A cross-sectional analysis of low-cost non-cancer registry recruitment sources for a hybrid type 2 trial evaluating a digital intervention for melanoma survivors.

Abstract

Background: Traditionally, cancer survivorship behavioral intervention research has recruited from cancer registries, which can be costly and time intensive. Over the last decade, the advent of social media recruitment, hospital electronic medical record portals, and large provider networks has offered the possibility of non-paid recruitment sources to recruit cancer survivors cost-effectively. In this study, we used three non-cancer registry recruitment sources for a digital intervention to improve skin self-exams and sun protection among melanoma survivors. The objective of this study is to address critical knowledge gaps in melanoma research recruitment by comparing enrollment rates and participant demographics across digital and clinical sources, including social media, electronic health messaging, and dermatology practice network outreach.

Methods: Adults aged 18-89 with a prior stage 0-III melanoma diagnosis, 3 months to 5 years post-surgery, no current disease, and nonadherent to thorough skin self-exams were eligible for this study. Recruitment occurred through paid Facebook advertisements targeted at high-melanoma-incidence cities, unpaid social media influencer outreach, EHR messages, and a national dermatology practice network. An online eligibility screener was disseminated across all sources, with safeguards to prevent fraudulent entries. Individuals deemed eligible were contacted to confirm eligibility and completed oral and digital consent before enrollment. Enrollment rates and demographics were calculated.

Results: Of 230 participants enrolled from non-cancer registry sources, 122 participants (53.0%) came from a dermatology provider network, 56 participants (24.3%) came from social media, 46 participants (20.0%) came from EHR messaging, and 6 participants (2.6%) came from other sources. Social media recruitment yielded younger, predominantly female, more frequently employed and privately insured participants, and a higher proportion diagnosed at higher stages, compared with other sources. Despite recruitment success, over 1,700 fraudulent entries were detected and removed.

Conclusions: Non-cancer registry sources can support enrollment for digital intervention research, but each approach carries unique trade-offs. Provider networks delivered high volumes of eligible participants with minimal burden, while social media provided broad reach but required intensive fraud monitoring and additional safeguards. These findings suggest that combining provider-based outreach with carefully managed social media efforts may offer an effective balance between reach and data integrity when recruiting cancer survivors.