Feasibility and Design Challenges for a 10-Year Treatment Follow-up Study Examining the LONG-TERM Effects of EARLY Intensive Behavioral Intervention

Background: In 2008 we conducted a novel, prospective study testing predictors of developmental outcome (IQ, adaptive behavior, ASD severity) after 12/24-months of early intensive behavioral intervention (EIBI) in a community-based sample of 71 children with ASD (age 20- 59 months). Predictors included social engagement and sensory-motor rituals. Generally, children with higher IQs had better outcomes. After controlling for age/ IQ/ EIBI site, social engagement but not sensory-motor rituals predicted IQ and adaptive functioning at exit. These data had implications for predicting short-term outcomes and better developmental trajectories in preschool. However, little is known about the long-term impact of EIBI, and whether early predictors continue to predict outcomes in adolescence (IACC, 2013). To address this gap, we conducted a 10-year longitudinal follow-up study designed to evaluate and compare the outcomes of children who originally received EIBI versus treatment as usual (TAU).

Objectives: 1) Report the feasibility of a 10-year follow-up study that was not planned a priori; and 2) Discuss the unique design challenges and experimental decisions that were made to help guide future research. We will address recruitment challenges, selection of assessment measures, and unique features of test administration.

Methods: From the initial pool of 136 participants (71 EIBI, 65 TAU), 110 met eligibility criteria for the 10-year follow-up study, at age 14-18 years. Recruitment letters were mailed; some participants did not have valid contact information and did not return calls (invalid numbers or ‘lost to follow up’). Additional recruitment methods were utilized (social media, flyers, last letter, data pull from EMR). Enrolled participants completed measures of cognitive ability, academic achievement, language proficiency, and autism diagnostic status. These included adolescent measures, parent-report measures, and teacher-report measures. Enrollment is ongoing until June 2019.

Results: To date, thirty-eight of the original 70 EIBI participants, and 22 of the original 65 TAU participants enrolled in the study (see CONSORT diagram). Feasibility outcomes: so far, we met goals and recruited approximately 54% (EIBI) and 34% (TAU). Recruitment barriers were addressed using a wide range of strategies, including letters, flyers, social media, and word of mouth. We conducted assessments in the home to minimize transportation barriers, and scheduled visits during vacations that brought participants to our region. Assessments were selected to maximize the validity of comparisons to our original study, without overwhelming participants.

Conclusions: A 10-year follow-up study can provide important information about long-term treatment outcomes. In the literature on follow-up studies in the US, re-enrollment rates are highly variable; with many studies reporting percentages in the low sixties to low eighties. Although we are still enrolling, our current rates may reflect inherent challenges for parents and children with ASD, who are reported to have more stressors and daily hassles than parents of other populations. To promote successful longitudinal studies, specific methodological challenges must be addressed proactively (e.g., recruitment, visit accessibility, selection of measures). Lessons learned from the current study can help to guide future researchers in preparing proactively for design challenges that may arise