McLean Hospital 115 Mill Street Belmont, MA 02478
The beginnings of 2016 have been exciting times for the OCDI Office of Clinical Assessment and Research, as we expand the scope, scale, and impact of our work. Recent weeks have seen the lab make a big dissemination push at McLean Research Day and at the upcoming ADAA conference, along with efforts to statistically validate and publish internally-developed symptom and process-related measures. We have also started data collection on our study of psychophysiological indicators of emotion processing and initialized collaborations with scholars both within and beyond the McLean community in an effort to contribute to developing a better understanding of the genetic underpinnings of OCD and related disorders.
All the while, we have continued to conduct weekly patient assessments in an effort to first and foremost understand our patients’ experience and evaluate treatment outcomes. Because of this regular data collection, we are able to study a patient’s trajectory through treatment. Although we already know that, on average, patients do see significant symptom reduction during their time with us (on the order of a 40% drop in symptoms) our anecdotal observations cause us to suspect that there is wide variability in terms of the trajectory (or “path,” “journey”) to symptom change that patients experience along the way. Because we track patients’ symptom changes in a fine-grained, week by week manner, we are well-positioned to examine the nature of these trajectories.
So, a question that we seek to answer is: do all patients tend to follow the same general trajectory through treatment? Or, do different groups of patients travel along different trajectories? Further, what is the shape of these trajectories? What might predict whether a given patient takes one trajectory versus another?
Formal statistical analytical techniques exist that allow us to answer these questions. An example is called growth mixture modeling. In a nutshell, this analysis looks at all of our weekly symptom data (measured using the YBOCS scale) and tells us how many unique symptom-change trajectories exist in the data, the shape of those trajectories (i.e., variables like slope and intercept), and which trajectory each patient in the analysis has traveled.
The results of this analysis were interesting. We found that there were three distinct trajectories of symptom changes in the sample we studies (220 patients). As you can see in the figure, the first group showed blunted change—their “path” was quite flat—and they showed the least amount of overall change. The second group was a bit better, showing a pronounced initial decline in symptoms followed by a significant slowing of that change during the back half of treatment. The third group showed the most robust changes, with the steepest initial decline followed by a slowing but still significant change.
What sorts of individual differences might predict which trajectory a patient traveled during his/her treatment? We are currently analyzing a number of such factors, including degree of severity in specific symptom presentation (i.e., contamination, checking, not-just-right experiences, intrusive thoughts), variations in response inhibition, and patients’ willingness to approach difficult experiences during their exposure work. Early results indicate that each of these factors may significantly shape a patient’s journey.
These data represent a step towards identifying not only the rate and amount of improvement in our patients, but also those patients who are struggling the most in treatment. Ultimately, we hope that our investigations will allow us to develop more refined and targeted interventions to better help individuals grappling with this debilitating disorder—both here at the OCDI and beyond.