Developing a Dynamic Progress Feedback System v1
Research type
Research Study
Full title
Development of a dynamic progress feedback system to guide psychological treatment in primary care.
IRAS ID
233799
Contact name
Jaime Delgadillo
Contact email
Sponsor organisation
Cumbria Partnership NHS Foundation Trust
Duration of Study in the UK
1 years, 8 months, 3 days
Research summary
In England, psychological therapy is often accessed via the National Health Service’s (NHS) Improving Access to Psychological Therapies (IAPT) services. However, between 2015-2016, only 46.3% of people attending IAPT saw reliable and clinically significant improvement (RCSI), where their symptom severity scores showed statistically significant improvement and moved to below clinical cut-off levels by the end of treatment (Community and Mental Health Team for the NHS, 2016).
Understanding who is at risk of poor outcomes is therefore important so that clinicians can act quickly to ameliorate this risk, and there are currently two approaches to this. ‘Patient profiling models’ such as the Leeds Risk Index (LRI, Delgadillo et al., 2016) make a prediction of risk at the outset of therapy, based on variables such as symptom severity, age, disability, employment status, expectations and functional impairment. ‘Outcome feedback models’ use session-by-session data to compare an individual’s scores with normative data and provide risk spikes where the person’s scores fall outside ‘on track’ boundaries.
These approaches reply on static snapshots of a person’s scores, however, which do not fully accommodate an individual’s pattern of recovery. The current study therefore proposes to combine the initial profiling capability of the LRI with session-by-session progress data to develop a state-of-the-art dynamic progress feedback system. The model will calculate an initial prognosis based on the LRI. However it will also ‘learn’ from incoming weekly progress change scores as determined by the GAD-7 for anxiety and PHQ-9 for depression. This means it adjusts the person’s prognosis on a weekly basis, providing the clinician with an updated percentage risk of not achieving RCSI. The model will be built using archival data from 4000 patients who accessed psychological therapy in IAPT Leeds. It will then be tested on data from 1000 patients who accessed IAPT Cumbria, to assess its generalisability.
REC name
West Midlands - Coventry & Warwickshire Research Ethics Committee
REC reference
18/WM/0012
Date of REC Opinion
9 Jan 2018
REC opinion
Favourable Opinion