Routine outcome monitoring in psychotherapy treatment using sentiment-topic modelling approach

Despite the importance of emphasizing the right psychotherapy treatment for an individual patient, assessing the outcome

of the therapy session is equally crucial. Evidence showed that continuous monitoring of patient's progress could significantly improve

the therapy outcomes to an expected change. The patient's progress can be tracked closely to help clinicians identify not progressing in

the treatment by monitoring the outcome. This monitoring can help the clinician consider any necessary actions for the patient's

treatment as early as possible, e.g., recommend different types of treatment or adjust the style of approach. Currently, the evaluation

system is based on the clinical-rated and self-report questionnaires that measure patients' progress pre-and post-treatment. While

outcome monitoring tends to improve therapy outcomes, there are many challenges in the current method, e.g., time and financial

burden for administering questionnaires, scoring, and analyzing the results. Therefore, a computational method for measuring and

monitoring patient progress throughout treatment is needed to enhance the likelihood of positive treatment outcomes. This paper

focuses on developing a computational method using a Dynamic Joint-Sentiment-Topic model (dJST) to measure and monitor the

patient treatment outcome by tracking patient’s current and recurrent views of topic and sentiment. We identified the sentiment and

topic trend evolved throughout treatments for each therapy session on the author's level. Our results show that this computational

method could potentially lead to an inexpensive clinical monitoring tool to evaluate patients' progress during psychotherapy.