Skip to main content

Predicting Response to Depression Treatment

Periodic Reporting for period 4 - PReDicT (Predicting Response to Depression Treatment)

Periodo di rendicontazione: 2018-10-01 al 2019-11-30

Depression is the leading cause of disability in the world today. With its high prevalence and treatment costs, its role as a risk factor for suicide and its impact on workplace productivity, depression places an enormous burden on the healthcare system and the wider economy. The scale of the problem is substantial; over 350 million people (of which 11% are EU citizens) suffer from depression, it takes 4-6 weeks after starting antidepressant therapy before a physician can detect whether the treatment is working, and more than 50% of patients fail to respond to the first antidepressant prescribed, meaning it can take several months before an effective treatment is identified. During this time a patients’ ability to work and function socially is impaired.

Our solution to this problem, the P1vital® PReDicT Test (PReDicT), is a web based digital tool able to measure subtle changes in the way patients interpret emotional information in the form of facial expressions. In combination with a depression questionnaire and a proprietary machine learning algorithm, PReDicT provides a rapid and objective measure of whether a patient is responding to their treatment or not within 1 week of starting treatment and, therefore, has the potential to exert multiple benefits.

The objective of the project was to conduct clinical and commercial research to validate the performance of PReDicT; ultimately to support the launch of our novel medical device to improve the treatment and management of depression in primary care.
The project spanned 4 years during which we focussed on (i) development of the medical device, (ii) conduct of the randomised controlled clinical trial, (iii) stakeholder engagement and dissemination, and (iv) commercialisation. Project Management was a constant to ensure the project was completed to scope, on time and within budget.

PReDicT was developed and registered as a CE marked Class 1 medical device, and deployed successfully in the clinical trial.

Ethics and regulatory approvals were secured to enable the clinical trial to be ran in 5 European countries: UK, France, Germany, Spain and The Netherlands. ~1000 patients were recruited over 2 years; half from the UK and the remainder across the mainland European countries.

Patients received a baseline assessment prior to starting their antidepressant medication, and completed PReDicT again 1 week later. Within the PReDicT arm, the results were reviewed by the GP who determined subsequent treatment. If a change in medication was recommended (either an increase in dose or a switch to a new antidepressant), the patient completed the PReDicT Test again the following week and so on.

The study was divided into an 8 week clinical phase and a 40 week follow-up phase. After week-8, clinical efficacy and acceptability were measured and compared between the PReDicT group and the control Treatment-as-Usual (TaU) group. During the follow up phase, clinical efficacy and health economics data were taken from each arm.

Compliance rates at all phases were good; over 95% during the clinical phase, falling to 59% & 51%, respectively, at the two follow up time-points (week-24 and week-48).

The main results of the trial are summarised as follows:
1. The algorithm was 56% accurate, with 70% of patients predicted to respond vs. 30% who were not.
2. PReDicT influenced GP prescribing behaviours; a significant proportion changed medications if their patients had received a negative response prediction.
3. PReDicT resulted in an increased proportion of depressed patients showing a response to treatment at week-8 compared to TaU, however this was not statistically significant.
4. A greater reduction in anxiety symptoms at week-8 was observed in the PReDicT group;
5. A functional outcomes improvement at week-24 was shown, suggesting PReDicT may bring forwards functional recovery. A similar improvement was also observed in capability wellbeing outcomes.
6. Positive feedback for PReDicT was received by both clinicians and patients;
7. Cost data were mixed with a borderline medication cost decrease observed with PReDicT;
8. European cost-effectiveness QALY thresholds supported an economic benefit from a full societal perspective.

We engaged with several stakeholders including patients, physicians, potential payers, and policy makers to optimise healthcare adoption. We also used multiple communication media including website and newsletters, alongside dissemination activities including attendance at multiple conferences, trade fairs and pitch events to publicise the project, our novel device, and the results, from which the level of interest was high.

We conducted market research in the UK, Germany and US, and compiled a Commercialisation Plan outlining our intended go-to-market strategy, which includes generating real world evaluation data of which a UK pilot is ongoing.
The expected positive benefits were realised, in the main.

• Significant reduction in time spent on ineffective medication and associated reduced risk of experiencing unnecessary side effects by identifying non-response early in treatment.
• Improvement in patient outcomes, notably by reducing anxiety symptoms and bringing forwards functional recovery and/or capability wellbeing, thereby shortening functional disability and positively impacting patients and their families in their social and working lives.
• Significant decrease in antidepressant non-adherence by providing patients with an early indication that the antidepressant they are taking is working which will encourage them to continue taking their medication even if they are suffering from unpleasant side effects.

• Facilitated decision-making; clinicians were able to follow the nationally recommended treatment pathway, but were able to evaluate response after 7 days rather than the typical 4-6 weeks, and use these data to support their clinical decision making.
• Reduction in standard face-to-face GP appointment time; several patient interactions were held using email or telephone consultations, which aided GP resilience.

Payers / Economy
• Significant reductions in healthcare costs did not materialise, unfortunately.
• Reductions in societal costs were part-realised, with the cost-effectiveness threshold data suggestive of favourable consideration by mainland European payers (marginally less so by the UK), and confirming that the main economic benefits of PReDicT fall on broader societal costs.
Schematic of the P1vital® PReDicT Test
Summary of the PReDicT value proposition