Statistics from Altmetric.com
We live in a time with an abundance of clinical data. Federal investments have led to the development of research data networks for addressing clinical questions (eg, the National Patient-Centered Clinical Research Network,1 All of Us,2 and the Electronic Medical Records and Genomics Network3). Additionally, there is a growing trend towards public data sharing through initiatives by the National Institute of Health.4 Such sharing of data increases transparency of findings and provides opportunities for generating new hypotheses.5
Physician scientists may be best positioned for secondary analysis because of their clinical knowledge. However, many do not have programming and analytical skills to take full advantage of these opportunities. Analyzing existing data offers advantages: it is less time and resource intensive, and it can provide preliminary data for grant proposals.5 6 Despite these advantages, and given the limited availability of resources, secondary analyses remain uncommon among physician scientists.
Traditionally, physician scientists are trained in primary data collection to study clinical phenomena—often involving rigorous designs, patient-oriented interventions, and experimental manipulations. However, strong data management and analytical skills beyond primary data collection are required. Physician scientists must learn to design investigations involving all or part of publically available data, as well …
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.