The term Big Data is an elusive term with a definition that is not commonly agreed upon. According to a recent definition, the term would encompass any data that is around a petabyte (1015 bytes) or more in size. 1 In health informatics research, Big Data of this size is quite rare. Nevertheless, data used for health informatics research can be considered Big Data in terms of their richness of clinical information and complex structure and may represent an important source of information

more »

... t to the clinical decisionmaking in real-world patient settings. 2 Administrative data bases Rapsomaniki et al. 3 used samples of national, ongoing, unselected record sources to assess three outcomes: all-cause death; a composite of myocardial infarction (MI), stroke, and all-cause death; and hospitalized bleeding. Patients aged 65 years and older entered the study 1 year following the most recent discharge for MI. The data bases used were the Echantillon Generaliste des Beneficiaires from France, the CALIBER research platform of primary care linked via Myocardial Ischaemia National Audit Project (MINAP), the Hospital Episodes Statistics database, and the nationwide cause-specific mortality database from England, the National Inpatient Register, the Swedish Prescribed Drug Register, and the Cause of Death Register from Sweden, and the Medicare database from the USA. Findings The information from these large administrative databases provided an overview of the effectiveness of specific procedures. For example, coronary artery bypass graft and coronary angioplasty results were described in the four populations. The study found that compared with patients from France (64.6%), patients from Sweden (61.3%), England (42.8%), and the USA (59.8%) were less likely to have undergone revascularization either by coronary artery bypass graft or percutaneous coronary intervention (PCI). Unadjusted 3-year cumulative risk of death was considerably lower among English patients (19.6%) than the corresponding figures from the USA (30.2%), Sweden (26.9%), and France (22.1%). After adjustment for age, sex, comorbidities (diabetes, heart failure, peripheral artery disease, renal disease, and chronic obstructive pulmonary disease), and revascularization, mortality rates were equivalent among England, Sweden, and France, but the mortality rates for the USA remained significantly higher even after adjustment for these variables. Thus, compared with Europeans, Americans had an increased risk of death. Adding age and sex as covariates Looking at the data in different formats enables people to see new information and may explain why Americans had an increased risk of death. The prognostic importance of age and sex appeared greater than that of comorbidities or revascularization in Europe than in the USA. Conversely, being affected by comorbidities increases the likelihood of adverse outcomes in the USA but not in Europe. This is shown in the following data provided by Rapsomaniki et al. 3 Sweden was chosen in reference to the outcomes. The lowest unadjusted relative risk of death was observed in England (RR, 0.78; 95% CI, 0.70 -0.88) followed by France (RR, 0.84; 95% CI, 0.70 -1.00) and the highest in the USA (RR, 1.37; 95% CI, 1.15-1.64). With adjustment for age and sex, there was a 20% relative risk increase for England and France, and post-MI survival was no longer significantly associated with better outcomes when compared with Sweden (RR, 0.95; 95% CI, 0.79-1.15, and RR, 1.12; 95% CI, 0.91-1.40, respectively). All results remained unchanged after adding comorbidities and revascularization use to the model (RR, 0.84; 95% CI, and RR, 1.15; 95% CI, for England and France, respectively). In contrast for the USA, age and sex did not account for the higher risk of death in post-MI patients (RR, 1.37; 95% CI, 1.26 -1.49). With further adjustment for comorbidities, there was a 14% relative risk reduction, even though post-MI survival was still significantly associated with poorer outcomes (RR, 1.13; 95% CI, 1.03-1.25). Results remained unchanged after adding revascularization use to the model (RR, 1.14; 95% CI, 1.04-1.26). Many questions remain unanswered. Are there differences in interpretations of ethnic groups within the databases? South Asian immigrants to North America have been noted to have 1.5 -4.0 times higher coronary heart disease mortality rates than indigenous populations and tend to be more comorbid. 4 What is the potential impact of revascularization facilities on patient outcomes? The numbers reported The opinions expressed in this article are not necessarily those of the Editors of the European Heart Journal -Quality of Care and Clinical Outcomes or of the European Society of Cardiology.