Communicating Risk (pt 1) - Patient Information Forum event: health literacy and data visualisation sessions

Sorry for any typos, written in haste before heading off to see chums so am just getting it up online and will correct any errors later :) This is just my notes from the first two sessions as I'm about to get on a train.

---

Last Tuesday I went to PiF (Patient Information Forum)'s event on communicating risk in health information (you can download the presentation slides). It was very good and inspiring and I thought I'd type up some of my notes from the first couple of sessions. 

Incidentally I retweeted this picture a year or so ago which I think is interesting, on the use of everyday words and how they can be misunderstood (although jargon words are often incomprehensible they do have the advantage of heralding that you might not understand them, whereas an everyday word might make you falsely believe that you have - in a way plain English might actually cause more problems!). I've written about this elsewhere on this blog.

From a paper about communicating climate change, PhysicsToday 2011

Tuesday 26th January 2016, 10.30-4pm (with very nice food) at NCVO offices, King's Cross

Session One - Janet Solla, Community Health and Learning Foundation
The first session was from CHLF and it touched on a topic I'm particularly interested in - people being able to understand information about health matters, including the language used.
[slides, PDF]

She talked about the sheer numbers of people with very poor basic literacy and abysmal mathematical literacy (numeracy), and how there's less stigma about being vaguely innumerate ("I'm no good with numbers, me") compared with that for being unable to read. Confidence is also an issue - if you're feeling that you know less than everyone else you're perhaps less likely to pluck up the courage to ask and get help.

There was a sad example of someone with a problematic cough whose GP referred him for a hospital X-ray, but on arriving at the hospital he was presented with signs saying Oncology, Paediatrics, Pathology, Radiology, Gynaecology etc but nothing saying 'X-ray'. So he went home and didn't get it done.

People who have lower levels of health literacy may also find it harder to engage with decisions about their healthcare - another example was given of a gardener that might be invited for a knee replacement operation in summer (when it would be better for them to have this non-urgent surgery and the recovery period in winter as summer's a busier time).

Similarly it's usually the well-engaged and literate (or health-literate) who get involved in consultations about service changes, so there's poorer representation of people who aren't generic middle class.

She also gave an example of a patient who was told that his blood test was positive and who took this to mean that all was well so was baffled when the doctor began talking about treatments. A woman receiving chemotherapy infused into one arm (which happened to be the same side as the breast cancer for which she was having the treatment) became distressed when the infusion site was switched to the opposite arm - worried that the drugs wouldn't get across to the other side.

This also made me realise how difficult it is for someone who is super-literate to imagine the many ways in which their kindly interventions might fail, and how things might be misunderstood. Training and awareness may be needed.

I thought of the recent news story that medical student applications tend to come from the most privileged pool of people and risks future healthcare teams that aren't representative of a lot of the people they're serving. The speaker, who does run training and awareness sessions for healthcare professionals about health literacy, pointed out an example of a doctor who'd felt that 'dumbing down' information for people was patronising (no, making that information accessible to more people is a good thing, it can be made clearer, not dumbing down) but sadly another thought that if people weren't prepared to learn to read etc then it was their own fault. This is a worrying attitude.

Even those of us who are plentifully literate might be marinated in fear when we find we're in a hospital being given bad news. It's easy to forget important information if you don't write it down (or are unable to, or didn't have the planning ability to remember to bring pen and paper).

Session Two - Eluned Hughes, Breast Cancer Now (formed from merger of Breakthrough Breast Cancer and Breast Cancer Campaign)
[slides, PDF]

This was a great session on using data visualisation to explain risks, particularly related to the risks and benefits of breast cancer screening for women - what should they be told about screening? Eluned's team use 'icon arrays' (think of lots of those little men and women icons that you find on the front of loo doors) which let them show numbers of people affected by a particular thing, and how taking an intervention (or not) might affect those numbers.

Eg 200 icons showing women in a population, as a 20x10 grid, then make 15 of them blue (scattered among them rather than the last 15 in your grid, the scattering can also be used to indicate randomness of breast cancer incidence) and then add a red one to indicate that in every 200 women 1 extra might experience a particular problem if they take this drug or don't take that one.

She also stressed the importance of talking in absolute numbers rather than relative risk (nothing wrong with both, but need absolute) and to make sure that if you use percentages that you also have the "17 in 200" explanation there too. Also keep your denominator (the 200 in the previous example) the same, so you're not talking about different things which adds confusion.

Eluned also thought that these data visualisations didn't have to be too shiny, nothing wrong with keeping things simple and perhaps even a little rough around the edges, don't overengineer. She also recommended actively using words like 'roughly' and 'about' to highlight the uncertainty of data.

Their visualisations are tested for understanding with various audiences including people with breast cancer (as well as experts in screening and risk communications). David Spiegelhalter was speaking later in the day and he thought these icon arrays were a clear way of communicating risk information.

Breast Cancer Now use a mixture of online and printed tools to reach different audiences. I piped up in the Q&A that when working at Diabetes UK we'd surveyed people and found that many of them were given our information by their nurses rather than contacting us directly for info, so if you're thinking about how to reach your target audience you can struggle to do that if you're not also working with the gatekeepers.

There were some good questions from the audience about how you recruit people for testing, how you evaluate the impact, what the survey questions might include (and if people's knowledge is tested or if they just like and understand the information), what decisions did the people make after using this information and is the tool available in languages other than English. Alas I made note of the questions but not the answers, and I'm just about heading out the door... sorry!