Back

Developing a research ready population-scale linked data ethnicity-spine in Wales

Akbari, A.; Torabi, F.; Bedston, S.; Lowthian, E.; Abbasizanjani, H.; Fry, R.; Lyons, J.; Owen, R. K.; Khunti, K.; Lyons, R. A.

2022-11-29 public and global health
10.1101/2022.11.28.22282810 medRxiv
Show abstract

IntroductionEthnicity information is recorded routinely in electronic health records (EHRs); however, to date, there is no national standard or framework for harmonisation of the existing records. Methods and analysisThe national ethnicity-spine uses anonymised individual-level population-scale ethnicity data from 26 EHR available through the Secure Anonymised Information Linkage (SAIL) Databank. A total of 46 million ethnicity records for 4,297,694 individuals in Wales-UK over 22 years (between 2000 and 2021) have been compiled in a harmonised, deduplicated longitudinal research ready data asset. We serialised this data and compared distribution of records over time for four selection approaches (Latest, Mode, Weighted-Mode and Composite) across age bands, sex, deprivation quintiles, health board, and residential location, against the ONS census 2011. The distribution of the dominant group (White) is minimally affected based on the four different selection approaches. Across all other ethnicity categorisations, the Mixed group was most susceptible to variation in distribution depending on the selection approach used and varied from a 0.6% prevalence across the Latest and Mode approach to a 1.1% prevalence for the Weighted-Mode, compared to the 3.1% prevalence for the Composite approach. Substantial alignment was observed with ONS census with the Latest group method (kappa= 0.68, 95% CI [0.67,0.71]) across all sub-groups. ConclusionWe provides a reproducible EHR based resource enabling the investigation and evaluation of health inequalities related to ethnic groups in Wales. This generalisable method informs opportunities for the transferability of this methodology across the UK to platforms with comparable routine data sources. Ethics and disseminationThis work was supported by the Con-COV team funded by Medical Research Council, Health Data Research UK, ADR Wales funded by ADR UK through the Economic and Social Research Council, and the Wales COVID-19 Evidence Centre, funded by Health and Care Research Wales. O_TEXTBOXStrengths and limitations of this study O_LIThis is the first comprehensive approach for retrieving ethnicity records from linked EHR. C_LIO_LIThe methodology presented here creates an anonymised longitudinal, population-scale individual-level linked ethnicity spine for the population of Wales. C_LIO_LIEnabling investigation and evaluation of health inequalities related to ethnic groups in Wales. C_LIO_LIEthnicity spine is derived from 26 data sources across 22 years, complementing existing ethnicity data in census by providing a reproducible maintainable RRDA. C_LIO_LIAs this is a data driven approach, the algorithm is limited to the reported ethnicity and actual diversity of the population. C_LI C_TEXTBOX Article summary

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.