Background. Research has identified a wide range of evidence-based interventions for key behavioral health
risks such as poor diet, smoking, unhealthy alcohol use, or a sedentary lifestyle. However, to be truly
successful, behavioral intervention science should fulfill at least three key criteria: (1) demonstration of
cumulative increases in intervention efficacy; (2) provision of interventions that reach a high proportion of those
in need; and (3) demonstration of a meaningful population impact/reduction in disease burden. It currently
meets none of these. Some reasons for this include (1) over-reliance on an imprecise delivery mechanism
(people; usually therapists or coaches of some kind) that is difficult to train to fidelity on a large scale, and that
cannot be manipulated with precision; (2) use of sample sizes that may be far below what is needed to
accurately characterize heterogeneity of response; and (3) cross-study variability in therapist characteristics,
sample characteristics, and measurement strategy. All of this combines to create a science that lacks evidence
of cumulative improvements upon prior benchmarks. Proposed solution. Mobile technology shows significant
promise as an intervention delivery mechanism that is replicable, transparent, modular, and precise. However,
progress in the development and implementation of mobile interventions has been slowed by factors such as
the tremendous time and money needed to develop an intervention; limitations in cross-platform compatibility
and interoperability; and lack of a consistent system around which to collaborate. To address these needs, the
PI developed the Computerized Intervention Authoring System (CIAS), which facilitates behavioral
intervention science by allowing investigators to directly develop sophisticated and interactive mobile
applications without programming. CIAS is already being used by investigators outside of the PI’s own lab,
a process that has revealed significant interest in this software from a wide range of NIH-funded investigators,
as well as significant limitations. Current aims. The proposed application will address these limitations, making
CIAS into a significant, open-source, and virtually unique non-commercial research resource. In Aim 1, we will
engage in sustained user experience testing designed to make the tool far more intuitive to use. This process
is expected to result in a dramatic reorganization of the investigator interface, as well as in a complete set of
evidence-based user training and support materials. In Aim 2, we will add a range of features and capabilities
to make CIAS more powerful, flexible, and interoperable (e.g., by building to SMART Health IT standards, as
well as FHIR open specifications to facilitate integration with Electronic Health Records). In Aim 3, we will
engage in focused efforts to promote the use of CIAS as part of multi-lab collaborations using open science
practices (e.g., via integration with the Open Science Framework). Importantly, we will engage in all of the
above with the assistance of a highly accomplished panel of advisors who will help ensure that the final
product is broadly relevant, future-facing, and usable to a broad range of behavioral scientists.
