Spatial Exploration of Deprivation Driven Inequalities in Opioids Prescribing in English Primary Care: a Cross Sectional Analysis

Problem

Over 23 million opioid prescription items were prescribed in English primary care in 2018-2019 and the levels of prescribing are not distributed equally across regions and populations. A greater understanding of the geographical differences and social drivers of opioid prescribing is needed to guide policy responses and interventions. The aim of this study is to explore regional variations in opioid prescribing in England and the role of socio-economic deprivation in driving inequalities in opioid prescribing.

Approach

Cross-sectional study using national prescription data from primary care settings. The prescribing of opioids in each general practice was quantified by Defined Daily Doses (DDDs) and then attributed to 32,844 Lower Layer Super Output Areas (LSOAs), the geographical units representing approximately 1,500 people. Linear regression was used to model the effect of socio-economic deprivation, measured using the Index of Multiple Deprivation (IMD) 2019 quantiles, on levels of prescribing, measured by DDDs/1,000 populations/day. The model accounted for the proportion of females and people over 65-year-old, rurality/urbanity classification of the area, and prevalence of cancer, palliative care, depression, severe mental illness (SMI), rheumatoid arthritis, and obesity. Within higher organisational areas (Clinical Commissioning Groups, CCG), adjusted DDD estimates were obtained and compared at each deprivation level.

Findings

In total, 5,350,151,699 DDD of opioids were prescribed. Prescribing of opioids was not distributed randomly between LSOAs and varied between 4 and 1,905 DDD/1000 population/day. On average, 93.77 (SD: 1.90, p < 0.001) more DDD/1000 population/day were prescribed in the most deprived areas compared to the least deprived areas. Prevalence of cancer and SMI were negatively associated with opioid prescribing, and the remaining conditions had a positive association. Smaller LSOAs and LSOAs with smaller proportion of females prescribed fewer opioids. A higher proportion of population over 65 was associated with higher prescribing. On average, opioid prescribing was 33.6 DDD/1000 populations/day (SD: 2.16, p < 0.001) higher in urban compared to rural areas. Within individual CCGs, the difference between prescribing in the least and most deprived areas varied by up to 200%.

Consequences

Even after adjusting for differences in population structure, opioid prescribing practices differ significantly across the country, and geospatial deprivation is a key prescribing driver. Future policies and interventions should focus on high-prescribing low-geography areas and tackling deprivation driven inequalities.