School of Healthcare FACULTY OF MEDICINE DENTISTRY

School of Healthcare FACULTY OF MEDICINE, DENTISTRY & HEALTH "Reading it makes my brain just want to shut down“ Improving the readability of clinical trial participant information sheets. Peter Knapp (With thanks to Theo Raynor, Jon Silcock & Brian Parkinson)

Consent to trial participation - context • Push to include more NHS patients in trials o How can trials recruit more participants? o (Jenkins et al, 2010: 83% cancer patients would participate) o Is valid consent compromised? • The issues for trials apply to research generally • …and to notions of shared decision making about treatments

Consent to trial participation - health literacy • Information sheets being good or bad is a matter of provision • …. ie imparting knowledge (functional HL) • However, NRES reasoning suggests something more… • Information should enable potential participants to ask…to question…to make an informed decision (interactive or critical HL) • Do potential participants and recruiters have a shared view?

Concern over consent § 40% did not know they could withdraw (Lynoe et al 1991) § 20% did not know the name of the trial drug (Griffin et al 2006) § 33% did not know the trial was primarily for research (Lavori et al 2007) § Important information not given; some aspects interpreted incorrectly by patients (Jenkins et al 2011)

Are Patient Information Sheets too difficult? § Understanding particularly poor in older people and fewer years of education (Sugarman et al 1998) § Written at the “college graduate level“ (Paasche-Orlow et al 2003) § Most require a level of literacy that is too high (Burman et al 2003) § Risks and benefits not understood well (Cox 2006)

After the TGN 1412 (“elephant man”) trial in 2006… UK Department of Health Expert Scientific Group, 2006 • Clarity of information communication between investigators & trial participants - extremely important and high priority Report on First in Man Studies - Royal Statistical Society 2007 • Good communication of risk essential to conduct of trials • Risk should be expressed numerically ABPI Guidelines for Phase 1 Clinical Trials, 2007 • Give subjects oral and written information that is free of jargon and easy to understand UK National Research Ethics Service, 2007 • Use simple non-technical terms a lay person will understand easily

How might we involve patients in improving information? User Testing in brief: Select key points • Relevant to safety and effectiveness Design & pilot a questionnaire which tests: • Finding each piece of information • Understanding (express in own words) Recruit 10 or 20 people from target patient group • People who are potential users or participants • Interviewed individually • For medicine leaflets: 80% to find and understand each point

Try, Try and Try Again User Testing is iterative and developmental • Draft material • Test material • Identify problems • Remedy problems, applying: • research evidence • good practice in writing & design • Test again

User Testing trial PIS: TGN 1412 § TGN 1412 Patient Information Sheet - Phase 1 trial of human monoclonal antibody - Sheet obtained from the web § 11 pages - >5, 500 words § Difficult language - ‘‘pharmacokinetics‘‘, ‘‘immunogenicity‘‘ § Poorly presented - Lack of useful headings

Applying User Testing to the PIS Research design: • • • Test original version Re-write and re-design Test revised version Amend as required Test further revised version • Finding & understanding for each point of information • Proportion of participants having a ‘clear round’

User Testing the TGN 1412 sheet Participants recruited: • Men aged 18 -40 • Range of occupations & education 21 key points selected • • nature and purpose of the trial (3) aspects of consent (6) trial procedures (7) safety and efficacy of the medicine (5) Questionnaire designed • Individual interviews • Q. What phone number should you ring in an emergency? • Q. How is it decided which treatment a participant receives in the trial? • Followed by opinions on the tested sheet

TGN 1412: Results Difficulty with 6 of the questions: § Presence of a placebo group § Action needed if medically insured § Pre-drug fasting § Informing GP of participation § Number of follow-up clinic visits § Emergency telephone number Would have failed if a leaflet for a licensed medicine

TGN 1412 Sheet Revision Design § Presented as an A 4 eight page folded booklet Wording § Added a summary of most important points and brief table of contents to first page § 10 section headings § Shortened sentences (where necessary) § Lay language § Removed text repetition § Related information brought together under relevant heading § Use of ‘bullets’ to indicate lists.

Clinical Trial Patient Information Sheets

Revised wording Each treatment regimen will involve slightly different timelines Each type of treatment lasts a different amount of time First time into human study This is the first time it will be tested in humans

TGN 1412 PIS testing paper Improved scores on finding and understanding information; No impact on reading time. P Knapp, DK Raynor, J Silcock, B Parkinson. Journal of Medical Ethics 2009; 35: 573 -578.

Poor Responders? § The ‘Poor Responders’ Phase 3 trial of In-vitro Fertilisation § For women for whom initial IVF treatment had not been successful § 7 pages § Information sheet obtained from the published paper § Study design: test original, then revise, test revised

‘Poor Responders’ Trial: participant comments (1) Original version Medical terms: "The technical terms are a bit hard to understand" (P 6); "It's a bit confusing to work out which drug's having what effect on the body" (P 17). Introducing the study: "There should be more information about what's involved in taking part in the study at the beginning. . . it launches into quite scientific explanations" (P 13); "It's quite confusing to begin with all the FSH, Gn. RH and so on" (P 16).

‘Poor Responders’ Trial: participant comments (2) Original version Layout: "The way it is laid out is very intimidating" (P 26); “Reading it makes my brain just want to shut down straight away" (P 25).

‘Poor Responders’ Trial: participant comments (3) Revised version Appearance: "It doesn't look like it's a scary thing to read" (P 40); "It looks. . . like a leaflet you'd want to pick up and read" (P 37). Organisation & wording: "If you look on the front and you need to find out where stuff is you go through each section and I think the headings just sum up what's in it" (P 33); "(There are) no huge big medical words there, they were explained in laymen's terms" (P 28); "(It) seemed to flow. . . ” (P 38).

‘Poor Responders’ Trial: participant comments (4) Revised version Appearance: "It makes you feel like it's not just come from a printer, it's been produced with a lot of thought and care" (P 24); "Possibly it seems like there is less information in (revised version) but I think actually it gives you more. Because of the way it's laid out you pick up more" (P 22); "If I was paying for my own treatment I may prefer (the original version) because I may think that all my money is going to produce glossy pamphlets like (revised version), but in terms of being easier to work your way through (revised version) is better" (P 31).

Phase 3 trials testing papers ‘Poor Responders’ Trial: P Knapp, DK Raynor, J Silcock, B Parkinson. Trials 2009; 10: 79. SHARP (cholesterol reduction intervention) trial: P Knapp, DK Raynor, J Silcock, B Parkinson. British Pharmaceutical Conference, 2009. (Abstracted into Int J Pharmacy Practice)

3 trial sheet studies… Promising results, but: • Small numbers of participants • Relatively weak study design (can we sure the ‘original’ and ‘revised’ participants are similar? )

RCT of original vs revised versions of a PIS: AML 16 Research design: • Test original version • Re-write and re-design • Test revised version • Then run parallel groups RCT of original vs. revised versions • Powered to find a doubling of the proportion of participants having a ‘clear round’ (finding and understanding all points of information) • n=116 • Stratified by age; education; previous experience of testing; regular use of documents • Participants sent sheets at least a day before testing

RCT of original vs revised versions of a PIS: AML 16 trial: series of Acute Myeloid Leukaemia trials • MRC-funded • AML 16 focussed on older people Participants: • 123 randomised; 116 tested (55 original; 61 revised) • • • Male 43; female 73 Aged 55 -85 (mean 64. 9) 29 (23. 6%) educated to graduate level 36 (29. 3%) used written documents in their work 83 (71. 5%) had attended readability testing before

RCT of original vs revised versions of a PIS: AML 16 ‘Clear round’ (finding & understanding all points): • Original 8 / 55 • Revised 40 / 61 • Chi 2= 31. 5; p<. 001; Odds ratio = 11. 2 Reading time (minutes): • Original 27. 6 (sd 11. 4) • Revised 28. 4 (sd 12. 9) • F = 0. 30; p=. 86 Participant preference: • Original 15 / 116 • Revised 101 / 116 • Sign test p<. 001

Testing clinical trial information sheets – where next? • Consent can be improved with brevity and clarity (Jefford & Moore 2008) • But systematic review: amending information does not increase recruitment (Caldwell et al 2010) • Does one offset the other?

Testing clinical trial information sheets – where next? • What are the effects within an actual trial? (working with U of York CASPER trial) • What impact is seen on participant recruitment and retention? • How should we assess the impact on consent, understanding, recall, etc? (ie what patient outcomes matter most? ) • Can HL help to inform this work?