EPR research
There is evidence that people with severe mental illness (SMI) in long-term care are at increased risk of incorrect and incomplete diagnoses and inadequate treatment. This can hinder recovery in various areas of life and increase dependence on long-term care. To improve care for these individuals, insight is needed into the factors associated with effective recovery-oriented care. However, such knowledge is scarce for people with SMI because they are often excluded from scientific research, and when they do participate, they frequently drop out due to the severity and complexity of their mental health issues. To date, there is insufficient knowledge about how the recovery process for this patient group unfolds and which factors act as facilitators or barriers to the recovery process, making it difficult to make evidence-based recommendations for more appropriate recovery-oriented care.
Electronic patient records (EPRs) represent a valuable source of information for this purpose but have remained underutilized. EPR data is available for all patients without imposing additional burdens on them, is described in real-time and in the patients’ daily context, and is particularly well-suited for explorative research. In addition to the structural Routine Outcome Monitoring outcomes, EPRs contain a wealth of qualitative information in the form of natural language. Using innovative machine learning techniques such as natural language processing, qualitative data from the EPRs can be analyzed in order to map recovery patterns across multiple life areas. Trained models can then be used to analyze data from other organizations without needing to exchange such data, allowing outcomes on factors that are associated with recovery to be compared across organisations. These outcomes may be used to make recommendations for the future in the field of research and practice regarding long-term care for patients with SMI.