GP prescribing in Northern Ireland by deprivation index: retrospective analysis

In the present study, we sought to explore the relationship between socioeconomic status and prescribing magnitude and cost in primary care throughout Northern Ireland.

We performed a retrospective data analysis of general practitioner (GP) prescribing using open-source databases with data collected from May to October 2019 to determine the number of prescriptions and cost of drugs and drug classes by area, ranking these by deprivation index. We used Kendall's tau to quantify the relationship between prescribing and deprivation.

We analysed open-source data collected from 325 GP practices in Northern Ireland during the period from May to October 2019.

We analysed a total of 2 764 303 prescriptions signed during our study period.

Our study indicates a clear trend of increased overall spending per patient (r=-0.1232, p=0.02) and number of prescriptions per patient (r=-0.3440, p<0.001) in areas of higher deprivation. The mean cost per item was higher in less deprived areas (r=0.3809, p<0.001). Overall, £13.79 more was spent and 3.5 more items were prescribed per patient in the most compared with the least deprived decile, although more expensive items tended to be prescribed in areas with lower socioeconomic deprivation (£11.27 per item vs £9.20 per item). We found a statistically significant correlation of prescribing of key drug classes, such as bronchodilators, antidepressants and drugs used to treat diabetes, among others, with greater deprivation. Prescribing of vaccines and drugs used in the treatment of glaucoma was correlated with lower deprivation.

We provide an exploration of the correlation of prescribing with deprivation by analysing all prescriptions signed within a 6-month period in Northern Ireland. Our study broadly agrees with published literature, although a few notable exceptions are highlighted. We provide evidence of discrepancies in medication cost between areas of differing deprivation and suggest possible explanations for these trends. This information will be valuable for future investigation of disease prevalence, as well as targeting of patient education and future funding.