Promising Use of Big Data to Increase the Efficiency and Comprehensiveness of Stroke Outcomes Research.

The gold standard for outcome assessment in stroke is provided by longitudinal population-based incidence studies with full case-ascertainment and repeated outcome assessments over time. This enables a comprehensive assessment of the burden of stroke that includes mortality, disability, adverse events, quality of life, and functional status and the effectiveness of the health system in relation to healthcare utilization. However, such purposeful collection of continuous data over a period of time for stroke surveillance is costly, and some patient-reported outcomes such as readmissions can be unreliable. It is, therefore, unsurprising that the number of currently ongoing incidence studies is limited. With the explosion of digital information, a broad range of health outcomes after stroke can be evaluated using new strategies for integrating administrative data with data from clinical studies or registries. In this article, we aim to describe approaches for the collection of outcomes data after stroke and their strengths and limitations. We further examine the utility of linking datasets to add value to stroke research, including ongoing stroke surveillance and outcomes reporting, and describe barriers to data linkage and to the use of administrative data in stroke research.