An application of matching algorithms to generalize small-area estimates of chronic pain prevalence to neighbourhoods across England

Quinn, L; Yu, D; Lynch, M; P Jordan, K; Wilkie, R; Peat, G

Abstract

Background
Local decision-makers lack granular data on the prevalence of chronic pain in their populations. We applied matching methods to generalize estimates from one local survey in England to other neighborhoods across the country with a similar sociodemographic composition.

Methods
We used propensity score matching to match lower-layer super output areas (LSOA) across England with 230 surveyed LSOAs in North Staffordshire by age, sex, ethnicity, deprivation, and rurality. LSOA-specific crude prevalence of chronic pain and high-impact chronic pain in adults aged 35+ years were estimated and mapped for matched LSOAs.

Results
Satisfactory matching was achieved for 24 871 of 31 580 LSOAs (79%). The 6709 LSOAs identified as either “off common support” or unmatched were principally inner-city neighborhoods with younger, more ethnically diverse populations. LSOA-specific estimates of chronic pain and high impact chronic pain ranged from 14% to 52% and from 4% to 31% respectively. Integrated Care Board estimates ranged from 27% to 38% and from 10% to 18%, respectively. Estimates for England were 31.9% and 12.6%, respectively.

Conclusions
Using matching methods we have produced the first detailed map of the distribution of chronic pain in England but with several strong assumptions. Our estimates highlight substantial variation in prevalence within ICBs.

Citation

Quinn, L., Yu, D., Lynch, M., P Jordan, K., Wilkie, R., & Peat, G. (2025). An application of matching algorithms to generalize small-area estimates of chronic pain prevalence to neighbourhoods across England. Journal of Public Health, 1-9. https://doi.org/10.1093/pubmed/fdaf002

• Journal Article Type: Article
• Acceptance Date: Jan 10, 2025
• Online Publication Date: Jan 23, 2025
• Publication Date: Jan 23, 2025
• Deposit Date: Jan 30, 2025
• Journal: Journal of Public Health
• Print ISSN: 1741-3842
• Electronic ISSN: 1741-3850
• Publisher: Oxford University Press
• Peer Reviewed: Peer Reviewed
• Pages: 1-9
• DOI: https://doi.org/10.1093/pubmed/fdaf002
• Keywords: health intelligence, epidemiology, public health
• Public URL: https://keele-repository.worktribe.com/output/1051859