The 'patient information leaflet' (PIL) is a key source of medicines information for patients. EU legislation now requires that such leaflets are supplied with all medicines¹. However, there is little evidence that the leaflets are evaluated before implementation. The PIL could be content tested (e.g. against a checklist) or a readability score calculated. However, such methods gives little indication of what the medicine user will do with the PIL and how useful it will be to them. The EU now recommend an alternative approach: 'user-testing', where use is tested with people in the target group for the medicine (but who are not taking it)².

The aim of this research was to evaluate the process of user-testing. It involves asking the participant to find and to explain, in their own words, 15 pieces of information in the PIL. It is designed to be used iteratively during leaflet development, with a series of re-tests and amendments. The aim is a PIL where 16 out of 20 people can find and explain all 15 points. In this study, we used user-testing to test whether leaflets currently in use reach this target.

Method

We tested 3 PILs: for Levonelle (Schering) and 2 nifedipine preparations: Adalat (Bayer) and generic Nifedipine (Norton). Each was tested with 20 participants: consecutive attenders at a general practice (Levonelle) or community pharmacy (Adalat, Nifedipine) and in the target group for the medicine (Levonelle: women 18-49; Adalat, Nifedipine: people aged 65+). All could read, gave consent and had never used the medicine. In a quiet interview room they were asked to find and then to explain 15 pieces of information about the medicine.

Results

On the Levonelle PIL, the women found an average of 13.2 items (of 15) and correctly explained 9.2 (of 15) items. None found and explained all 15 items of information. Scores varied between items; some items were consistently found and explained less well.

Information in the Norton PIL was found and explained better than that on the Bayer PIL (finding means 13.3, versus 11.6; explaining means 11.5 versus 9.7). Two (of 20) participants found and explained all 15 items on the Norton PIL. No participant did so for the Bayer PIL. As for the Levonelle PIL, item scores on both nifedipine PILs varied greatly and consistently.

There was evidence of education effects on scores on all 3 PILs, with participants with educational qualifications doing better than those without.

Discussion

User-testing was easy to do, with a high consent rate and all participants completing the task. We were able to identify items in PILs that presented more or less difficulty to patients. It was also sufficiently robust to identify significant differences in scores between two PILs for the same medicine. Education level had an effect on scores and so this and perhaps other factors must be monitored carefully in user-testing to prevent biased scores. None of the leaflets came close to the target level of 16/20 users successfully finding and explaining all 15 points.

References

1. Dickinson D, Raynor DK, Duman M. Patient information leaflets for medicines: using consumer testing to determine the most effective design. Pat Educ & Couns 2001; 43: 147-159.
2. Sless D, Wiseman R. Writing about medicines for people. 2^nd Edition. Canberra, Communication Research Institute of Australia, 1997.