Browse/search for people

Publication - Professor Richard Martin

    Factors associated with trial recruitment, preferences and treatments received were elucidated in a comprehensive-cohort study

    Citation

    Donovan, J, Opmeer, B, Young, G, Mills, N, Martin, R, Lane, JA, Metcalfe, C, Peters, T, Davis, M, Turner, E, Walsh, E, Neal, D, Hamdy, FC & , 2019, ‘Factors associated with trial recruitment, preferences and treatments received were elucidated in a comprehensive-cohort study’. Journal of Clinical Epidemiology.

    Abstract

    OBJECTIVE: Recruitment to pragmatic trials is often difficult, and little is known about factors associated with key participation and retention decisions. These were explored in the ProtecT (Prostate cancer testing and Treatment) study.

    STUDY DESIGN AND SETTING: Baseline socio-demographic, patient-reported outcome, clinical history, and prostate cancer biopsy data were collected for all patients eligible to take part in the ProtecT trial, in a comprehensive-cohort design. Men who rejected randomization specified a preferred option and were followed-up identically to the randomized cohort. Factors associated with participation decisions, patient preferences, and reasons for changing treatment were explored.

    RESULTS: Of 2,664 men with clinically localized prostate cancer, 997 (37%) rejected randomization. Their treatment preferences, and subsequent treatment choices/changes in both randomized and treatment-choice cohorts, were strongly associated with prostate cancer risk-features: toward active monitoring for low-risk disease and toward radical options with higher-risk prostate cancer. Among many factors measured, only a small number of weak associations were found for occupation groups and some patient symptoms. Similar percentages changed from the random allocation and initially stated preference.

    CONCLUSIONS: The comprehensive-cohort design provided new insights into trial recruitment and participation decisions. Opportunities to improve recruitment by supporting recruiters with equipoise and patient preferences were identified.

    Full details in the University publications repository