Non-participation in clinical research: Barriers, motivators, and recruitment strategies in an ovarian cancer study

Clinical studies have made significant contributions to disease prevention, early diagnosis, and new therapies for many diseases, including cancer. However, recruitment of participants is a major challenge in clinical studies on cancer; older adults, minorities, rural residents, and individuals of lower socio-economic status are traditionally difficult to enroll, despite these groups having higher cancer incidence and mortality. The ongoing clinical study in Montreal (Canada)—Diagnosing Ovarian cancer Early (DOvE)—aims to determine whether providing fast-track diagnostic testing to symptomatic women over 50 leads to early diagnosis and better prognosis. Results from the pilot phase showed a 'high' prevalence of ovarian cancer among participants and indicated an increased likelihood of diagnosing ovarian cancer early while still completely resectable. However, these promising findings were observed in a highly selected population, with a higher proportion of younger, highly educated and Anglophone women, which raised the concern of volunteer bias. The benefits of the intervention could be overestimated if those who volunteered to participate in DOvE would have been diagnosed earlier even in the absence of DOvE (the study has no control group, given that all participants have symptoms). The overarching objective of this manuscript-based thesis is to examine different aspects of non-participation in clinical cancer studies. To characterize non-participation, it is essential to have knowledge about the underlying target population. However, in the case of DOvE, the target population was unknown (i.e. women aged 50 and older living in Greater Montreal and having symptoms with the specified duration). Thus, the first paper was to estimate the prevalence of symptoms determining eligibility to DOvE and their associations with socio-demographic characteristics. The object of the second paper was to identify factors associated with intention to participate and investigate motivators and barriers among potential DOvE participants. The third paper was to evaluate the effectiveness of strategies for increasing participation of underrepresented groups. Between May 2011 and April 2014, DOvE investigators opened five additional centers in areas with a dense population of older, Francophone women. The third manuscript examines the success of this strategy in terms of participant characteristics and study accessibility.A postal survey was sent to a random sample of 3000 women aged 50 and older living in Greater Montreal, with up to two reminders to non-responders. Due to the low response (28%), the inverse probability weighting was applied to correct for non-response. Older women (70+) were less likely to respond and, among the responders, less likely to report any symptom. Prevalence of symptoms was high, even when limiting duration to a specific time window (59.7 % reported at least one symptom), and the crude and weighted estimates were very similar. Having more symptoms and not having a family doctor was significantly associated with intention to participate, while being older, living in rural areas, having lower income and being in excellent health was associated with intention to not participate. The majority of those who did not plan to participate preferred to be examined by their own physician or thought that they were not at risk of ovarian cancer. "Inconvenience" was also a commonly reported reason. Strategically placing satellite centers across the city improved accessibility and resulted in a less selected study population.The findings of this thesis not only provide us with a better picture of the underlying population for DOvE, but also suggest several strategies with the potential of enhancing recruitment of hard-to-reach groups. Efforts to include the subgroups that disproportionally experience higher rates of cancer in clinical research continue to be an important issue in need of further research.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.