OBJECTIVES : To explore whether the OMOP common data model enables reliable estimation of healthcare utilisation and cost data in COPD across Europe.

METHODS : We explored this using healthcare data from the UK (CPRD, HES) and the Netherlands (IPCI). The OMOP common data model standardises the structure, content, and coding systems of otherwise disparate datasets and allows the sharing of common analytical code across a data network without sharing data.

RESULTS : Costs are rarely observed in European healthcare datasets, and instead must be derived by applying unit costs to utilisation data. Where costing is at the level of healthcare visits this is simple to implement assuming the visit type is represented in OMOP. Standard visit concepts in OMOP are available for many common visit types including outpatient and inpatient visits. Currently, contacts in CPRD and IPCI are mapped to ‘outpatient’ visits but more nuanced concepts are often needed to reflect healthcare delivery, for instance, primary care physician or nurse practitioner visits. Further, the nature of the visit may matter, e.g. whether in clinic, by telephone, or home visit. There are also important differences in the scope of source data concepts from IPCI and CPRD mapped to a single visit concept in OMOP, with IPCI adopting a more restrictive selection. Sometimes costs may depend in complex ways on various source concepts spread across OMOP tables, e.g. inpatient care costs in England. In such situations there may be value in costing source data prior to mapping.

CONCLUSIONS : OMOP has the potential to support the modelling of healthcare utilisation and cost data across European healthcare databases, but currently there is considerable heterogeneity in the mapping of source data to OMOP, limiting the reliability of standardised analyses. To maximise the value of OMOP to HTA, greater consideration must be given to the mapping of healthcare utilisation and cost data.