Wednesday Sep 09, 2015
Information Overload: New Gadgets Provide More Data, What Else?
Eko Devices Inc.(ekodevices.com) recently achieved FDA approval for Eko Core, which is a digital attachment for an analog stethoscope. Although it's not the first digital stethoscope on the market, it is the first to connect with a smartphone app, and it joins several hundred other devices used to collect data from patient encounters.
Essentially, Eko's device captures audio waveforms from direct auscultation and transmits them to a phone, tablet or computer. Software compliant with the Health Insurance Portability and Accountability Act allows for recording and storage of the collected waveforms, even depositing them directly into a patient's electronic health record.
© Eko Devices Inc.The Eko Core stethoscope attachment allows physicians to stream heart sounds to a HIPAA compliant smartphone app and deposit them directly into a patient's electronic health record.
More than just a gadget, however, this device is another example of the increasingly complex data collection that's becoming ubiquitous for physicians and patients. From exercise tracking to lab values to genome sequencing, we add more complex measurements and data points every day. We measure our inputs and outputs. We track height and weight and body fat percentage. Many patients monitor glucose or INR levels at home. Add to that the reams of data we collect in physician offices and hospitals each day, like laboratory measurements of renal and hepatic function, cholesterol, and EKG data. Terabytes of storage are consumed by imaging data alone.
This glut of data is only going to increase as we develop newer, more specific technologies for tracking our biological processes. We can measure specific disease markers far more efficiently. A little more than a decade ago researchers sequenced the first human genome. Today, you can have your genome sequenced in a few hours for about $1,000.
In 2005, the CDC estimated that 7 billion to 10 billion lab tests(www.cdc.gov) are performed each year in the United States. A safe estimate with no inflation in the number of performed tests still amounts to at least 70 billion new data points in the last 10 years. On top of clinical data, consumer technology allows recording of what were once considered medical data. Glucose monitoring has become commonplace. Pulse oximeters are available at most big box stores. Most of us carry in our pocket enough computing power and storage to track every nuance of our daily lives.
As advanced as our data gathering methods have become, however, we still don't have direct application for most of the mountain of data we collect about each patient. Although the stethoscope attachment mentioned above provides quantifiable, measurable data about the function of a patient’s heart, it still requires context for it to mean anything.
That’s not to say that there is no benefit from recording heart sounds and waveforms. We simply don’t know yet how useful it will be. Will recording data about the heart sounds of this patient allow me to eventually detect changes during future visits? Will those changes have clinical significance? Will these measurements allow extrapolation of conclusions for other patients? These are the questions we should be -- and are -- asking, not just recording for the sake of one more piece of information in the EHR.
This also assumes the stethoscope is a medical device in the hands of a trained clinician. Not all data we collect are recorded by trained scientists. Large proportions of data now come from smartphones and exercise trackers. How significant are the data points collected by patients with home monitoring equipment? How accurate are these data? What, if any, is the context for these data points? Is there a point to all this collection? What do we do with all these datasets?
The answer is simple, but far more complex in execution. We study them. We record all the points we can, with all the context we can obtain, and we analyze them for patterns and correlations. Those correlations are not the end, but just the beginning, giving direction for further study. The real benefit from this mountain of data is a course heading for further randomized-controlled research. Rather than jumping to conclusions based on limited and often context-poor data, we can now make educated decisions about resource allocation for research using the tools at our disposal. Computer systems like Watson comb through millions of data points looking for algorithms that may improve our diagnostics and clinical decision-making.
As primary care physicians, we will be on the front lines for data collection, but we will also reap the benefit of more focused research. In this early phase, that often means information overload. We will be recording data points that may not make sense or have direct application at the moment. If used correctly, however, these measurements will eventually allow for more targeted treatment and better patient care.
Gerry Tolbert, M.D., is a board-certified family physician who practices in northern Kentucky. A lifelong technophile, his interests include the intersection of medicine and technology. You can follow him on Twitter @DrTolbert(twitter.com).
Posted at 12:01PM Sep 09, 2015 by Gerry Tolbert, M.D.