Call for evidence outcome

Equity in medical devices: independent review - response to call for evidence

Updated 11 March 2024

Overview

The Equity in medical devices: independent review ran an 8-week public consultation from 11 August to 6 October 2022 in the form of a call for evidence through an online survey hosted on GOV.UK. It also accepted submissions by email.

Prior to its launch, the call for evidence was reviewed by stakeholders - including representatives of patient charities, industry representatives and the Medicines and Healthcare products Regulatory Agency (MHRA) - and revised based on their feedback. Among other changes, questions were introduced on views and insights for those who may not have access to evidence.

Scope of the call for evidence

The call for evidence invited information concerning all medical devices across the entire device life cycle, as well as ethnic and other unfair biases. It asked key open-ended questions on bias in medical devices, and respondents could upload attachments to support their responses. It asked not just for evidence, but also insights and views, to allow a broad range of individuals to contribute to the consultation.

The panel welcomed any data and evidence, including preliminary evidence and information on ongoing data collection and research or evaluation projects, regarding existing equity concerns or biases with respect to ethnicity or other socio-demographic characteristics, and any mitigating solutions. This could include:

• data and evidence related to ethnic and other unfair biases in relation to the development and use of medical devices
• examples of effective, successful or evidence-based approaches to mitigate these types of unfair biases and risks
• ideas for potential approaches to improving equity in the development and use of medical devices

Audience

The call for evidence was open to the public and shared with the media through the Department of Health and Social Care (DHSC) press office as well as directly through DHSC communication channels to relevant stakeholder groups.

Respondents could contribute:

• as individuals sharing personal views and experiences
• as individuals sharing professional views
• on behalf of organisations

The panel welcomed contributions from all interested parties, but expected these topics to be of particular interest and relevance to:

• organisations representing diverse communities and patient groups, including charities
• community leaders
• academia, research, think tank and funding institutions
• clinicians and other healthcare professionals
• NHS and health improvement organisations
• industry, including medical device developers, manufacturers and innovators
• device regulators and standard-setters, including legal teams

Patients and members of the public were invited to share their views and insights.

If patients had concerns about the healthcare received, they were provided with information about the relevant channels and support available through NHS England. Additionally, they were encouraged to report suspected side effects from medicines, vaccines and e-cigarettes, medical device incidents, and defective or falsified (fake) products through the MHRA’s Yellow Card reporting site.

Results

The consultation received 44 completed submissions through the online survey, and a further 7 submissions by email.

Contributors

Of the 44 online submissions, 9% of respondents were individuals sharing personal views and experiences, 43% were individuals sharing professional views, and 48% were submissions on behalf of an organisation.

The responses came from a wide range of organisations and sectors, including the device industry, healthcare providers, research, academia, standards and practice, and a variety of medical specialties.

Submissions

At the start of the submission, contributors could mark which categories their submissions fall under (multiple answers were possible). Below is a list of how many submissions were self-classified as related to a specific medical device type (noting that some submissions referred to more than one type of device):

• medical devices not enabled by artificial intelligence (AI) - 29

• AI-enabled devices - 22
• polygenic risk scores - 5
• none or I don’t know - 2
• all other responses - 6

Below is the number of submissions classified by the respondents as falling into different categories of bias (noting that some submissions referred to more than one characteristic):

• ethnicity - 32
• gender and sex - 8
• age - 5
• disability - 4
• socio-economic background or deprivation - 10
• none or I don’t know - 6
• all other responses - 4

There were few submissions from individual members of the public, and most of these were outside the scope of this review, covering issues surrounding access to and quality of medical care and devices - such as access to dental treatment and bias of healthcare professionals giving preferential treatment to white patients.

Evidence included in submissions

Of the online submissions, 17 said they had evidence relating to bias in design and development of medical devices, and 17 had evidence relating to bias in the use of medical devices. There was some degree of repetition between these answers and related attachments.

A total of 18 respondents said they knew of evidence-based solutions in the design and development of medical devices, and 14 of evidence-based solutions in the use of medical devices.

Most evidence provided drew on published research and academic papers.

Key themes

Key themes emerged, which are listed below. The findings for each were synthesised to inform the independent review panel’s thinking as they developed their problem definition tables and recommendations, which were published in the Equity in medical devices: independent review - final report. Multiple themes may have appeared in a single submission, and several submissions spoke to the same themes.

Theme 1: issues of bias in the design and development of devices

This included:

• unrepresentative or excluded populations in clinical trials, research and biobanks​
• poorly worded regulations and standards, which leave room for inadequate or insufficient design and development practices​
• economic and practical challenges for companies in designing and developing equitable medical devices (with a focus on the majority of users who will sustain business models)​
• challenges in developing single all-encompassing AI models (arguments for targeted solutions that may not meet equity thresholds)​
• poor standardisation and monitoring of devices

Theme 2: issues of bias in the use of devices

This included:

• postcode lottery in access to medical devices (and care in general) - complex routes to inequality​
• many socio-demographic and economic characteristics contributing to inequities in access to and use of medical devices (including mental health, disabilities, language barriers or cultural factors)​
• lack of instructions and/or calibration to different physical or biological parameters​
• limited or poor quality of data collection within devices or while using them - challenges for monitoring and evaluation​
• poor standards or regulations for risk control, such as warning labels​
• off-label use or use of medical devices offered for free by NHS undergoing less scrutiny​
• the immaturity of the market and limited evidence or data available​
• debate over effectiveness or equity leads to uncertainty and poorer quality

Theme 3: solutions in the design and development of devices

This included:

• representation and diversity in data sets, user-groups, clinical trials and so on, including demographic characteristics and lived experience of condition​
• involvement of end users at development stage​
• clarity in guidelines or regulation​
• incentivising collection of demographic data by manufacturers​
• the importance of clean data (distinguishing between signal and noise in data)​
• accessibility features such as braille, audio description and language options
• the need for manufacturers to conduct research to identify under-served groups and bias or disparities

Theme 4: solutions in the use of devices

This included:

• using tools such as equality health impact assessments
• training or guidance on using devices among different groups - healthcare and patients​
• considering ethnicity, children, the elderly, intersex and transgender people​
• post-deployment monitoring or data ​collection
• recommendations of specific devices and tools​
• incentivising collection of demographic data by manufacturers​
• publication of evidence on how devices work among different groups​
• clear guidance or a ‘companion guide’ to reduce educational, literacy or digital access barriers to use​

How insights from the call for evidence were used

Responses were collated for the panel’s review. Additionally, the secretariat highlighted key themes from the written submissions and presented these to the panel to inform their thinking as they defined the key equity issues and developed draft recommendations.

Attached and referenced papers submitted were added to the landscape mapping document, which aimed to collect core materials on topics pertinent to the review. Suggestions and recommendations made by the respondents were extracted from the submissions and mapped against each of the solution types (for example, data representativeness and fidelity or optimising deployment), following the same process as the resources from the landscape mapping.