Using structural topic modelling to reveal patterns in reports on opioid drugs in a pharmacovigilance database

Background: Adverse drug reaction (ADR) reports in pharmacovigilance databases often contain coded information and large amounts of unstructured or semi-structured information in plain text format. The unstructured format and sheer volume of these data often render them neglected. Structural topic modelling (STM) represents a potentially insightful way of harnessing these valuable data and to detect grouping or themes in spontaneous reports to aid signal detection. Purpose: This was an explorative study of the potential for structural topic modelling to identify useful patterns in ADR reports involving opioid drugs in a pharmacovigilance database. Methods: A dataset of ADR reports on opioid drugs reported to the Netherlands Pharmacovigilance Centre Lareb from 1991 to December 2020 was used, comprising a total of 3069 unique reports. Qualitative text analysis was combined with STM, an automated text analysis method, to examine these data. Results: In reports submitted directly by patients and healthcare professionals, 11 meaningful topics were identified, whereby patient experience reports, particularly in relation to pain (relief), and the timing of intake and ADRs of tramadol and paracetamol, were the most common. Of the 12 topics identified in reports received via marketing authorization holders, patch and skin-related side effects, addiction and constipation were the most prevalent. Conclusions: The STM-based analysis identified information that cannot always be captured by coding with the Medical Dictionary for Regulatory Activities (MedDRA®). The identified topics reflect findings in the literature on opioids.