The promise of big data for precision population health management in the US.
2020
OBJECTIVES As we enter the year 2020, health data in the United States (US) is still in the process of being curated into a usable format. With coordinated data systems, it becomes possible to answer, with relative certainty, what preventive and medical interventions work in the real world and for whom they might work. STUDY DESIGN This is a non-systematic expert review. METHODS A non-systematic expert review was undertaken to identify relevant scientific and gray literature on the current state and the limitations of evaluation of health interventions and the health data infrastructure in the US. This review also included the literature on nations with unified data systems. We coupled this review with non-structured interviews of data scientists to gain insight into the progress in establishing the components necessary to support a unified data system and to facilitate data exchange for evaluations, as well as further guide our review. Our goal was to produce a critical analysis of the existing attempts to standardize and use data collected during patient encounters with physicians for public health purposes. RESULTS Data obtained from electronic health records are produced in a way that is challenging to use and difficult to compile across platforms in the US. One response to this problem has been to encourage the exchange and standardization of health record information through Distributed Research Networks and Common Data Models (CDMs). These data can be combined with mobile health, social media, and other sources of data to radically transform what we know about the prevention and management of disease. However, issues with the variety of CDMs and growing sense of distrust of institutions that maintain data continue to impede medical progress. CONCLUSIONS We present a framework for data use that will allow public health to answer a swath of unanswered research questions that can improve public health practice.
Keywords:
- Correction
- Source
- Cite
- Save
- Machine Reading By IdeaReader
35
References
4
Citations
NaN
KQI