Business intelligence in health marketing and mental health: A structural topic modeling study.

Abstract

Mental health disorders represent a major global challenge, driving demand for scalable, evidence based strategies in early diagnosis, prevention, and population surveillance. Business intelligence (BI) research in this domain remains conceptually and methodologically fragmented, impeding interdisciplinary synthesis. This study investigates: the main themes in BI applications for mental health and healthcare over the past 30 years; patterns within and across abstracts of BI relevant articles, including integration potential; and themes with the strongest citation. By doing so, we contribute to health marketing scholarship by synthesizing fragmented BI applications into a Service-Dominant Logic SDL informed, market-shaping intelligence perspective. This reframing is an interpretive synthesis: it is informed by the STM findings and developed through their integration with established health marketing theory, rather than being a direct empirical output of the topic model. Using structural topic modeling (STM) on 2,227 peer reviewed articles from the Web of Science Core Collection (1990-2025), after PRISMA based screening of 20,634 records, the analysis reveals 20 latent topics that trace the evolution of BI, data mining, and dashboard analytics in mental health. We contribute to the literature by offering a novel perspective on Business Intelligence in health marketing, showing how fragmented analytical streams can be synthesized into a Service Dominant Logic informed Mental Health Intelligence Platform under conditions of interdisciplinary fragmentation and temporal evolution Results reveal a shift from individual psychological foci toward systemic, computational approaches underpinning Mental Health Intelligence Platforms (MHIPs). This study outlines a research agenda for empirically evaluating Mental Health Intelligence Platform MHIP designs in future health marketing research.