Plain English Summary

Background: 

Many decisions in healthcare are made by considering the costs and health benefits of different options. An important issue for decision-making bodies, like the National Institute for Health and Care Excellence (NICE) and the Department of Health and Social Care, is to assess the health benefits of different devices, medicines, vaccines and procedures in a consistent way, even though they affect different types of patients and health conditions. Questionnaires asking about broad health aspects, such as EQ-5D, are commonly used as the basis of measures of health benefit. In the case of EQ-5D, the questionnaire asks patients about their health in five areas: mobility, ability to self-care, usual activities, pain and anxiety/depression. 

EQ-5D works well in many situations. However, there is a concern that this type of measure, chosen for its simplicity and applicability to a wide range of disease areas, may miss some important health benefits that more detailed disease specific measures can capture. Decision-makers have limited tools available to assess whether these claims are true and, if they are, no way to judge how incorrect estimates of health benefits might be. This is obviously unsatisfactory because decisions are based on these assessments and overall patient care may suffer as a result. Therefore, there is a need for improved methods to identify when EQ-5D may not be fully appropriate and to consider alternative approaches that can be taken when shortcomings are identified.

We have developed methods to address these issues and tested them out in a case study of breast cancer. Another area where it is unclear whether EQ-5D captures health benefits accurately is in mental health. There are lots of new digital health applications being developed to help people with complex mental health conditions, so it is important decision makers can assess health benefits in relation to these diseases.

Aims and objectives:

• To apply previously developed methods to assess if EQ-5D suitably covers aspects of quality of life that are relevant to people with psychosis.

• To calculate how biased results might be from using EQ-5D if it does not fully cover the aspects of quality of life that matter to patients. 

Methods:

We will use data from 406 people that took part in a clinical trial that aimed to improve the health of people with established psychosis. Psychosis is one type of mental health condition where there are digital apps in development that may require evaluation by organisations like NICE. 

We will also use data from another trial if we are able to negotiate access.  

The statistical methods we will use, that were developed and demonstrated by the same researchers in data from people with breast cancer, compares EQ-5D to other measures of disease severity, and considers how much they move in the same direction. 

Policy relevance and dissemination:

This work is of direct relevance to NICE as it will help them in future assessments of health interventions for people with mental health conditions. 

We will consult with NICE and the EEPRU Public Involvement panel to help produce descriptions of our work and the findings in the clearest ways.