New research has been published in BJPsych Open by the Cassel Hospital Charitable Trust’s (CHCT) collaboration between King’s College London, and the University of Bristol.

The project used novel approaches to explore measures of severity of ‘personality disorder’ using routine data from electronic health records (EHRs).

Novel approach to health service data

Research lead Dr Jonathan Monk-Cunliffe of Bristol University explains: “Personality disorders are a contested diagnosis, and people have different ideas about how we should understand or describe the distress that people who may be given the diagnosis experience. Our research didn’t set out to address this, but we did want to see if we can learn something about the experience of people who have been given the diagnosis.

“People diagnosed with personality disorder can experience significant distress, and there is evidence that people with more severe symptoms might have worse health outcomes. New diagnostic criteria for personality disorder include an assessment of severity, which is important because of its link to future health, but there is not a simple way to capture this in routine clinic practice. We wanted to see if there was a way to measure this using data that is already collected, which might help us understand what people are experiencing and help them get support earlier.

“We explored whether this could be done using a measure called Health of Nation Outcome Scales (HoNOS), which is routinely recorded in electronic health records and includes items relating to a number of personality disorder symptoms. We looked at the association between HoNOS scores and health service use, using data for 6,912 people available in Clinical Records Interactive Search (CRIS), an anonymised and ethically approved version of electronic health records.

“We found that for people diagnosed with personality disorder, the self-harm HoNOS domain might be one potential indicator of the severity of personality disorder, because it is linked to future service use. As with any research there are limitations to our work, and to be clear we are not saying the HoNOS measure of self-harm should be used to make a diagnosis of personality disorder or to give a severity rating – these should only be considered with thorough assessment. However, it might be one helpful indicator for some people.

“We looked at HoNOS as this is one of the only measures that is routinely recorded, but it might not capture the things that are important to people, such as relationships or daily activities. We didn’t find any association when looking at HoNOS scores for these things which we think is an important finding. New measures like ReQOL or DIALOG might do this better as it is really important that research can find a way to measure the things that are meaningful to people.

“There are also limitations to using health record data, but some of these are addressed when looking at summary information for a large number of people.

“One difficulty we had was in combining a number of different outcomes into a single measure so that we could run an appropriate analysis. To do this we used cost to combine different outcomes together – this wasn’t because we were interested in cost itself, but because it provided a way to bring different data together.

“The word limit of the short report also meant there were lots of other things we would have liked to explain in more detail, but weren’t able to.”

Supporting early career researchers

This project fits within the wider strategy of the CHCT Research Programme to promote research within the field of personality disorders, which is often a neglected area, alongside supporting the development of early career researchers.

Dr Monk-Cunliffe continues: “On a personal level I learnt a lot about the application of complex statistical methods to a large dataset, which has given me the skills to conduct further research in this area. It was a good learning experience to present our preliminary results at a conference for the International Society for the Study of Personality Disorders, and I am very pleased that the findings have now been published.

“I’m grateful for the support and mentoring from colleagues within the team at a formative stage in my career as a psychiatrist and researcher.”

”
Next steps

This study completes Phase 2 of the research collaboration between the CHCT, King’s College, London and Bristol University. The research team will continue to answer clinically relevant questions about personality disorder, using the rich resources available in Clinical Records Interactive Search (CRIS) and other big data sources. This is closely linked to other work exploring ideas around ‘recovery’.

The study was led by Dr Jonathan Monk-Cunliffe from the University of Bristol and Dr Giouliana Kadra-Scalzo from King’s College London, with support from Professor Paul Moran (University of Bristol), Dr Richard Hayes (KCL) and Dr Oliver Dale (Clinical Director, Community Mental Health Transformation Programme, Sussex Partnership NHS Trust).

With thanks to the Cassel Hospital Charitable Trust and the Luff Foundation for their funding support.

This article was updated on 1 October to provide more information on the research.