On September 11, Olly Kravchenko successfully defended her doctoral thesis, Predicting health and socioeconomic outcomes in patients with common psychiatric disorders, at Karolinska Institutet. The thesis brings together Olly’s work on precision psychiatry, treatment outcome prediction, psychiatric genetics, and machine learning.
The overarching aim of the thesis was to investigate whether health and socioeconomic outcomes following internet-delivered cognitive behavioural therapy (ICBT) could be predicted using information available before and during treatment. Across four studies, Olly worked with a uniquely rich combination of clinical and sociodemographic information, nationwide Swedish register data, treatment-course data, and genetic data. The research focused primarily on patients receiving ICBT for depression, panic disorder, and social anxiety disorder, with one genetic study also including patients treated for OCD.
Overall, the thesis showed that treatment and socioeconomic outcomes could be predicted to a moderate and potentially useful degree. Information from national registers provided some additional predictive value beyond routinely collected clinical information, while polygenic scores consistently added little or no predictive value. Across the studies, some of the strongest signals came from patients’ accumulated history of psychiatric illness, functional difficulties, and socioeconomic circumstances. The work also highlighted that improvement in psychiatric symptoms and longer-term participation in working life are related but distinct aspects of recovery.
One of the thesis studies developed and temporally validated machine-learning models to predict clinically meaningful improvement after ICBT. These models achieved moderate predictive performance and provide a foundation for future prospective validation before such tools could potentially be used to support clinical decision-making.
Olly’s principal supervisor was Professor Christian Rück, with John Wallert and Ralf Kuja-Halkola as co-supervisors. Professor Joseph Hayes from University College London joined the defence as opponent. (Karolinska Institutet News)
To mark the occasion, we asked Olly a few questions about her time as a PhD student at Rücklab and what comes next.
What has been the most rewarding part of doing your PhD at Rücklab?
The opportunity to work with a truly unique dataset: real psychiatric patients treated in routine clinical care, followed longitudinally, with several different types of data. The linkage to Swedish registers allowed me to learn both about the enormous research opportunities they provide and the intricacies of working with complex, large-scale register data. Having access to genetic data also gave me the chance to deepen my knowledge of psychiatric genetics, which I find absolutely fascinating.
Which study, finding, or part of your PhD work are you most proud of, and why?
Probably the study in which we developed prediction models for treatment outcome. I encountered far more methodological and data-related challenges than I had anticipated, and at times it felt as though there was no end to the tasks that needed solving. Working through them taught me an enormous amount and gave me a certain confidence that even an intimidating research project can be completed with enough persistence.
What are you most looking forward to in the next step of your research career?
More challenges. I am looking forward to starting with a new research question, contemplating how to operationalise it and developing a study design that is best suited to answering it.
A huge congratulations to Olly on successfully completing her PhD — we look forward to seeing what comes next! 🎉
Read Olly’s thesis in KI Open Archive
DOI: 10.69622/32263065
















