The scope of this guideline is to provide recommendations for studies reporting key health and justice outcomes for discharged forensic patients and people released from prison settings returning to the community. Specifically, the guideline aims to provide recommendations for standardising the way in which post discharge and release community outcomes, including reoffending, readmissions and mortality, are defined, measured, recorded, analysed, and reported. A standardised approach will improve comparability between studies, settings, and samples, allowing future meta-analytic syntheses.
Why a new reporting guideline is needed In undertaking a systematic systematic review and meta-analysis of post-discharge outcomes for forensic mental health patients, it was evident that published studies varied enormously in approaches to outcome reporting. In particular, the following problems were identified:
Most studies provide sufficient demographic details for the main sample, but often fail to report the same details consistently when reporting on the sub-samples that might be different from the main cohort (e.g., patients who remained in hospital vs those who were eventually discharged into the community)
Reoffending outcomes are often inconsistently defined, with some studies using reoffending and reconvictions interchangeably, for example.
Outcomes are sometimes not defined in studies. For example, some authors may provide details on what is counted as “violent offending”, while others do not explicitly define it.
There is sometimes insufficient data transparency regarding the sample/cohort. Information about the location of secure hospitals, catchment areas, and whether the sample is drawn from a previously established dataset is often lackingFollow-up durations are inconsistently reported. This makes it difficult to calculate person-years at-risk and prohibits comparisons being made between cohorts.
Post-release reoffending outcomes for people being released from prison settings are reported without the information required to calculate person-years at risk, which further prevents comparison. These outcomes tend to be reported as cumulative prevalence rates (with varying durations of observation) only.
Registered: 26 October 2026
Last updated:
Contact: James McLauchlan Email: [email protected]