First published January 29, 2019; updated June 20, 2019
It is generally understood that it is difficult to translate ASD research into better outcomes for children with ASD if they are not promptly and properly diagnosed. In this context, our persistent failure to close gaps in timely and accurate identification of ASD in children is truly remarkable. Consider the Centers for Disease Control's most recent update regarding ASD's prevalence: the wide variations in reported prevalence from state to state (from 1 in 34 in New Jersey to 1 in 77 in Arkansas), the high median age of the earliest diagnosis (52 months), and the low percentage of children who had received a comprehensive developmental evaluation by 36 months of age (42%).
These and other marginal improvements in ASD identification over the past decade are especially alarming given the increasing investments in research in ASD. Between 2008 and 2015 alone, more than $2.5 billion was mobilized for research in the US through the Combating Autism Act and other related campaigns, with $277 million dedicated specifically to improving ASD identification. How has this happened? Do we know where we are going? How we will get there? What kind of tools we will need? When we can hope to see delays in diagnosis really begin to decrease for all children, regardless of who they are or where they live? In another 5 years... 10 years... 20 years... ever?
One answer may lie in the kind of research we have prioritized. In a series of presentations in 2018 and 2019 at the International Society of Autism Research (IMFAR), I reviewed all projects focused on ASD identification, and found that a remarkably small proportion actually involved applied research intended to immediately improved identification. An even smaller proportion actually yield publications in peer-reviewed journals that reported the results of the project's principal objectives. The gaps in applied research are part of a broader pattern of research funding that has heavily favored basic research (e.g., into the causes, characteristics, and course of ASD).
So what can we do? One solution lies in the creation of a roadmap that describes precisely what we need to learn from research to translate better understanding of ASD into better practice and better outcomes. In this series of essays, I outline the rationale and structure for such a research roadmap to address ASD screening. This roadmap focuses on research conducted on the most widely-studied and recommended tools for ASD screening; those derived from the Checklist for Autism for Toddlers (CHAT), and its variants. This includes the original CHAT, the Modified CHAT (M-CHAT), and the Quantitative CHAT (Q-CHAT), grouped together here as the CHAT family (CHAT+). This roadmap offers an important new perspective for viewing research by breaking research down into specific stages and phases to track progress and identify gaps. This can help program and policy leaders set new priorities and re-allocate resources as needed, ensuring that research can more effectively help people with ASD to achieve important life outcomes.
Listed below are individual essays in this series: I draw out a comprehensive research roadmap, describe the underlying rationale and structure, and illustrate how the stages of research are aligned with other elements in an integrated network (like training, policy, and services). Other essays will highlight the milestones and watersheds between different stages of research (for example, from basic to applied research) and the phases within each (for example, from clinical to implementation research within the stage of Applied Research). These also reveal how each stage of research requires different resources and expertise. Finally, we have begun to pilot the roadmap through a systematic analysis of a specific body of research: applied research studies into the CHAT+ family of screening measures.
This series of essays is written to reach beyond researchers and experts in ASD, to a broader audience of professionals and policy leaders. It is intended to supplement other presentations and publications in preparation for a more traditional academic audience. This roadmap sets the stage for other studies described in greater detail elsewhere on this site, including a review of all $277 million of research funded between 2008 and 2015 to improve ASD identification in the US.
Driving without a roadmap is never a good idea! We might know where we want to go, but we do not know how we will get there. Maybe we do not end up driving recklessly off a cliff in pursuit of a dream, but we just drive slowly to the only possible destination we can imagine... more OJ than Thelma and Louise. Or we might know where we want to go, but need a better car to get there... more Jed Clampett than Jason Bourne. Click on the following to see what we might hope to eventually achieve with the right vehicle and the right map.
A Screening Research Roadmap
This summary diagram illustrates how each stage of research on ASD screening practices can help to ensure continued progress towards important outcomes for people with ASD.
An integrated network
De-constructing the potential interplay between services, training, policy, and advocacy reveals the unique role of research in improving outcomes within an integrated network.
Stages of research
Each stage of research has different impacts on different elements of this network, with cascading effects on knowledge, practices, policies, and eventually population outcomes.
Piloting the roadmap on the CHAT+
We use the roadmap to systematically analyze more than 25 years of research and more than 90 original, peer-reviewed applied research studies into the CHAT+ family of screening measures.
Defining the specific steps within each stage, and the practices, desired outcomes for the entire roadmap, further break progress down, to potentially be addressed one study at a time.
When are the results of research ready for use in the clinic? For widespread dissemination? These watersheds define the designs, participants, and measures needed to progress towards impact.
Each stage requires different types of scholarship, partnerships, funding, and training. These differences reveal how traditional researchers must adapt to undertake implementation research.