Self-harm and suicide are common among prison inmates, but less is known about these phenomena in those with psychosis.

The aim of this study was to examine self-harm behaviour in New South Wales (NSW) prisons in Australia among inmates diagnosed with psychosis. This study also examined self-harm-related alerts applied by Corrective Services to assist staff with the management of the security and well-being of inmates.

A retrospective case-control data-linkage study was conducted using administrative data collections in NSW, Australia.

The study included all individuals diagnosed with psychosis and incarcerated between 2001 and 2020 in NSW as cases and an age and sex matched control group with no such diagnosis with a record of incarceration in the same time period.

The primary outcome measure was self-harm among the cases and controls. The secondary outcome measure was the application of alerts by Corrective Services in relation to self-harm incidents.

Multivariate regression analysis was used to examine predictors of self-harm in prison. Prisoners with psychosis (n=14 900) were more likely to self-harm than controls (n=2713), with 15.0% versus 3.6% engaging in self-harm (highest odds of self-harm observed in those with schizophrenia and related psychoses, aOR=4.84, 95% CI: 3.93 to 5.98). Those of Aboriginal heritage had an increased risk of self-harm (aOR=1.58, 95% CI: 1.43 to 1.75). Factors associated with a lower risk of self-harm were male sex and older age (≥25 years) at the time of their first incarceration. 35.6% of those released from prison with a prior psychosis diagnosis had at least one alert applied during incarceration compared with 10.1% of prisoners without a diagnosis of psychosis. Overall, 35 individuals with psychosis and 1 individual from the control group died while in prison between 2001 and 2020. 17 prison suicides were recorded from the study population; all occurred in the psychosis group.

Given the heightened risk of self-harm in those with histories of psychosis, consideration should be given to sharing mental health information between agencies to improve the care and management of this group during incarceration. Prison alerts may be a useful tool to help staff manage inmates’ well-being if used appropriately.

This study aims to determine key workforce variables (demographic, health and occupational) that predicted National Health Service (NHS) staff’s absence due to illness and expressed intention to leave their current profession.

Staff from 18 NHS Trusts were surveyed between April 2020 and January 2021, and again approximately 12 months later.

Logistic and linear regression were used to explore relationships between baseline exposures and four 12-month outcomes: absence due to COVID-19, absence due to non-COVID-19 illness, actively seeking employment outside current profession and regularly thinking about leaving current profession.

22 555 participants (out of a possible 152 286 employees; 15%) completed the baseline questionnaire. 10 831 participants completed the short follow-up questionnaire at 12 months and 5868 also completed the long questionnaire; these participants were included in the analyses of sickness absence and intention to leave, respectively. 20% of participants took 5+ days of work absence for non-COVID-19 sickness in the 12 months between baseline and 12-month questionnaire; 14% took 5+ days of COVID-19-related sickness absence. At 12 months, 20% agreed or strongly agreed they were actively seeking employment outside their current profession; 24% thought about leaving their profession at least several times per week. Sickness absence (COVID-19 and non-COVID-19 related) and intention to leave the profession (actively seeking another role and thinking about leaving) were all more common among NHS staff who were younger, in a COVID-19 risk group, had a probable mental health disorder, and who did not feel supported by colleagues and managers.

Several factors affected both workforce retention and sickness absence. Of particular interest are the impact of colleague and manager support because they are modifiable. The NHS workforce is likely to benefit from training managers to speak with and support staff, especially those experiencing mental health difficulties. Further, staff should be given sufficient opportunities to form and foster social connections. Selection bias may have affected the presented results.