Abstract
AbstractObjectivesTo address the lack of individual-level socioeconomic information in electronic health care records, we linked the 2011 census of England and Wales to patient records from a large mental healthcare provider. This paper describes the linkage process and methods for mitigating bias due to non-matching.SettingSouth London and Maudsley NHS Foundation Trust (SLaM), a mental health care provider in southeast London.DesignClinical records from SLaM were supplied to the Office of National Statistics (ONS) for link-age to the census through a deterministic matching algorithm. We examined clinical (ICD-10 diagnosis, history of hospitalisation, frequency of service contact) and sociodemographic (age, gender, ethnicity, deprivation) information recorded in CRIS as predictors of linkage success with the 2011 Census. To assess and adjust for potential biases caused by non-matching, we evaluated inverse probability weighting for mortality associations.ParticipantsIndividuals of all ages in contact with SLaM up until December 2019 (N=459,374).Outcome measuresLikelihood of mental health records’ linkage to census.Results220,864 (50.4%) records from CRIS linked to the 2011 census. Young adults (Prevalence ratio (PR) 0.80, 95% CI 0.80-0.81), individuals living in more deprived areas (PR 0.78,0.78-0.79), and minority ethnic groups (e.g., Black African, PR 0.67, 0.66-0.68) were less likely to match to census. After implementing inverse probability weighting, we observed little change in the strength of association between clinical/demographic characteristics and mortality (e.g., presence of any psychiatric disorder: unweighted PR 2.66, 95% CI 2.52, 2.80; weighted PR 2.70, 95% CI 2.56, 2.84)ConclusionsLower response rates to the 2011 census amongst people with psychiatric disorders may have contributed to lower match rates, a potential concern as the census informs service planning and allocation of resources. Due to its size and unique characteristics, the linked dataset will enable novel investigations into the relationship between socioeconomic factors and psychiatric disorders.Article summaryStrengths and limitations of this studyThis is the first time mental healthcare electronic records have been linked to ONS census at the individual-level in England. Due to its scale, ethnic diversity and demographic characteristics, and abundance of detailed information on a variety of socioeconomic and demographic indicators acquired through the linkage to census records, this dataset will enable novel investigations into the causes, trajectories and outcomes of psychiatric disorders.A significant strength of the study is that we could assess and adjust for potential biases caused by non-matching related to age, gender and deprivation.Whilst we observed differences between individuals that matched to census, and those that did not, our weighted analyses were able to show that these differences did not substantially alter associations with mortality outcomes.Due to the nature of the deterministic linkage algorithm, we could not determine the causes of non-linkage.
Publisher
Cold Spring Harbor Laboratory