Lay Summary

• We linked data from two datasets, namely electronic mental health care records from people who received mental health treatment from the South London and Maudsley (SLaM) NHS Foundation Trust, with benefit records from the Department for Work and Pensions (DWP). • We found that 84% of people with linked data had received benefits, with men, older people, those who had died, those who had received a primary psychiatric diagnosis and those living in an area of higher deprivation were most likely to have received benefits. • We will use this linked data to further explore the impact of the benefits system on people with mental health problems.

Background

1.8 million people face long-term sickness absence of four weeks or longer, costing the UK society £100 billion annually. Mental disorders are one of the most common causes of sickness absence. Yet, little is known about the interrelationships between employment status, benefit receipt and mental disorders. This study describes the outcomes of a data linkage between mental healthcare records from SLaM and administrative records from DWP.

Methods

448,404 IDs of patients who accessed secondary mental healthcare services at SLaM were sent to the DWP, including personal identifiers. A deterministic linkage approach was applied. Data from SLaM covers years 2007-2019, whereas data from DWP covers years 2005-2020.

Results

A linkage rate of 92.3% was achieved. Women, younger people and people from ethnic minority groups were less likely to be successfully linked. Patients who had died, had received a primary psychiatric diagnosis, and had a higher number of historical postcodes available were more likely to be successfully linked. Of the patients who were successfully linked, 83% had received benefits. Men, older people, those who had died, had received a primary psychiatric diagnosis and those living in an area of higher deprivation were most likely to have received benefits.

Conclusion

This novel data linkage is the first of its kind to demonstrate the use of routinely collected mental health and benefits data. We will examine patterns of benefit receipt by sociodemographic, clinical and treatment factors and how these vary over time and by psychiatric diagnosis. We will also investigate how these factors predict return to work and the effectiveness of specific interventions on the likelihood and timeliness of return to work, and how this varies by psychiatric diagnosis.