Why consent forms fail the people who read them
A consent form has to carry everything the law requires and still be understood by the person who decides from it, and often it is not understood. The failure starts with who that reader is, shows in the evidence that forms are too hard to read, grows as forms lengthen, and compounds when the form is translated.
Who the form has to work for
Medical writers prepare the consent form in parallel with the protocol, but the protocol is read by doctors and nurses and the form by patients, so the form needs a very different kind of language. The form draws on complex information from the protocol and related materials, so it has to be adapted and recontextualised for a lay reader. Its readers may often be elderly, they may have a mental illness, they may be children or the parents of children, and they come with different levels of education and different severities of illness. A patient lying very ill in a hospital bed needs a document that is very clear and written in language they can understand. At the same time the form is legally required, so it has to carry all the information the patient needs to understand what will happen to them if they agree to take part, what the risks are and what the advantages will be. The patient signs it to say they agree.
A potential participant is often in distress when given the consent document, which puts them in a vulnerable position that may detract from their ability to understand complex written information. Many people who seek to take part in research have a low literacy level, may not be native English speakers and may not be well versed in science.
One participant in several clinical trials, a bioengineer with a PhD who understands many complex terms, found his consent document very dense and still sometimes wondered what its terms actually meant and how to apply what it said to his own life.
Readability of forms as written
Consent forms have been found to tend to be long and difficult to read and understand. A growing body of research shows that consent forms often exceed what their lay readers can understand. Several factors contribute, from length to a lack of visual aids, but studies claim that the complexity of the text itself is the main obstacle to comprehension. Specialised terminology and jargon, sentence length and structure, and the volume of information all play a role in limiting how well readers process what a form contains, and this in turn makes the form less effective as a vehicle for an informed decision.
The forms tend to include excessively long sentences, with multiple subordinate clauses in equally long paragraphs, which are typical of expert-facing texts but not appropriate for lay readers, who may struggle to follow them. The sentences can also be incoherent or inconsistent in order and structure.
Lexical complexity, distinct from technical jargon, is the other wording and phrasing that makes the forms difficult to understand, and it includes, among other issues:
- nominalisations and a disproportionate use of nouns over verbs, often with the passive voice, as in the investigation will be conducted for we will investigate
- compound nominal and adjectival phrases, such as phase IV, multicentre, open-label, double-blind, randomised clinical trial and recurrent major depressive episodes
- phrasal verbs, such as follow up, fill out, slow down and rule out
In one study of a sample of the Italian and English consent forms in use for seven common cardiology imaging examinations (coronary angiography, percutaneous coronary intervention, myocardial perfusion imaging, cardiac positron emission tomography, cardiac computed tomography, cardiac radiofrequency ablation and stress echocardiography), the forms were poor in overall quality and readability, too complex and poorly organised, and did not properly highlight the most relevant information. The forms were analysed against the recommendations of scientific societies, with quality judged on three criteria: content and its organisation, text construction and layout, and the development process. Once revised using reference standards such as the federal plain-language guidelines and analysed again, the forms were more readable.
Length of forms
Given their legal and ethical functions and the health risks a participant consents to, consent forms contain a substantial amount of information, from medical processes and procedures to legal provisions on the individual's rights and data protection. Their authors include too much technical detail, partly to comply with established conventions for the forms and partly to avoid any potential liability by covering every base. The resulting wall of text can have the opposite effect, in that readers might struggle to process and retain essential trial information. Repetitions, verbatim or reworded, are also common and may confuse readers.
Trial designs are becoming more complex, with tests, procedures and questionnaires evaluated at multiple visits, and this lengthens the form. The form now carries a lot of medical information besides the sections regulators require on personal data protection, under the General Data Protection Regulation for trials in the European Union or the Health Insurance Portability and Accountability Act for trials in the United States. The need to include all the information regulators require on top of what ICH good clinical practice outlines makes the form bulky and lengthy. One editor of consent documents at a US cancer centre holds that strict federal regulations on the elements of informed consent a consent document must include make it difficult to communicate clearly with patients, and notes that the requirements of the Health Insurance Portability and Accountability Act have added many pages to consent forms.
A driving factor behind the length of consent forms is the lack of standardised guidelines on what they should look like. Although the principle of informed consent is recognised in international agreements and legislation, no overarching legal framework at the European or international level codifies exactly what information a form should contain or how it should be written in content, style and layout. Some international bodies, such as the WHO, provide templates, but there is no legal obligation to use them. Most pharmaceutical companies and contract research organisations running trials have their own templates and conventions, usually not publicly available, which they use in compliance with the requirements of regulatory authorities. These templates generally include similar types of information but vary from one company to another or across countries or regions. Nor are they standardised the way the European Medicines Agency's centralised procedure package leaflets are. Those leaflets, though imperfect for readability, help ensure uniformity across documents from different companies, manage expectations about what the documents contain and look like, and provide a tool for checking whether those expectations are met and, if they are not, for holding those responsible for running the trial accountable. Without such guidelines, the authors of consent forms and their translators face uncertainty and variation, which makes it all the more difficult to uphold and implement measures that help lay readers understand.
One organisation's forms show the length growing over time. Across the trials it ran over a ten-year period, the average consent form was 17 pages for pneumonia studies and 30 for breast cancer trials. For one investigational product in multiple myeloma, the master form grew from 12 pages in 2016 to 18 in 2017 for another study with the same product, and 25 in 2020, for the same drug and indication. The organisation first assumed the growth came from safety data accumulated over four years of trials with the product, but that assumption appeared to be wrong. The 2020 form carried more detailed information than the 2016 form on the procedures at each visit, such as Eastern Cooperative Oncology Group (ECOG) status assessment, copied exactly from the protocol along with medical terminology that makes no sense to a lay reader, such as "increase in blood level of creatine phosphokinase that could suggest heart damage". Its paragraph on the risks of the glucocorticoid dexamethasone ran to 15 lines, against 6 in 2016. Translation lengthens the form further: the 25-page master became 42 pages in Spanish for Argentina and in Malay for Malaysia, and translations into other languages ran to 34 pages for Hungary and Taiwan and 31 for Poland.
Section 4.8.10 of ICH E6(R2) says the form should explain the trial procedures to be followed, including all invasive ones, and in practice this is read as an instruction to list every study procedure in the form. But the protocol describes the procedures for the investigator and site team, and transferring that information directly into a document for patients seems unacceptable. Where the protocol requires ECOG status assessment at each visit, for example, there is no sense in mentioning it in the form, because it is the assessment of the patient's ability to carry out activities and care for themselves, which the site team does as part of the routine examination and which asks no additional procedure of the investigator or action of the patient. Section 4.8.6 says the information should be as nontechnical as practical and understandable to the subject. The requirements of ICH E6(R2) are not strictly followed, and the result is forms containing information patients find hard to understand, such as medical terms.
Excessive length may delay review and approval by ethics committees and regulatory authorities. Deficiencies in a form's content and structure might also harm patient enrolment, which would prolong study timelines.
Translated forms
In Spain, the participant information sheet and the informed consent form must be written in the subject's own language, so their translation is a mandatory legal step in the approval of any multinational study. Consent forms there are almost always translated from English, and some have run to more than 10,000 words. Their final readers are lay people, so the translations should be clear and understandable as well as accurate and adapted to the target legal and social framework. Yet professional translators often focus primarily on how faithful the translation is to the original rather than on how understandable the translated version is, forgetting that a message that works well with an audience speaking one language often may not work for an audience speaking another.
Where consent forms are translated and then back-translated to ensure that they keep their original meaning, one view holds that the emphasis on literal translation serves largely to satisfy the legality of the process rather than the information and comprehension needs of the community or of the individuals who may take part in the research.
Calque, a borrowing from or interference by the source language, can occur at the level of the word, producing word-for-word literal translations, or at the level of the sentence, where the translation follows the sentence structure of the source language and reads as stylistically unnatural and awkward. These problems help explain why ethics review boards complain about the poor readability and comprehensibility of consent forms and their translations.
A study in Spain of 101 sequentially selected participant information sheets and consent forms, a good proportion of them likely translations since the sample covered all trials approved in the previous two years, found that 97% required readers to have secondary to higher education. The study concluded that the documents were unacceptably difficult for readers, a situation that might even affect the validity of the consent process. Experts in Spain have also complained that many information sheets are poor translations from English, done by people who lack the necessary clinical experience, so that ethics committees are forced to review and rewrite them. Others have complained that consent documents are usually literal translations that are too long and carry too many technical words, which make them difficult for patients to understand, and that problems with terminology may come from translations that are too literal or not adapted to the local culture.