Objective

Few data are available to inform clinical expectations about course and prognosis of severe obsessive-compulsive disorder (OCD). Such information is necessary to guide clinicians and to inform criteria for invasive interventions for severe and intractable OCD. This study sought to examine course and prospective predictors of a chronic course in patients with severe OCD over 5 years.

Method

A selected subset of adults in the Brown Longitudinal Obsessive-Compulsive Study (BLOCS) was included. Adult BLOCS participants were enrolled between 2001 and 2006. All participants in the current study (N = 113) had DSM-IV OCD diagnosis, severe OCD symptoms at baseline, and at least 1 year of follow-up data.

Results

Cox proportional hazard models were used to examine the general pattern of course in the severe OCD sample based on Longitudinal Interval Follow-Up Evaluation (LIFE) psychiatric status ratings, as well as test predictors of chronically severe course. Results indicated that approximately half of patients with severe OCD at baseline had illness drop to a moderate or lower range of severity during 5 years of follow-up (50.4%) and that marked improvement was rare after 3 years of severe illness. The only unique predictor of a more chronically severe course was patient report of ever having been housebound for a week or more due to OCD symptoms (P < .05).

Conclusions

Findings of this study were 3-fold: (1) half of participants with severe OCD have symptom improvement over 5 years of follow-up, (2) the majority of participants that drop out of the severe range of symptom severity do so within the first 3 years of follow-up, and (3) patient-reported history of being housebound for 1 week or more due to OCD is a significant predictor of OCD's remaining severe over the 5-year follow-up.

The study objective was to apply machine learning methodologies to identify predictors of remission in a longitudinal sample of 296 adults with a primary diagnosis of obsessive compulsive disorder (OCD). Random Forests is an ensemble machine learning algorithm that has been successfully applied to large-scale data analysis across vast biomedical disciplines, though rarely in psychiatric research or for application to longitudinal data. When provided with 795 raw and composite scores primarily from baseline measures, Random Forest regression prediction explained 50.8% (5000-run average, 95% bootstrap confidence interval [CI]: 50.3-51.3%) of the variance in proportion of time spent remitted. Machine performance improved when only the most predictive 24 items were used in a reduced analysis. Consistently high-ranked predictors of longitudinal remission included Yale-Brown Obsessive Compulsive Scale (Y-BOCS) items, NEO items and subscale scores, Y-BOCS symptom checklist cleaning/washing compulsion score, and several self-report items from social adjustment scales. Random Forest classification was able to distinguish participants according to binary remission outcomes with an error rate of 24.6% (95% bootstrap CI: 22.9-26.2%). Our results suggest that clinically-useful prediction of remission may not require an extensive battery of measures. Rather, a small set of assessment items may efficiently distinguish high- and lower-risk patients and inform clinical decision-making.

There is a critical need for increasing the diversity representation among clinicians and researchers in academic medicine, including departments of psychiatry and psychology. Mentorship of under-represented groups has been identified as an important way to remediate diversity-related barriers in the field. This paper outlines the early development and pilot implementation of a diversity mentorship program at Brown University. In an effort to inform and guide future diversity programs, we discuss the challenges faced in creating the program, the successes experienced during the first year, and the future directions undertaken as a means for improving the program.