Background

A common method of recruiting for randomized trials is to send letters to potentially eligible patients inviting them to a screening appointment. In 3 consecutive UK studies the proportion attending from those invited fell from 49% [1] in 1994-1997 and 42%[2] in 1998-2001 to 13% at the beginning of an ongoing study in 2007. Procedures were similar in the 3 trials except that in 2007 the Patient Information Leaflet (PIL) was enclosed with the invitation letter. In order to understand whether the contents and/or style of invitation would explain the declining trend, 2 separate randomized comparisons were undertaken during the recruitment for the ongoing study.

Methods

Potentially eligible patients identified from hospital records were randomized to receive either an invitation letter enclosing the PIL (a 12-page A5 MREC approved booklet giving detailed information about the trial) or just a one page summary. A second comparison was made between a PIL modified after Focus Group discussions and the original PIL. Modifications included more colours, pictures and simplified language. The pre-specified endpoints for these assessments were the proportions of patients attending the screening visit, and entering the pre-randomization run-in period.

Results

Between July and October 2008, 20,759 personalized invitation letters were randomized to have the PIL or brief summary enclosed. There were no significant differences in either the proportions attending: PIL enclosed 1122/10,566 (10.6%) versus not 1181/10,590 (11.2%) [OR 1.06; 95%CI 0.97-1.16]; or in the proportion entering the pre-randomization run-in: 720/1181 (6.8% of those invited) versus 690/1122 (6.5%) [OR 1.05; (0.94-1.17)].

From November 2009 to January 2010, 12,164 patient invitations were randomized to enclose either the modified or the original PIL. A 17% higher attendance was detected for the modified PIL: 580/6104 (9.5%) versus 499/6060 (8.2%) for the original PIL [OR=1.17; (1.03-1.33): p=0.01). However there was no significant difference in the proportion entering the pre-randomization run-in: 373 (6.1% of those invited) vs 339 (5.6%) for modified versus original (OR 1.10; 0.94-1.28).

Conclusion

Whether the full PIL or brief summary was enclosed with the invitation did not affect the likelihood of attending or entering the run-in. Enclosing a more patient friendly PIL modestly improved the chance of attending, but not whether patients agreed to enter the study.