Piloting a minimum data set for older people living in care homes in England: a developmental study

Background: We developed a prototype minimum data set (MDS) for English care homes, assessing feasibility of extracting data directly from digital care records (DCRs) with linkage to health and social care data. Methods: Through stakeholder development workshops, literature reviews, surveys and public consultation, we developed an aspirational MDS. We identified ways to extract this from existing sources, including DCRs and routine health and social care datasets. To address gaps, we added validated measures of delirium, cognitive impairment, functional independence and quality of life to DCR software. Following routine health and social care data linkage to DCRs, we compared variables recorded across multiple data sources, using a hierarchical approach to reduce missingness where appropriate. We reported proportions of missingness, mean and standard deviation (SD) or frequencies (%) for all variables. Results: We recruited 996 residents from 45 care homes in three English Integrated Care Systems. 727 residents had data included in the MDS. Additional data were well completed (<35% missingness at wave 1). Competition for staff time, staff attrition and software-related implementation issues contributed to missing DCR data. Following data linkage and combining variables where appropriate, missingness was reduced (≤4% where applicable). Discussion: Integration of health and social care is predicated on access to data and interoperability. Despite governance challenges we safely linked care home DCRs to statutory health and social care datasets to create a viable prototype MDS for English care homes. We identified issues around data quality, governance, data plurality and data completion essential to MDS implementation going forward.