What have we learnt from involving and engaging with patients and the public in our health data science projects?

Embedding patient and public involvement and engagement (PPIE) in our analytical projects is still a relatively new way of working for us. PPIE allows us to involve patients and the public in decisions about what we do and how we interpret and communicate our analysis. This also means that research is done in accordance with the ethical principle of ‘nothing about us, without us’.

Last year we wrote about why PPIE is important, and some challenges we faced. Now, in this blog, I share the lessons we have learnt from some of our recent PPIE activities.

Our project focused on ethnicity and survival of people with multiple long-term conditions, using data from general practices to examine inequalities relating to ethnicity. We found that patients from minority ethnic groups with multiple long-term conditions had higher relative mortality than patients from white ethnic groups. This project benefited enormously from engaging with patients to improve our understanding of their experience and perspectives.

Four lessons from our PPIE work

1. If you’re new to this, work with experienced PPIE practitioners

From the outset of this project, we wanted to incorporate patient and public involvement and engagement in order to sense check whether we were asking the right questions and whether our results reflected real-world experiences. With the help of Human-Centred Health, an external team specialising in engagement and involvement, we refined the goals of our PPIE activities, identified the groups of patients and carers we wanted to involve, and tailored our questions to present our work appropriately.

We convened people from diverse ethnic groups with varying experiences of managing their own long-term health conditions or caring for someone with long-term health conditions. We used this open resource guide, selecting the activities that matched our engagement aims. We conducted one-to-one interviews with patients, and these discussions fed into our analysis plan. Participants suggested that we listen to a variety of experiences for our next activity, so we also ran a focus group session at a later stage of our project.

2. Incorporate lived experiences when developing and refining analysis plans

Participants weren’t surprised to hear our results. Our findings resonated with their lived experience, and the media had drawn attention to similar ethnic inequalities during the COVID-19 pandemic. The group suggested several possible explanations for these inequalities, including societal factors and aspects of health care which led to different socioeconomic opportunities and different capacity to self-manage. They pointed out that these have long-standing, even intergenerational, effects. The group gave us valuable insight into the numerous ways the inequalities we identified could arise. We have considered how participants’ ideas could be tested in future analysis and whether additional variables in the data and additional data sources may better represent their experiences.

3. Ensure the people you engage with understand what you’ll do with the findings

We do analysis to highlight an issue, understand the underlying causes of a problem, or test potential solutions. Our discrete pieces of analysis are part of a wider programme of analytical and influencing work, ultimately aimed at achieving change. However, the focus group participants were frustrated with hearing the same messages from researchers and wanted action. Next time, we will give participants more detail about the role of research and how we see our work leading to action through contributing to the evidence base which policymakers and others draw on. Ensuring that participants know what we’ll do with our findings may lead to greater discussion of how we could increase the influence of our work.

4. Make the findings relevant to patients and the public

Our research was high level and was primarily aimed at influencing national thinking and policy. Participants saw the value of this but also wanted to see more detailed analysis (eg for specific age groups or gender or specific health conditions). Although we were already aware of the value of detail, especially to help local teams dig deeper and find areas for improvement, we now see that this may also help engage and inform patients. Disaggregating our analysis into more detailed groups could mean that we are able to provide insights directly relevant to patients and the public.

Next steps for our PPIE work

We, like other organisations, are still learning about what makes for effective PPIE in health data science. We’d like to hear from you about your experiences from your own PPIE journey, what you’ve learnt and how you’ve implemented this within your ways of working.

We all share the group’s frustrations with the persistence of inequalities, which we know have been exacerbated by the pandemic. Having this relationship with patients and the public means we can be closer to aligning our work and intended impact with the issues that affect them the most.

Anne Alarilla (@AlarillaAnne) is a senior data analyst in the data analytics team at the Health Foundation.