Ab1671-Pare Rheumatoid Arthritis Patients’ Perspective on the Use of Prediction Models in Clinical Decision-Making

Background: A rapidly expanding number of prediction models are being developed that aim to improve care for rheumatoid arthritis (RA) patients. These models have the potential to decrease diagnostic delay and facilitate more effective treatment. However, few are actually implemented in clinical practice. Objectives: This study explores factors influencing the acceptance of prediction models in clinical decision-making by RA patients. Methods: A qualitative study design was used with thematic analysis of semi-structured interviews. We recruited patients from the outpatient rheumatology clinic of the University Medical Centre Utrecht in the Netherlands. Patients aged ≥18 years with a clinical RA diagnosis were eligible to participate. Purposive sampling was applied based on age, gender and technological proficiency to obtain a large variety of perspectives. The interview topic list was based on pilot data consisting of an open questionnaire, focus group and interview. The semi-structured interviews focused on factors influencing the acceptance of prediction models in clinical decision-making. Data was analysed using the six phases of thematic analysis as described by Braun and Clarke [1]. Data analysis started after the first four interviews and continued throughout. Initial codes that arose were collated into themes through subsequent analysis steps. Results: Data saturation was reached after 12 interviews. Patients were generally positive about the use of prediction models in clinical decision-making. Six key themes were identified from the interviews (Figure 1). First, patients have the need for information about prediction models. They preferred information about possible personal treatment consequences, and were less interested in technical details. Second, factors influencing trust in model-supported treatment were described. For example, patients indicated that if a model included variables capturing for example stress, the weather or diet, this would increase their trust. Patients however also feared that a model cannot capture emotions from patients adequately. Third, patients envision the model to have a supportive role in clinical decision-making. Ultimately, treatment decision are made by the caregiver and patient. Fourth, patients hope to personally benefit from model-supported treatment in various ways, e.g. by reducing the time to find an effective treatment. Fifth, patients are willing to contribute time and effort to contribute to model input (e.g. blood withdrawals), if within reasonable bounds. Lastly, patients described their envisioned effects of the relationship with their caregiver in model-supported treatment. Importantly patients who had trust in their current caregiver did not think that their relationship would change. We also identified relationships between the themes (Figure 1). Trust in model-supported treatment was influenced by the received information on said models. Second, patients envisioned a supportive role for the model due to a lack of trust in the model as a stand-alone system. Third, the willingness to contribute to model input was often motivated by hopes to personally benefit from model-supported treatment. Lastly, the current relationship with the caregiver influenced the trust in model-supported treatment. Conclusion: In this study RA patients were generally positive about the use of prediction models in their treatment given some conditions were met and concerns addressed. The results of this study can be used during the development and implementation in RA care of prediction models in order to enhance patient acceptability. REFERENCES: [1] Braun V, Clarke V. Using thematic analysis in psychology. Qualitative research in psychology 2006;3(2):77-101. Acknowledgements: NIL. Disclosure of Interests: None declared.