2015 American Journal of Gastroenterology Lecture

How Digital Health Will Transform Gastroenterology

Brennan Spiegel, MD, MSHS


Am J Gastroenterol. 2016;111(5):624-630. 

In This Article

Digital Health Technology No. 3: Wearable Biosensors

Ubiquitous broadband networks and penetration of smartphones across socioeconomic strata have spurred a technical revolution in remote patient monitoring. In addition to collecting PROs from smartphones, it is now possible to supplement PROs with PRIs, such as step counts, stairs climbed, caloric expenditures from exercise, heart rate, and sleep parameters found on commonly available biosensors. In addition to consumer biosensors, there are Food and Drug Administration (FDA)-approved, medical-grade sensors that monitor a range of physiological data, from heart rhythms[23] to medication adherence[24] to GI motility.[25,26]

In theory, remote monitoring of PRIs can potentially "fill in the blanks" by building a more complete and accurate picture of disease progression in the real-world environment. This capability is part of modern "population health," as defined in the post-ACA era. In particular, if remote digital monitoring can identify subtle but meaningful clinical changes that predict expensive outcomes across large patient groups, then it might improve population health management on a large scale.

Despite the hype and promise of using digital monitoring for patient care, many of its purported benefits have not yet been rigorously tested at scale. There is a pressing need for more research supporting population health monitoring with digital devices. It remains unclear whether PRIs from wearable biosensors meaningfully correlate with clinical outcomes, how they should be collected at scale for population health management, and how to interpret the results in the context of other outcome measures, such as PROs or laboratory markers. For these reasons, the NIH is now examining the role of wearable biosensors as part of the national Precision Medicine Initiative.

Our research team at Cedars-Sinai is also studying the role of wearable biosensors in patients with conditions across specialties. By remotely measuring PROs and PRIs at the same time while patients receive treatment, we can correlate biosensor data with patient symptoms and outcomes. For an example outside of gastroenterology of how PRIs may be deceiving, consider Figure 4, which shows the results of remotely monitoring pain scores (a PRO) against step counts (a PRI) in a 72-year-old woman with rheumatoid arthritis receiving biological therapy. Prior to the study, the patient explained that she refuses to let her disease stop her from her daily activities but regrets how it keeps her from her true passion: writing. She found relief during the treatment course of the study, and her pain, fatigue, and joint stiffness all improved dramatically over the course of 30 days. Yet, Figure 4 reveals an unexpected disconnect between the PRO and PRI data. There is a steady symptom improvement in the left panel, yet the subject moved less and less during the treatment period, shown in the right panel. This counterintuitive pattern became clear only after speaking with the patient: the symptom improvement allowed her to return to her work, writing. Stiffness and pain in her knees and ankles had previously prevented her from sitting at her computer for long periods of time, a necessary prerequisite for the type of work she was passionate about. As her symptoms lessened, sitting at her desk for many hours and registering little movement beyond typing was an "activity" she could return to with gusto. Diminished pain and stiffness in her knees and ankles allowed her to be less physically active and to achieve a greater quality of life in the process.

Figure 4.

Diverging data. A patient with arthritis receiving biological therapy achieves dramatic improvement in pain (left panel), yet nearly stops walking during the same 30-day time period (right panel). See text for an explanation for this seeming paradox.

Only by contextualizing the patient's PRO and biosensor data with a face-to-face interview did the data streams give a clear picture of her progress and its effects on her quality of life. Had we been monitoring this patient from a remote digital coordination facility, tracking her steps as a surrogate measure of illness severity, we would have been misled. Instead, her inactivity was evidence of her improvement. Our assumptions were violated, showing that we must be careful about overinterpreting wearable biosensor data and always maintain a human-to-human exchange.

Nonetheless, published reports show that wearable biosensors, when used in the right setting, can be highly accurate in monitoring patients. Medical-grade biosensors are validated to remotely monitor blood glucose, survey for falls, assess dementia, identify arrhythmias, track vital signs, predict asthma exacerbations, and measure levels of physical and emotional stress, among other applications. In gastroenterology and hepatology, however, there are very few examples, to date, of applying wearable biosensors. This is expected to change soon; investigators are beginning to collect PRIs in patients with chronic digestive diseases, such as inflammatory bowel disease or irritable bowel syndrome.

In the absence of validated wearable biosensors designed for patients with GI disorders, our research team developed, tested, and obtained regulatory approval for a device called AbStats.[25,26] Cleared by the FDA in December 2015, AbStats is a biosensor that non-invasively measures intestinal activity—similar to a "gut speedometer." The sensor is disposable, about the size of a large coin, adheres to the external abdominal wall, and has a small microphone inside that monitors intestinal sounds. A specialized computer analyzes the results and presents a value called the "intestinal rate," measured in acoustic events per minute. If the intestinal rate is very high, like 30 or 40 events per minute, then it indicates the gut is revved up and active. If it is very low, like 1 or 2 per minute, then it indicates the intestines are dormant or, possibly, even dysfunctional depending on the clinical scenario. Research shows that AbStats can predict which postoperative patients will develop ileus; this can help make evidence-based feeding decisions based on dynamic changes in postoperative intestinal rate. We are also evaluating the role of AbStats to track gastric-emptying time non-invasively, monitor outcomes in other functional GI and motility disorders, track food intake, and distinguish ileus from bowel obstructions, among other applications.