Randomised controlled trial assessing effectiveness of health education leaflets in reducing incidence of sunburn

Subjects, methods, and results

The study population comprised holidaymakers travelling on Air UK Leisure flights from Manchester airport during August 1993. The unit of randomisation was the flight. Flights were stratified into long haul (North America and Jamaica) and short haul (Europe) and randomly allocated to the intervention or control arm. Before boarding, the health education authority leaflet If You Worship The Sun, Don't Sacrifice Your Skin was placed in seat pockets on flights in the intervention arm but not in the control arm. Cabin crew distributed questionnaires to passengers on Air UK Leisure return flights to Manchester. A history of sunburn was elicited by the question “Did you suffer from any sunburn during your recent holiday?” and, if so, whether this was associated with one or more of: redness of the skin, blistering of the skin, pain for less than a day, pain for more than a day. Adults completed the questionnaire for children. The study endpoint, severe sunburn, was defined as any episode of sunburn which was either painful for more than a day or resulted in blistering. Randomisation by group was undertaken to reduce contamination between the study arms. A clustering parameter was calculated for the study endpoint. Brier's adjusted χ² was used for baseline comparisons,3 and 95% confidence intervals were constructed for the difference in proportions using methods appropriate to group randomised trials.4

Sixteen long haul and 62 short haul flights were randomised to the intervention arm and 15 long haul and 62 short haul flights to the control arm; 21611 questionnaires were distributed and 14956 (69%) returned. A total of 2483 questionnaires completed by passengers who had not departed from Manchester airport during the study period and 88 inconsistent or illegible questionnaires were excluded from the analysis, leaving 12385 evaluable questionnaires. The clustering parameter was 0.02. The study had a power of 90% to show a 5% difference between the two groups at the 5% two sided significance level. There was no significant difference between the two groups in the distribution of baseline characteristics or the proportion reporting severe sunburn (see table).

Comment

This leaflet did not seem to reduce the incidence of severe sunburn. Passengers were not asked if they had seen or read the leaflet as this might have influenced their response to the questionnaire. Therefore it is not possible to determine if the intervention failed because subjects had not read the leaflet, they had ignored its messages, or the messages were inappropriate. However, unsolicited distribution of leaflets is a common practice.

Randomised controlled trials provide the most compelling evidence of effectiveness. We have found no other randomised controlled trials of the effectiveness of health education leaflets in the prevention of sunburn and few trials of health education leaflets in other settings.5 Although leaflets are perceived as low cost and unlikely to cause harm, there are substantial opportunity costs associated with promoting strategies of unproved effectiveness. We commend the more extensive use of randomised controlled trials in the evaluation of primary prevention programmes.

We thank the management and staff of Air UK Leisure and Manchester airport for their invaluable help in administering the study and the Health Education Authority for supplying the leaflets.