Recruitment Strategies for SPARK in Oregon

Background: Given the heterogeneity of the ‘autisms,’ very large sample sizes are required to research more homogeneous subgroups stratified by common clinical characteristics and/or by genetic profiles. Simons Powering Autism Research for Knowledge (SPARK) was launched in April 2016 with the goal of assembling a research cohort of 50,000 individuals with ASD recruited by 25 academic centers in the USA, including Oregon Health & Science University. The priority was set on recruiting trios (ASD individual + 2 biological parents) who can then be genetically characterized using whole genome sequencing.

Objectives: OHSU established a number of strategies to disseminate information and encourage ASD subjects and their relatives to sign up for this novel national registry. We targeted newly diagnosed/follow-up patients in the Autism Clinical Program (ACP-p), OHSU patients having a previous ASD ICD code in electronic medical records (OHSU-p), we connected and worked with community organizations, liaised with statewide special education professionals, ran local social media campaigns, and organized special events out of state (OOS) to reach out to a larger population base. Here, our objectives were to identify which recruitment strategies succeeded in achieving completed trio participation (completed registration process, self-reported phenotypic characterization, and successful collection of 3 saliva samples for ongoing genetic analysis).

Methods: Several metrics are regularly issued for each site including: 1) individual accounts created; 2) ASD probands having consented to genetic sample collection; 3) ‘potential’ trios (missing some consent data forms and/or biological samples); and 4) ‘completed’ trios (all trio data including usable biological specimen received centrally). Of note, some participants and trios may have joined SPARK without sites being properly credited for their recruitment.

Results: Over 18 months, 2,280 accounts were created and allocated to the OHSU site. Of these, 1,181 ASD probands did consent (or assent) for genetic analysis, including 963 children, 135 independent adults and 83 dependent adults. We could form 801 ‘potential’ trios of whom 306 reached the final stage of full participation in SPARK. The process of registration lasted on average 110 days for ‘completed’ trios. The recruitment source for the 306 completed trios was: 33.2% for OHSU-p, 24.8% for OOS, 22.5% for ACP-p, and 19.5% for other sources. During the study period, approximately 375 patients were seen in the ACP for diagnosis/follow-up and individually targeted for SPARK participation. Less than 20% of those patients completed the registration despite repeated blast e-mails, systematic inclusion of flyers in family feedback sessions, banners posted in the hospital and in waiting rooms. By contrast, social media campaigns and out-of-state lectures/events accounted for a larger proportion of our recruitment (Figure). Recruitment analysis using other metrics and statistical models will be presented with longer run of data.

Conclusions: Considering Oregon epidemiological estimates (about 15,000 ASD subjects <20 years) and the pre-eminence of the OHSU ACP, relatively few ACP patients enrolled in SPARK. This is in marked contrast with research participation in other childhood diseases (e.g. cancer). Qualitative research to explore reasons for participation or not by our patient population is needed to address this gap.