Registered mental health therapists must have regular supervisory sessions with a clinical supervisor. The live supervisory sessions with clients can be uncomfortable for therapists, and there is a tendency to avoid them.
With mental health relapse rates after 18 months as high as 60% (Shehzad, et. al., 2017) and spontaneous recovery from most mental health problems in the 30 – 35% range (Whiteford, et al., 2013), the effectiveness of therapy is very questionable. Fifty-two plus thirty-five percent is 87% of therapy for clients is ineffective. Given that about 5% of serious mental health cases go on to different outcomes following an initial therapy session (regular in-patient visits, “falling through the cracks”, suicide, etc.), we are looking at a long-term success rate of about 8% for an average mental health therapist.
Although no specific studies are looking at the effectiveness of different mental health therapists in the success of treatment, it is stressed in training that there is a difference in outcomes based on the therapist’s competence. Research has shown that the relationship between therapist and patient contributes substantially to treatment effectiveness, independent of the specific type of treatment used (Cook, et. al., 2017: Kohrt, et. al., 2015).
AI offers an interesting new approach to this issue.
AI is currently capable of determining emotions from a one-minute clip at 75% (Antonov et. al., 2024). Longer exposure increases accuracy. In addition, the most recent AI models can highlight differences in peoples’ performances at levels as high as 99.4% (Sahoo et. al., 2023).
Why not use AI for live supervision? The therapist could be recorded by AI and the success of the therapist could be estimated by their 18-month relapse rate for clients. Across thousands of therapists, commonalities among the best therapists could be isolated and used to train poorer-performing therapists in techniques and treatment models.
Just a thought.
References
Antonov, A., Kumar, S.S., Wei, J. et al. Decoding viewer emotions in video ads. Sci Rep 14, 26382 (2024). https://doi.org/10.1038/s41598-024-76968-9
Cook, S. C., Schwartz, A. C., & Kaslow, N. J. (2017). Evidence-Based Psychotherapy: Advantages and Challenges. Neurotherapeutics : the journal of the American Society for Experimental NeuroTherapeutics, 14(3), 537–545. https://doi.org/10.1007/s13311-017-0549-4
Kohrt, B. A., Jordans, M. J., Rai, S., Shrestha, P., Luitel, N. P., Ramaiya, M. K., Singla, D. R., & Patel, V. (2015). Therapist competence in global mental health: Development of the ENhancing Assessment of Common Therapeutic factors (ENACT) rating scale. Behaviour research and therapy, 69, 11–21. https://doi.org/10.1016/j.brat.2015.03.009
Sahoo, Goutam Kumar, Das, Santos Kumar, and Singh, Poonam, (2023). Performance Comparison of Facial Emotion Recognition: A Transfer Learning-Based Driver Assistance Framework for In-Vehicle Applications, Circuits Syst. Signal Process, 42-7. Pages 4292-4319. https://doi.org/10.1007/s00034-023-02320-7
Shehzad Ali, Laura Rhodes, Omar Moreea, Dean McMillan, Simon Gilbody, Chris Leach, Mike Lucock, Wolfgang Lutz, Jaime Delgadillo (2017). How durable is the effect of low intensity CBT for depression and anxiety? Remission and relapse in a longitudinal cohort study, Behaviour Research and Therapy, 94: p. 1-8, https://doi.org/10.1016/j.brat.2017.04.006.
Whiteford, H. A., et al. (2013). Estimating remission from untreated major depression: A systematic review and meta-analysis. Psychological Medicine 43.8 (2013): 1569-85.
Comments
3 responses to “Clinical Supervision”
I love this idea. To give therapists an advantage would be wonderful. I have been to many and been misdiagnosed several times. This has the potential of being life saving for some people.
To further this point, a technology called Facial Emotion Recognition (FER) is a field of computer vision that focuses on detecting and analyzing human emotions from facial expressions in images and video content (Zi-Yu Huang et el: 2023).
FER systems often use Convolutional Neural Networks (CNNs) or other machine learning algorithms to classify facial expressions into basic emotions such as happiness, sadness, anger, surprise, fear, and disgust (Morphcast, 2024).
this could give Therapists tools to determine if the patient is representing their real emotions in therapy. That could lead to more effective sessions and better solutions.
Zi-Yu Huang, Chia-Chin Chiang, Jian-Hao Chen, Yi-Chian Chen, Hsin-Lung Chung, Yu-Ping Cai & Hsiu-Chuan Hsu (2023) : https://www.nature.com/articles/s41598-023-35446-4
Morphcast, (2024); https://www.morphcast.com/facial-emotion-recognition-faq/
Where could this end though? What if they start replacing therapists with Ai in general? Certainly more cost affective and available! It’s already happening. There’s sites where you can discuss your mental health and get suggestions:
https://www.clareandme.com/ai-for-mentalhealth-worries-and-overthinking#:~:text=proven%20behavioral%20health-,You%20can%20call%20or%20text%20Clare%20anytime.,exercises%20based%20on%20behavioral%20health.&text=Our%20AI%20will%20determine%20which,in%20order%20to%20support%20you.