Involving patients and the public in medical and health care research studies: An exploratory survey on participant recruiting and representativeness from the perspective of study authors

Research on patient and public involvement so far concentrates on defining involvement, describing its methods, and analyzing involvement practices in various individual research disciplines. There is little empirical data on the process of and aims for selecting (lay) PPI participants, and to what extend they can and should be representative of the population at large. To explore practices and perceptions on these issues and on future PPI conduct more generally, we sent an electronic survey to authors who published involvement activities as part of their studies in medical and social science journals. We identified such authors with a systematic search of five databases and applied descriptive statistics for analysis. Of those who returned the survey (n = 127 of 315; 40%), most had previously conducted involvement activities (73%). 45% reported more than one type of involvement, e.g. consultation and deliberation and participation (14%) and to have recruited more than one type of participant for their PPI activity (56%), e.g. 'lay publics' and 'expert publics' (33% of 71). Representativeness was often seen as a crucial objective when selecting PPI participants, while less than half found it very easy (9%) or rather easy (34%) to select participants. Many respondents considered achieving good representativeness difficult (52%) or very difficult (17%). They identified significant respective challenges and desired more guidance on various aspects of planning and conducting PPI (56%). 55% thought that the concept of involvement should be changed or improved. We conclude that recruiting lay people for PPI activities and deciding about and handling representativeness are controversial in current PPI practice, given the manifold challenges mentioned by the survey respondents. Our findings may inform further research particularly regarding the potentially many cases of unpublished PPI.