In this blog, EDIT Lab Postdoc Meg Skelton discusses our recent paper looking at patient trajectories of symptom and impairment during internet-based psychological therapy for anxiety and depression. 


Meg Skelton, EDIT Lab Postdoctoral Researcher

Psychological therapies can be effective in treating anxiety or depression. However, only about half of patients recover by the end of treatment (1,2). There are lots of things that are associated with whether a patient recovers or not, including characteristics of the patient (see our blog on therapygenetics) and of the therapy itself (1). Yet, it isn’t only that people differ at the end of treatment. People also differ in how they respond throughout treatment, showing different trajectories of response. Identifying potential subgroups of patients with similar trajectories and understanding the characteristics of patients within those subgroups is important for patient-centred treatment. For example, information on a patient’s expected trajectory could be used to monitor whether they are ‘on-track’ (improving as expected after a certain number of sessions), and also help to guide recommendations on treatment suitability. This would be especially important for any subgroup of patients who take longer than average to show clinically significant improvement, to avoid patients or clinicians making unnecessary treatment adjustments.

We previously looked at symptom trajectories in around 16,000 patients at South London NHS Talking Therapies (previously ‘IAPT’) services (3). We found that the different trajectories could be captured as four trajectory subgroups (‘classes’; see figure below). Three classes had initially moderate-severe symptoms and demonstrated 1) no change, 2) gradual improvement, and 3) fast improvement. A final class 4) showed initially mild symptoms and minimal improvement. What we found most interesting was that although the first three classes all start with similar symptom scores, they show very different patterns of response as treatment progresses. Within those classes, we found that patients in the two showing improvement, compared with the one that didn’t, tended not to have reported prescribed psychotropic medication or a disability (e.g., physical, learning) and to be in employment.

In our new paper, we analysed data from over 50,000 patients who had received internet-enabled cognitive behavioural therapy (CBT) via a confidential chat-based platform, to see if they showed the same trajectories as patients in more traditional face-to-face therapy (4). We were also able to look at a third outcome, alongside anxiety and depression symptoms: ‘functional impairment’, which is the extent to which symptoms impact the ability to perform everyday tasks like work and socialising. The trajectories for all three outcomes were very similar to each other, and to our previous study of face-to-face treatment! Across the outcomes, patients with moderate-severe baseline symptoms/impairment who showed improvement as opposed to no change were more likely to be employed and not to have obsessive-compulsive disorder (OCD). We were also able to identify variables associated with different rates of change: fast improvement was likelier than gradual improvement or no change for patients with older age, no disability, or lower comorbid symptom or impairment scores. The similarity in results indicates that findings from face-to-face therapy may be generalisable to remotely delivered therapies, which makes sense in light of evidence that they are similarly effective (see our blog on internet-delivered therapy here).

Using patient information at the start of treatment to identify their most likely outcome trajectory could inform patient and clinician expectations and decisions, and be used to monitor patient progress throughout treatment.


    1. Cohen ZD, DeRubeis RJ. Treatment Selection in Depression. Annu Rev Clin Psychol 2018; 14: 209–236.
    2. IAPT Team, NHS Digital. Psychological Therapies, Annual report on the use of IAPT services, 2020-21. NHS Digital, (2021, accessed 15 March 2022).
    3. Skelton M, Carr E, Buckman JEJ, et al. Trajectories of depression and anxiety symptom severity during psychological therapy for common mental health problems. Psychol Med 2022; 1–11.
    4. Skelton M, Catarino A, Brown S, et al. Trajectories of depression symptoms, anxiety symptoms and functional impairment during internet-enabled cognitive-behavioural therapy. Behav Res Ther 2023; 169: 104386.
Megan Skelton

Author Megan Skelton

More posts by Megan Skelton