ABSTRACT
Accurate transcription of audio recordings in psychotherapy would improve therapy effectiveness, clinician training, and safety monitoring. Although automatic speech recognition software is commercially available, its accuracy in mental health settings has not been well described. It is unclear which metrics and thresholds are appropriate for different clinical use cases, which may range from population descriptions to individual safety monitoring. Here we show that automatic speech recognition is feasible in psychotherapy, but further improvements in accuracy are needed before widespread use. Our HIPAA-compliant automatic speech recognition system demonstrated a transcription word error rate of 25%. For depression-related utterances, sensitivity was 80% and positive predictive value was 83%. For clinician-identified harm-related sentences, the word error rate was 34%. These results suggest that automatic speech recognition may support understanding of language patterns and subgroup variation in existing treatments but may not be ready for individual-level safety surveillance.
ABSTRACT
OBJECTIVE: To examine maintenance of recovery following treatment in an adult anorexia nervosa (AN) population. METHOD: One year follow-up of a randomized clinical trial with 122 participants treated with: cognitive behavioral therapy (CBT), drug therapy (fluoxetine), or a combination (CBT+fluoxetine) for 12 months. Participants were assessed at baseline, end of treatment, and follow-up. The primary outcomes were weight and the global scores from the Eating Disorder Examination (EDE) separately and combined. RESULTS: Fifty-two participants completed the follow-up. Mean weight increased from end of treatment to follow-up. Seventy-five percent (75%) of those weight recovered at end of treatment maintained this recovery at follow-up. Recovery of eating disorder psychopathology was stable from end of treatment to follow-up, with 40% of participants with a global EDE score within normal range. Using the most stringent criteria for recovery, only 21% of the completer sample was recovered. DISCUSSION: The findings suggest that while adults with AN improve with treatment and maintain these improvements during follow-up, the majority is not recovered. Additionally, further research is needed to understand barriers to treatment and assessment completion.
Subject(s)
Anorexia Nervosa/therapy , Cognitive Behavioral Therapy , Fluoxetine/therapeutic use , Selective Serotonin Reuptake Inhibitors/therapeutic use , Adult , Anorexia Nervosa/drug therapy , Anorexia Nervosa/psychology , Body Weight , Combined Modality Therapy , Female , Follow-Up Studies , Humans , Personality , Psychiatric Status Rating Scales , Self Concept , Surveys and Questionnaires , Treatment OutcomeABSTRACT
OBJECTIVE: The diagnostic criteria for anorexia nervosa (AN), bulimia nervosa (BN), and binge eating disorder (BED) establish symptom severity levels, which are used to separate full cases from partial cases. However, the value of these distinctions is unclear. METHOD: Three hundred eighty-five women with full or partial AN, BN, or BED were assessed at entry into a longitudinal study of eating disorders. RESULTS: Stepwise discriminant analysis revealed that full and partial BN were discriminated by the Yale-Brown-Cornell Eating Disorders Scale total scores (kappa =.46). However, it was not possible to discriminate between full and partial AN or BED. Discriminant analysis also demonstrated clear differences between full AN, BN, and BED. DISCUSSION: Full BN can be differentiated from partial BN by more severe eating disorder symptoms, whereas both full and partial AN and full and partial BED appear quite similar. These results emphasize the distinct nature of AN, BN, and BED, as well as the similarities between full and partial cases.
Subject(s)
Anorexia Nervosa/diagnosis , Bulimia/diagnosis , Surveys and Questionnaires , Adolescent , Adult , Anorexia Nervosa/epidemiology , Bulimia/epidemiology , Discriminant Analysis , Female , Follow-Up Studies , Humans , Middle Aged , Prevalence , Reproducibility of Results , Severity of Illness Index , SyndromeABSTRACT
OBJECTIVE: The study was designed with the aim of determining whether extending group cognitive-behavioral therapy (CBT) would enhance outcome among individuals with binge eating disorder (BED) who failed to stop binge eating after an initial 12-week CBT intervention. METHOD: Forty-six participants who met diagnostic criteria for BED were randomly allocated to either a 12-week group CBT intervention or a waiting list control condition. At the end of 12 weeks, treated participants who met clinical criteria for improvement subsequently received 12 sessions of behavioral weight loss. Remaining participants received 12 additional sessions of CBT for binge eating. RESULTS: Fifty percent of treated participants improved with the initial 12-week course of CBT. There was a strong trend for the extension of CBT to affect improvement in binge eating among initial nonresponders (6 of 14 subjects no longer met diagnostic criteria for BED). Overall, extending CBT led to clinical improvement in 66.7% of all treated participants, with treatment gains occurring through session 20. DISCUSSION: The results suggest that an extended course of CBT (i.e., longer than 12 weeks) will likely maximize the number of potential responders to treatment.