By Skylar Cohen
A multidisciplinary team at Stanford Medicine has envisioned Green Button, a proposed system designed to let doctors receive guidance for treating patients with combinations of characteristics that have not been examined in clinical trials. The idea was outlined in an article in the journal Health Affairs last year.
The Green Button would be a tool for doctors to search among millions of electronic medical records (EMRs) to retrieve valuable information. Doctors could examine similar cases and results of previous treatments. Although this would not provide the experimental conditions of a clinical trial, it would provide the doctor with experiences from many past doctors.
Green Button could provide critical help in a wide variety of scenarios. Christopher Longhurst, clinical professor of pediatrics in system medicine and chief medical information officer for Stanford Children’s Health, told Inside Stanford Medicine that in many situations there are not clinical trials applicable for a particular patient. Nigam Shah, assistant professor of biomedical research and assistant director of the Stanford Center for Biomedical Informatics Research, was also quoted in the Inside Stanford Medicine article saying that there is only about a four percent chance that a clinical trial will match the patient.
Green Button was inspired by a 2010 incident involving Jennifer Frankovich MF ’05 CRT ’08 MS ’09, in which a girl arrived to the Lucile Packard Children’s Hospital displaying a combination of lupus and several regions of severe inflammation. Frankovich did a real-time search of past EMRs for previous patients with similar symptoms. Using these records, Frankovich and her colleagues were able to determine the proper course of action (which involved immediately providing anticoagulants).
Technology today naturally encourages this type of project. Technical improvements and infrastructure developments allow for faster retrieval times and more comprehensive data.
The challenge, however, lies in getting a sufficiently large pool of patient data, given that the project by definition seeks to represent as many scenarios as possible. If combined, the EMRs of many institutions could provide a vast record: Stanford Health, the University of California system and Kaiser Permanente together have around 25 million EMRs. Questions remain, however, about whether enough institutions would be willing to contribute EMRs due to patient privacy concerns.
One solution would be to have a system that does not involve directly contributing EMRs, so as to better protect patient privacy. Rather than combing through all of the EMRs from a separate institution, a doctor might send an initial request for a particular type of patient, and an outside institution might just return a summary of the data from applicable EMRs.
Contact Skylar Bromley Cohen at skylarc ‘at’ stanford.edu.
A previous version of this article did not attribute information presented to an article by Inside Stanford Medicine. The Daily regrets this error.