A Better Way to Identify, Measure, and Treat PTSD

For a significant number of veterans, the symptoms of PTSD are an organizing feature of their lives. It is the feeling of hypervigilance, avoiding overstimulation or triggers, and the dread of going to sleep for fear of nightmares. These examples cannot capture the experience of living with PTSD, but they might provide a small insight into what 23% of  Operations Enduring Freedom and Iraqi Freedom veterans experience daily. To put it bluntly, veterans are no strangers to the symptoms of PTSD.

Still, the process of being diagnosed can be anxiety provoking. Standard practice has historically required clients to travel to a busy clinic where they will meet with a clinician to go through a structured assessment for an hour or more, and each of these tasks—from travel, entering a clinic, to sitting with a clinician for an hour—is a hurdle that sits between a veteran seeking help and receiving treatment. The work that my collaborators and I recently published in JAMA Network Open seeks to radically transform this process.

Making Assessment Accessible

In collaboration with the Veterans Health Administration, we built a set of tools based on computerized adaptive testing and diagnosis (CAT and CAD) and multidimensional item response theory (MIRT) and machine learning (extremely randomized trees) that strip away each of the barriers discussed above. With the CAD- and CAT- PTSD, clients have the ability to complete the assessments in spaces where they feel comfortable and without fear of judgement from an interviewer—the only requirement is that they have access to a phone, tablet, or computer with internet access. Importantly, these tools faithfully reproduce lengthy structured clinical interview diagnosis taking an hour or more, but have low client burden, requiring a fraction of the time (less than 2 minutes) to complete. 

The CAD-PTSD is an adaptive diagnostic screener that requires a maximum of 6 questions and can be administered in an average of 35 seconds. The CAT-PTSD provides a continuous PTSD symptom severity score with an average of 10 questions in 59 seconds. In practice, we can begin with the CAD and if positive measure symptom severity using the CAT. As such, most people will complete the assessment in less than a minute. 

A key benefit of these tools is that they allow clinicians to change how they spend their time with clients. The CAD-PTSD provides clinicians with a probability that denotes the likelihood of a positive diagnosis of PTSD, and the CAT-PTSD provides a severity score and a category (such as none, mild, moderate, severe), that forms the basis of treatment selection and triage. When assessment happens outside a clinic or office, clinicians can jumpstart a session with this knowledge, enhancing their ability to build a meaningful client relationship, supporting clients in developing coping skills, and engaging in interventions. Moreover, the CAT-PTSD is optimized for regular, longitudinal assessment. Clinicians can use weekly or bi-weekly assessments to guide treatment plans, track severity over time, and collaborate with clients about their own sense of progress and goals. Taken together, these changes present an opportunity to practice in a way that blends the benefits of both data-driven practice and clinical expertise.

Assessment in Policy and Research

The CAD- and CAT- PTSD are powerful on the individual-level, but their power is no less notable in the realms of policy and research. PTSD, and mental health generally, is difficult to measure in all its complexity. Prior endeavors to evaluate mental health have often required a tradeoff between long, labor-intensive structured clinical interviews and less-informative methods that do not come close to the accuracy and precision found with these tools. The combination of machine learning for diagnosis and computerized adaptive testing based on multidimensional item response theory for severity measurement, provides a tangible solution—the accuracy of lengthy structured clinical interviews, and increased precision of severity measurement using half the items as provided by extant PTSD scales such as the PCL-5.  

The CAD-PTSD and CAT-PTSD make reaching new horizons in policy, research and implementation possible. The scalability of these tools creates the possibility of regular population-level screening for PTSD among veterans and active-duty troops. A policy of routine screening will aid in early detection of PTSD or its precursors, allowing mental and behavioral health care for veterans to be proactive as opposed to reactive. The CAT-PTSD is ideal for use in research that seeks to measure the efficacy of therapeutic interventions. Because it is scalable, affordable, and built with longitudinal measurement in mind, the CAT-PTSD makes it possible to look beyond whether a client is simply less symptomatic at the end of an intervention to whether the effects of the intervention are statistically significant, even at the level of the individual participant. Researchers will be able to ask much more targeted questions such as:

• Are there periods of time where the treatment, on average, increases symptom severity for clients?
• Are there differences in how long it takes for a client to achieve benefits of treatment? 
• Is the current treatment working or should it be augmented or switched?

These are questions that allow clinicians to tailor their interventions to specific clients, and the CAT-PTSD has the capacity to make this information much more widely available.

These tools represent a paradigm shift in psychiatric measurement. Through adaptive testing, they increase the precision of measurement and decrease the burden of measurement. Optimal items are selected in real time for each person, targeted to their level of severity, which is learned from their responses to the adaptively selected questions. Unlike traditional mental health measurement which fixes the items and allows the measurement uncertainty to vary, CAD and CAT fix the precision of measurement and allow the items to vary. Measurement-based screening and measurement-based care are the steppingstones to effective treatment and the CAD/CAT -PTSD make this a reality. The development of these tools through collaboration with the Veterans Health Administration paves the way for more effective interventions, more supportive systems, and better outcomes for those who have served our country.

Disclaimer: The views and opinions expressed are those of the author(s) and do not necessarily reflect the official policy or position of the Population Health Learning Network or HMP Global, their employees, and affiliates. Any content provided by our bloggers or authors are of their opinion and are not intended to malign any religion, ethnic group, club, association, organization, company, individual, or anyone or anything.