Now that some mental-health Trusts and organisations have recently changed from using GRiST in paper format to accessing eGRiST, the GRiST database is steadily accumulating. At present there are approximately 20,000 anonymous patient risk profiles and clinicians associated risk judgments available for analysis, with the database filling at a rate of about 1400 new assessments per month. This volume will increase as more Trusts begin to use e-GRiST.
These empirical data will permit on-going analysis of risk assessment decisions: how patient data influences clinicians risk judgements, which are the most important pieces of information, how consistent clinicians are, and how to improve the clinical decision support system so that it targets the relevant data at the right time and provides the most appropriate advice. No other tool is able to do this because they do not have a formal specification of risk data at the right level of detail and they do not automatically record clinical risk judgements alongside the data in a database that facilitates mathematical analyses.
These analyses will highlight where and how clinical practice can be improved. In particular they will highlight where there is evidence of any health disparities related to different patient groups, and they will also help with risk prediction. Further, the holistic nature of GRiST allows us to examine relationships between patient cue clusters (e.g. depression, female gender, menopausal age) and clinicians risk judgments, and how these link to interventions and outcomes. This will begin to provide answers to questions about which interventions are best for which patients in which circumstances, and about intervention and outcome expectations for each mental health care cluster. Answers to these questions are currently not available, but they will provide much greater clarity in the planning and costing of episodes of mental health care as part of the Payment by Results initiative.