Mobile health (mHealth) is defined by the Word Health Organization as “medical and public health practice supported by mobile devices, such as mobile phones, patient monitoring devices, personal digital assistance, and other wireless devices” [1,2]. mHealth is considered the future of health care [3,4], and many health care organizations have embraced mHealth as part of their patient-centered initiatives. Specifically, according to a US News & World report, 18 of the top 20 medical centers have adopted and widely celebrated mHealth technologies, at least in a review we conducted of their websites [5]. The National Institutes of Health funding for developing, testing, and implementing mHealth interventions grew from US $16.8 million in 2014 to US $39.4 million in 2018 [6].

Despite the growth and hype of mHealth technologies, there is a paucity of effectiveness evidence in the literature to support such widespread implementation [6]. Several recently published studies and consensus statements have demonstrated that there is only modest (and in many cases, low-quality) evidence that mHealth can improve patient clinical outcomes such as lengths of stay or reduction of readmissions [6-9]. There is also uncertainty as to whether mHealth can improve patient-centered outcomes such as patient engagement or patient satisfaction [6-10]. One principal challenge behind the “effectiveness” research is a lack of shared understanding about what it means to be effective in the digital space (ie, what should constitute a relevant outcome and how best to measure it) [2]. This lack of a shared understanding can likely be attributed to the multidisciplinary, multifaceted nature of mHealth, often involving disciplines such as engineering, data science, systems science, human-computer interaction, behavioral sciences, public health, and medicine [10-13]. Each discipline tends to frame its effectiveness research in terms of its own specialized knowledge [13], thereby limiting the generalizability of research results across domains.

In this viewpoint, we call for interdisciplinary, conceptual clarity on the definitions, methodologies, and patient-centered outcomes frequently used in mHealth research. There have been recent, important calls in the literature for conceptual clarity about engagement [6,10-13]; however, as we demonstrate below, engagement is only one facet of patient-centered care [14] and it is not always clear that sustained engagement is required to achieve mHealth outcomes [2]. Stated differently, a shared understanding about engagement is necessary but not sufficient for an interdisciplinary research approach. Conceptual clarity on engagement can only go so far to reduce the current fragmentation of research efforts [12]. Instead, what is needed is consideration beyond one measure of patient-centered care to include several measures of patient-centered care and the methods for evaluating them. By creating a common pathway linking definitions with outcomes and methodologies, we hope to draw a comprehensive (but not necessarily exhaustive) outline of a possible interdisciplinary approach to evaluating mHealth technologies.

To formulate our recommendations, we used a snowballing approach to identify relevant definitions, outcomes, and methodologies related to mHealth. To begin, we drew heavily upon the work of Perski et al [12], Yardley et al [6,13], and Short et al [10], among others [14,15], whose detailed frameworks enumerate the definitions and measurements of engagement [15]. We built upon their frameworks by extracting other relevant measures of patient-centered care, such as patient satisfaction, patient experience, and patient activation, drawing on the quality and patient safety literature—a literature base that has, thus far, to our knowledge, been largely untapped in the mHealth context.

Quality health care refers to care that is safe, effective, patient-centered, timely, efficient, and equitable [11,16]. In the mHealth context, most of the work comes from informaticians rather than health care quality practitioners. Theory-based informatics research must be informed by the frontline real-world, hospital environment where health care quality takes place, because patients are different than consumers or research participants [11,17]. Patients experience a constellation of complex, emotionally laden perspectives during their use of mHealth technology that may not be considered when conducting informatics research outside of the hospital setting [18]. The field of health care quality can help advance mHealth research in its evidence-based emphasis on the patient, his or her experiences, and how the continuum of care can influence outcomes that matter to the patient [19,20].

A data extraction table was used to sort, explore, and synthesize existing research (see Multimedia Appendix 1). We filtered and coded our findings based on whether a definition was proposed, whether outcome measures were discussed (and what the outcome measures were), and methodologies used to assess the outcomes. We reran the search queries that were performed in 5 recent, frequently cited systematic reviews [6,12,14,21,22] using the Medline, PsycINFO, PubMed, and Google Scholar databases. Although we did not intend to perform a systematic review for this Viewpoint, we believe that this research approach was comprehensive and could be used to formulate recommendations based on gaps and inconsistencies in the literature.

A survey of the literature highlights the myriad terms used in mHealth and the various ways in which they are defined. The terms “mHealth,” “telehealth,” “eHealth,” and “digital technologies” are often treated synonymously [23], although, in reality, they could be different in that eHealth or digital technologies can encompass devices that are not supported by mobile means, such as hospital check-in or registration portals to more advanced technologies designed to enhance patient understanding through education and communication such as Smart Boards and Smart TVs located on hospital units [24].

Digital behavior change interventions (DBCIs) are a subset of eHealth, defined as “a product or service that uses computer technology to promote behavior change,” which can be delivered through computer programs, websites, mobile phones as text message, smartphone apps, or wearable devices [6].

Patient-centered care is a health care quality indicator proposed by the Institute of Medicine (IOM) [25]. Patient-centered care has generally been poorly defined, with authors often conflating several distinct concepts such as using the term “patient-centered care” when they are usually referring to a specific outcome measure of patient-centered care [26]. As the IOM explains, patient-centered care is care that is “respectful of and responsive to individual patient preferences, needs, and values, and ensuring that patient values guide all clinical decisions” [25]. Through this lens, patient-centered care can be conceptualized as an overarching, broad concept that puts emphasis on respecting patients as a means of honoring their dignity and worth [26,27]. By extension, patient-centered mHealth technologies can be defined as technologies supported by mobile devices designed to promote patient-centered care. There are several outcome measures for patient-centered care.

“Patient experience” is likely the strongest, most proximate outcome measure of patient-centered care. Patient experience refers to any process observable by patients, including subjective experiences (eg, pain control) and their objective experiences (eg, wait time) [28,29]. A primary feature of patient experience is that it reflects actual health care experiences [30]. Another important feature of patient experience is that it refers to a time sequence. Specifically, patient experience can refer to the first touchpoint within an episode of care (eg, being assessed for a left knee replacement) to the last touchpoint within that episode (eg, last follow-up appointment following surgery for a left knee replacement). Alternatively, patient experience can refer to the whole continuum of care—their first encounter with Hospital A to their last encounter with Hospital A.

“Patient engagement” and “patient activation” are two outcome measures of patient-centered care that are distinct from patient experience. In particular, “patient engagement” has been defined in various ways in the literature [14,30,31]. Recent proposals for integrative definitions of engagement have defined engagement as consisting of objective and subjective components [6,12,13]. The objective component is the extent (eg, amount, frequency, duration, and depth) of usage of the mHealth technology [12]. With respect to what constitutes sufficient objective engagement, the literature demonstrates a palpable lack of consensus [15,31]. Some researchers have proposed that a certain empirical threshold of engagement must be met to show sufficient engagement with the intervention to achieve intended outcomes [13], whereas others suggest that one critical point of engagement or fluid, ebbing, and flowing engagement may be sufficient [32,33]. The subjective element of engagement is often characterized by the user’s attention, interest, and affect—their overall experience in engaging with the specific mHealth technology [10,12].

“Patient activation” is a patient-centered care outcome measure that refers to patients’ “willingness and ability to take independent actions to manage their health,” such as avoiding health-damaging behaviors and adopting healthy lifestyle choices, including exercising regularly, eating well, or monitoring their glucose levels [34]. DBCIs usually attempt to evaluate patient activation [34-36], although they may indicate that they are designed to assess adherence or engagement. “Adherence” is often confused with “engagement” and has been defined in at least three separate ways: adherence could refer to whether the intervention is used as intended by the developers [10,32], the usage of mHealth [37] (which is more accurately called engagement), or the patient’s willingness and ability to adhere to the recommendations provided by their physician or health care provider (which is likely the most common use of the word “adherence” in medico-legal parlance) [35,38-40].

Taken together, patient engagement typically refers to patients’ engagement with an intervention itself, whereas patient activation refers to patients’ physical and mental health-related activities based on what they learned from an intervention [34,41]. Patients can be engaged by, for example, reading digital messages on the importance of exercising and still not be activated to start exercising [34].

Patient engagement and activation are distinct outcome measures from patient experience in that patient experience refers to patients’ perceptions of others’ actions, whereas patient engagement and activation refer to patients’ actions [27,42]. Patient experience, engagement, and activation are all measures of patient-centered care in that they put emphasis on the patient playing an integral role in their outcomes [43,44], with the ultimate decision-making authority resting with patients. Further, all three concepts can be evaluated empirically.

Finally, “patient satisfaction” is the most attenuated outcome measure of patient-centered care. Patient satisfaction is a term that is often used in health care quality parlance and yet is frequently misunderstood [27] owing to one key feature of satisfaction that is often ignored: satisfaction has little to do with quality [22,27,29]. Patient satisfaction only refers to whether patients’ expectations were met [22,27]. Patients can be satisfied with care that is low quality and yet be dissatisfied with high-quality care. For this reason, it is incorrect to say that the Hospital Consumer Assessment of Healthcare Provider and Systems survey, the first standardized, publicly reported survey of patient perspectives [45], measures patient satisfaction when, in actuality, the survey uses patient experience as its primary measure for patient-centered care [27,28,46].

As the above discussion illustrates, to date, we have not achieved a shared understanding of important mHealth constructs, or how to conceptualize and operationalize them [10]. Thus, if the mHealth community is to continue to promote the use of patient-centered mHealth technologies (ie, mHealth technologies that are designed with the goal of promoting patient-centered care, as measured by patient experience, patient engagement, patient activation, or patient satisfaction), then we need precision in our terminology to extrapolate generalizable, transferable results [6,31].

To validly measure a concept, there must be a tight linkage between the patient-centered care construct (ie, the outcome measure) and items developed for measurement [47]. Below, we describe several methodologies that can assess each patient-centered outcome.

Above, we have described several definitional inconsistencies for key constructs in the mHealth literature. In an effort to achieve clarity, we teased apart several patient-centered care outcomes, and we outline methodologies appropriate to measure each of the patient-care outcomes (Table 1). In what follows, we provide recommendations for evaluating mHealth patient-centered technologies. Our recommendations relate to the patient-centered constructs in that we advocate for patients taking center stage in mHealth research and collaboration efforts to enhance patient experience, patient engagement, patient activation, and patient satisfaction.

In this Viewpoint, we called for interdisciplinary, conceptual clarity on the definitions, methodologies, and patient-centered outcomes frequently used in mHealth research. In doing so, we advocate for consideration of several measures of patient-centered care, and we outline various methods for evaluating them. By creating a common pathway linking definitions with outcomes and methodologies, we provide a possible interdisciplinary approach to evaluating mHealth technologies.

To that end of creating an interdisciplinary approach, we also provide several recommendations that we believe can advance mHealth research and implementation. For instance, if an international collaboration were created to develop evaluative criteria, using the guidance provided here to ground criteria development, then low-quality digital technologies would likely be excluded. Transparency and precision would be promoted, large-scale published evidence would be encouraged [3], and mHealth technologies could finally flourish within a high-quality, patient-centered landscape.