation between RN and RC). If a picture is worth a thousand words, then a conversation with a clinical expert who will be conducting the study must be at least as valuable, and perhaps more so. FDA should then mandate an oral consent process, and return responsibility for the relationship between physician/healthcare provider and patient back to those individuals rather than rely on a paper document. In this fashion, a tailored approach can be adopted and those who require very detailed, very simplistic and very thorough explanation of particular items of concern to them can receive it. Conversely, those who may require or desire a less time-consuming approach that still covers the essential elements can receive that – it is only through a patient-provider relationship and knowledge of the individual patient’s needs that such a judgment can be appropriately made. To give the impression that in order to combat decreased literacy levels further reductions in reading levels are necessary, rather than adopting an approach whereby patient understanding can be enhanced, is unlikely to be a successful strategy, especially when the very act of decreasing the “reading level” may not even be possible to convey the information required – a fact acknowledged by some leading IRBs.7
Our fundamental duty as researchers must have its basis in good ethical behavior. If there are means by which we can enhance patient understanding of what it means to be a subject, then we should actively pursue and consider adopting those means, as indeed the Agency is attempting with this latest Draft Guidance. It is likely that in a discussion process a healthcare professional can make a far better determination of a patient’s suitability for participation, their commitment to meeting their visit and treatment obligations, and their propensity to dropout or become lost to follow-up than in a process where a prospective subject merely reads a form, is given the opportunity to ask questions, and then signs. Costly patient attrition and trial delays may then be avoided.
Pursuing the same approach, but more of it, to Informed Consent is not the answer, if the NAAL research is indicative of the situation at large in 2014. FDA should seriously consider a distinct shift in thinking and a different approach.
The arguments presented here are the subject of a peer-reviewed paper where they are developed at much greater length with extensive references, charts and tables. Please click the contact information listed below to be notified when the full manuscript is available online. If you would like to submit comments to FDA on their Draft Guidance, please click here by September 15, 2014.
- Subscribe to our blog to stay informed on industry news and insights.
- Contact us for information on the informed consent paper when available.
1. http://www.cancer.gov/clinicaltrials/conducting/simplification-of-informed-consent-docs/page1/AllPages#fn123 accessed 22 July 2014.
2. Pandiva A, Readability and comprehensibility of informed consent forms for Clinical trials.Perspect Clin Res (2010) 1(3):98–100.
3. Albala L, Doyle M, Appelbaum PS. The evolution of consent forms for research: a quarter century of changes. IRB (2010) 32(3):7–11.
4. Sharp SM, Consent documents for oncology trials: does anybody read these things? Am J Clin Oncol (2004) 27(6):570–5.
5. Beardsley E, Jefford M, Mileshkin L, Longer consent forms for clinical trials compromise patient understanding: so why are they lengthening? J ClinOncol (2007) 25(9):e13–4 in Isles AF, Understood consent versus informed consent: A new paradigm for obtaining consent for pediatric research studies, Front Pediatr, 2013; 1: 1-4.
6. Winslow EH & Hagan P, Making Research Consent Forms More Readable, 1999, Revised 2003,https://www.ttuhsc.edu/research/hrpo/irb/files/informed_consent_article.pdf accessed 23 July 2014.
7. http://www.hopkinsmedicine.org/institutional_review_board/guidelines_policies/guidelines/informed_consent_ii.html accessed 23 July 2014.