The Open Mind: The Open Mind
Smartphones have been one of the most rapidly adopted technologies in the history of humankind.1 When these devices are coupled with the near-ubiquitous availability of wireless Internet connectivity, each of us holds previously unimaginable access to information literally in the palm of the hand. This shift in access to information has had far-reaching effects on every aspect of society from our social connections to commerce to health care. Personalized medical data through companies such as 23andMe2 provide the motivated patient an unprecedented ability to learn much about their medical conditions, perhaps more than their physician, depending on the rarity of the conditions. In fact, 23andMe recently partnered with Pfizer to conduct a genome-wide association study of over 75,000 individuals with depression to discover 15 previously unknown genomic regions associated with higher rates of major depression.3 More and more, this decentralization and democratization places the patient in the driver seat and relegates the physician to the role of a (trusted?) mentor. Surely, many of us have encountered at least 1 patient with a differential diagnosis made in consultation with “Dr. Google.” This sort of revolution currently is limited to the diagnostic realms because licensed providers must still be used to legally prescribe any medication or perform any interventional procedures, although it does certainly represent a sea change in the dynamic of the doctor–patient relationship.
The smartphone is fast becoming the technological lynchpin of this revolution in health care. The device itself contains a processor capable of impressive levels of data analysis, given its size. Furthermore, the other onboard technologies such as lights, cameras, microphones, and accelerometers allow for further data acquisition that can contribute to more robust analysis. Some innovators have developed external hardware attachments that allow more specific data acquisition and thus further analytical capabilities by the phone’s central processing unit. Tying all these hardware devices together are specific software programs to turn the data captured into useful information for the user. We know these unifying software programs as applications or “apps.” Most Americans spend about 37.5 hours a month on their smartphones and use about 26 apps per month.4 Even though much of this time is used for entertainment purposes such as games, music, and video, the use of apps specific to health and wellness is soaring as well. The market for such apps is expected to generate US$26 billion in 2017.5 The presence of such a large market has attracted the attention of the innovator and regulator alike.
The Food and Drug Administration (FDA) has had difficulty keeping pace with innovators and has had to rethink its regulatory approach to this rapidly shifting landscape to achieve its mission of “protecting consumers and enhancing public health” while still allowing innovation to flourish. In February 2015, the FDA published a statement to clarify its position on the status of what it terms “mobile medical apps,” which it defines as a software application for a mobile device that is intended to be used as an accessory to a regulated medical device or to transform a platform into a medical device. Given the vast number of applications that could fall under such a broad definition, though, the FDA indicated that they intend to apply this regulatory authority only to “those mobile apps whose functionality could pose a risk to a patient’s safety if the mobile app were to not function as intended.” These guidelines are derived from older FDA policies governing computer software in which if software is used in a medical device, as a medical device, or to manufacture a medical device, then that software is a medical device. In its current guidelines, the FDA states that “the intended use of a mobile app determines whether it meets the definition of a ‘device,’” where intention refers to both the labeling and marketing efforts surrounding a potential device. Realizing that this could still apply to a vast number of application-based products, the FDA further stated that they intend to exercise their “regulatory oversight only to those mobile medical apps that are medical devices and whose functionality could pose a risk to patient safety.” For instance, the FDA has an extensive, but not exhaustive, list of the types of applications for which they intend to “exercise enforcement discretion” (ie, not require FDA approval at this time). These include mobile medical apps that provide tracking/organizing functions of health care data, routine clinical calculators (eg, body mass index or Apgar score), or access to reference data. The focus of the FDA’s regulatory oversight is for those mobile medical apps that “use a sensor or electrode attached to the mobile platform or tools within the mobile platform itself” for specific functions that could pose a risk to patient safety if a malfunction were to occur.6
Broadly speaking, from the user standpoint (as opposed to the regulatory classifications), there are 3 types of health care apps (Figure 1). First, there are the apps that are simply another means of accessing factual information. Gas Guide7 is a good representative of this type of application that is specific to anesthesiology; it presents topical information on variety of domains within the specialty. In a recent study, McEvoy et al8 showed that these types of apps could be coupled with decision support tools to help improve adherence to clinical practice guidelines (Figure 2). On the basis of the FDA’s definitions, these applications fall under the classification of mobile apps that are not medical devices.
Another more innovative type of app uses other hardware attachments to add functionalities to the smartphone/app combination. One recent example is Eko Core (Figure 3), which is an FDA-approved device that attaches to a stethoscope and then amplifies and records heart and lung sounds via an associated app.9 The recordings can then be uploaded to the electronic medical record or shared with other users. Another is Kardia (Figure 4), which turns the smartphone into a single-lead electrocardiogram, and the waveforms can be saved or uploaded to providers for analysis.10 Finally, GlucoWise (Figure 5) is a noninvasive glucometer combined with an iPhone application that measures, organizes, and uploads blood glucose measurements for analysis by one’s health care provider.11 It is currently in clinical trials. These devices use external hardware attached to the smartphone to allow unique data acquisition in conjunction with analysis using the smartphone and application to add new functionalities to the phone and are therefore considered mobile medical apps by the FDA. Furthermore, because the functionality of these devices could result in patient harm if malfunction were to occur, the FDA has required them to undergo regulatory oversight. One could conceive of the potential benefit to patients and potential cost savings to health systems if such applications were used in an outpatient setting to assist with shared medical decision making between a patient and a provider before an exacerbation of an underlying condition led to a costly hospital admission.
The final category of medical apps uses only the onboard technology inherent to the phone, such as the microphone, accelerometer, light, and camera, for data acquisition and the smartphone processor for data processing. On the basis of the FDA’s definitions in which functionality and intent define medical devices, a cursory look at available health care applications show many such applications that are available for download by consumers without any FDA oversight, despite the fact that the published FDA guidelines clearly state that the use of tools within the mobile platform itself or attached sensors both qualify as medical devices based on functionality and intent. In fact, the regulatory process is platform agnostic (Personal communications with Mr. Bakul Patel, Associate Director for Digital Health, FDA), so a specific search cannot be performed to look specifically for mobile medical apps that use only onboard technology that have been submitted for FDA oversight. Despite the fact that many of these apps that use only the onboard smartphone technology engage in marketing and labeling consistent with medical devices, they have somehow escaped FDA regulatory oversight. This situation is akin to how the FDA regulates prescription pharmaceuticals, but not supplements (eg, vitamins), despite the various claims made by supplement manufacturers as to their effects.
It is this final category of apps, then, that requires validation studies to confirm the proper functioning to mitigate risk to patients. The European Society of Cardiology task force on the use of smartphones for cardiovascular signal acquisition and analysis state that “in the absence of robust validation results, clinicians are hesitant to trust measurements by apps or recommend specific apps to their patients.”12 They go on to specifically mention commercially available apps that purport to measure noninvasive blood pressure in the absence of a cuff or measure pulse oximetry in the absence of an infrared light source. They call into question the means by which such applications could even derive such data and lament the lack of validation studies. To date, no studies have been published in which authors validate apps such as these. A small study under laboratory conditions showed excellent agreement for heart rate (HR) between electrocardiography and an app developed by the study investigators.13 Alternatively, other investigators looking at apps purporting to measure HR or respiratory rate showed poor performance in clinical settings.14,15
In his book, The Patient Will See You Now, Dr. Eric Topol1 begins with the fictional example of the character Elaine from the television show Seinfeld having difficulty getting an appointment with a dermatologist to have a skin rash evaluated. He creates a tantalizing vision of a world to come where a patient like Elaine could use her phone to photograph the rash and the app would diagnose and recommend treatment options to her. In such a world, though, if the app cannot accurately diagnose Elaine’s rash, then the revolution dies. In our world, there are apps available to consumers to diagnose melanoma based on photographs. One study found that such apps failed to diagnose melanoma in about 30% of cases,16 and patients anecdotally report that such apps erroneously diagnose benign lesions as malignant, understandably causing significant anxiety.17 Therefore, if we are to move from our world to such a world Dr. Topol envisions, then we must ensure the apps that we, and our patients, rely on actually provide accurate information.
If the FDA will not undertake the regulation of these devices, then validation studies must come from the medical community to prove or disprove the claimed functionalities of these apps. The European Society of Cardiology specifically mentions the apps “Instant Blood Pressure” (Figure 6) and “Pulse Oximeter” (Figure 7) because of their claims to measure noninvasive blood pressure and pulse oximetry/HR, respectively, using only the technology inherent to the smartphone (ie, without external hardware attachments such as Eko, Kardia, or GlucoWise). It is necessary to understand how would one measure noninvasive blood pressure without the use of a cuff or pulse oximetry without the use of an infrared light source. Validation studies to test such apps would then need to compare data derived from the apps with that of the standard medical devices. For example, data from “Instant Blood Pressure” should be compared with findings from sphygmomanometry or oscillometric noninvasive blood pressure monitors, and data from “Pulse Oximeter” should be compared with findings from photoplethysmography.
In fact, the authors have performed studies of these applications, and several similar ones all were found to be inaccurate.18 Furthermore, another recent study also showed that Instant Blood Pressure, in particular, was inaccurate.19 It is interesting to note that Instant Blood Pressure was removed from the app store on July 30, 2015, but was downloaded thousands of times by consumers and was in fact 1 of the top 50 downloaded apps for >150 days before its removal. Given the sheer number of apps, it is daunting to consider that each one must be validated individually before its use can be recommended to our patients, although some researchers have made efforts to make that process more manageable.20
As anesthesiologists’ traditional role as perioperative physicians expands to include the entire perioperative patient experience, it is imperative that our specialty learn to show our value to hospitals and health care systems. One of the means by which we will show value is by designing better systems of preoperative disease modification to improve outcomes while decreasing costs. The data needed to implement these changes increasingly will be generated, analyzed, and controlled by our patients and their devices. In addition, as value-based payment models become more ubiquitous, the patients who have more quantifiable data will become increasingly valuable as it allows more opportunities for disease modification to prevent acute (ie, expensive) care episodes. This growing democratization of digital health data provides a unique opportunity to partner with our patients in designing and implementing these systems. In the absence of adequate regulation by the FDA, the onus of validation falls on those of us in the health care community to use our expertise to bring some order to the anarchy of the iTunes store.
Name: John C. Alexander, MD.
Contribution: This author helped design the study and prepare the manuscript.
Conflicts of Interest: John C. Alexander declares no conflicts of interest.
Name: Girish Joshi, MBBS, MD, FFARCSI.
Contribution: This author helped design the study and prepare the manuscript.
Conflicts of Interest: Girish Joshi has received honoraria from Baxter, Cadence, and Pacira Pharmaceuticals.
This manuscript was handled by: Maxime Cannesson, MD, PhD.
3. Hyde CL, Nagle MW, Tian CIdentification of 15 genetic loci associated with risk of major depression in individuals of European descent. Nat Genet
. 2016;[epub ahead of print].
5. research2guidance. Mobile Health Market Report 2013–2017. March 4, 2013.
6. Mobile Medical Applications: Guidance for Industry and Food Drug Administration Staff. February 9 2015.
8. McEvoy MD, Hand WR, Stiegler MP, et al. A smartphone-based decision support tool improves test performance concerning application of the guidelines for managing regional anesthesia in the patient receiving antithrombotic or thrombolytic therapy. Anesthesiology. 2016;124:186198.
12. Bruining N, Caiani E, Chronaki C, Guzik P, van der Velde E; Task Force of the e-Cardiology Working. Acquisition and analysis of cardiovascular signals on smartphones: potential, pitfalls and perspectives: by the Task Force of the e-Cardiology Working Group of European Society of Cardiology. Eur J Prev Cardiol. 2014;21:413.
13. Matsumura K, Yamakoshi T. iPhysioMeter: a new approach for measuring heart rate and normalized pulse volume using only a smartphone. Behav Res Methods. 2013;45:12721278.
14. Ho CL, Fu YC, Lin MC, Chan SC, Hwang B, Jan SL. Smartphone applications (apps) for heart rate measurement in children: comparison with electrocardiography monitor. Pediatr Cardiol. 2014;35:726731.
15. Nam Y, Lee J, Chon KH. Respiratory rate estimation from the built-in cameras of smartphones and tablets. Ann Biomed Eng. 2014;42:885898.
16. Wolf JA, Moreau JF, Akilov O, et al. Diagnostic inaccuracy of smartphone applications for melanoma detection. JAMA Dermatol. 2013;149:422426.
18. Alexander JC, Minhajuddin A, Joshi GP. Comparison of smartphone application-based vital sign monitors without external hardware versus those used in clinical practice: a prospective trial. J Clin Monit Comput. 2016; [epub ahead of print].
19. Plante TB, Urrea B, MacFarlane ZT, et al. Validation of the instant blood pressure smartphone app. JAMA Intern Med. 2016;176:700702.
Copyright © 2016 International Anesthesia Research Society
20. Singh K, Drouin K, Newmark LP, et al. Developing a framework for evaluating the patient engagement, quality, and safety of mobile health applications, The Commonwealth Fund, February 2016.