Regional variation, health inequalities, ethnic density, and antipsychotic prescribing: insights from a national observational study

Abstract Introduction Antipsychotic medications are a cornerstone of treatment for severe mental disorders. Global trends show increased antipsychotic use, yet little evidence exists on how prescribing varies geographically or relates to health inequalities and ethnic density in the UK. Aim To examine national, regional, and Integrated Care Board (ICB)-level trends in antipsychotic prescribing across England between April 2019 and March 2024, and to explore associations with health inequalities and ethnic density. Methods Using routinely collected prescription data from OpenPrescribing,1 we conducted a population-based observational study. OpenPrescribing compiles monthly data on prescriptions dispensed in primary care, published by the NHS Business Services Authority (NHSBSA). It excludes prescriptions issued outside England or dispensed in hospitals or prisons. Prescription data for each antipsychotic were extracted by chemical name and analysed at the ICB level, with 42 ICBs grouped into seven regions. OpenPrescribing has been widely used in previous research.2 Data on health inequalities, ethnicity, and population were obtained from NHS publicly available documents, including NHS England health inequality indices, NHS Digital’s Hospital Episode Statistics, and the Waterfalls analysis. Pakistani ethnic density was the focus because this group shows distinct patterns in mental health outcomes and psychotropic use in prior research, and exhibits strong geographical clustering, allowing clearer estimation of ethnic density effects at ICB levels. Temporal trends were modelled using linear regression, while generalised additive models (GAMs) examined associations between prescribing rates, ICB-level inequalities, and proportions of the British Pakistani population. Results Over the five-year study period, antipsychotic prescribing increased from 183.2 to 197.3 prescriptions per 1000 population. This growth was driven by second-generation antipsychotics (SGAs), rising from 168.3 to 184.4, while first-generation antipsychotics (FGAs) declined from 15.0 to 12.8. Regional growth varied: London showed the highest annual increase (4.4%, 95% CI = 3.8–5.0), whereas the South-West had minimal change (1.0%, 95% CI = −0.3–2.2). At ICB level, changes ranged from a 5.9% rise (95% CI = 4.2–7.6) in Bedfordshire, Luton and Milton Keynes to a 2.0% fall (95% CI = −2.5−−1.7) in Staffordshire and Stoke-on-Trent. GAMs revealed complex, non-linear associations between prescribing patterns, socioeconomic inequalities, and Pakistani ethnic density. Total prescriptions varied non-linearly with health inequalities (Estimated Degrees of Freedom EDF = 7.17, p 0.01). Areas with higher Pakistani ethnic density showed a general but non-significant decline in prescribing (EDF = 2.62, p = 0.07). For FGAs, both health inequalities (EDF = 1.00, p 0.01) and Pakistani ethnic density (EDF = 3.14, p = 0.02) were significant, with the latter showing a non-linear pattern. In contrast, SGA use decreased linearly with increasing Pakistani ethnic density (EDF = 1.00, p = 0.04). Conclusion Antipsychotic prescribing in England has increased in recent years, with marked regional and ICB-level variation. Prescribing rates showed complex, non-linear associations with deprivation, and tended to be lower where Pakistani ethnic density was greater. These patterns highlight the importance of considering social and community contexts when addressing inequalities in mental health treatment. A key strength is the multi-tiered analysis offering deeper insights into geographical patterns, while a limitation is the absence of patient-level data, such as clinical indication, which limits the ability to distinguish between appropriate and potentially inappropriate prescribing.