Internet Interventions 2 (2015) 91–101

Contents lists available at ScienceDirect

Internet Interventions

j ourna l homepage: www. invent - journa l .com/

From black box to toolbox: Outlining device functionality, engagementactivities, and the pervasive information architecture ofmHealth interventions

Brian G. Danaher a, Håvar Brendryen b, John R. Seeley a, Milagra S. Tyler a, Tim Woolley c

a Oregon Research Institute, Eugene, OR, USAb Norwegian Centre for Addiction Research, University of Oslo, Oslo, Norwayc IEQ Technology, Springfield, OR, USA

http://dx.doi.org/10.1016/j.invent.2015.01.0022214-7829/© 2015 The Authors. Published by Elsevier B.V

a b s t r a c t

a r t i c l e i n f o

Article history:Received 29 November 2014Received in revised form 22 January 2015Accepted 28 January 2015Available online 7 February 2015

Keywords:mHealtheHealthToolboxBlackboxInternet interventionsPervasive information architecture

mHealth interventions that deliver content via mobile phones represent a burgeoning area of health behaviorchange. The current paper examines two themes that can inform the underlying design of mHealth interventions:(1) mobile device functionality, which represents the technological toolbox available to intervention developers;and (2) the pervasive information architecture of mHealth interventions, which determines how interventioncontent can be delivered concurrently using mobile phones, personal computers, and other devices. We positthat developers of mHealth interventions will be able to better achieve the promise of this burgeoning arena byleveraging the toolbox and functionality ofmobile devices in order to engageparticipants and encouragemeaningfulbehavior change within the context of a carefully designed pervasive information architecture.

© 2015 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license(http://creativecommons.org/licenses/by-nc-nd/4.0/).

Contents

1. Background & aims . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 922. Defining the domain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92

2.1. Taxonomy for defining smartphones . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 922.2. Taxonomy for defining mHealth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92

3. mHealth toolbox . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 933.1. Short message service (SMS) text messaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 933.2. IVR automated calls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 943.3. Smartphone apps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 943.4. Email . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 943.5. Program content display using onscreen text . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 943.6. Text notifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 943.7. Audio and video notifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 953.8. Recording pictures, audio, and video . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 953.9. Sensor functionality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 953.10. Discussion forums and blogs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 953.11. App-specific design and engagement activities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96

4. Pervasive information architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 974.1. Information architecture of interventions delivered on personal computers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 974.2. Combining mHealth with interventions delivered on personal computers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 974.3. Pervasive information architecture of mHealth interventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98

5. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99

. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

92 B.G. Danaher et al. / Internet Interventions 2 (2015) 91–101

1. Background & aims

eHealth interventions have been shown to be effective in encouraginga broad range of health behavior change (e.g., Myung et al., 2009;Wantland et al., 2004) including, for example, interventions for smokingcessation (Civljak et al., 2013; Graham et al., 2011; Strecher, 2007),curbing alcohol consumption (Riper et al., 2011), andmanaging depres-sion (Griffiths et al., 2010; Spek et al., 2007; Titov, 2011). The promise ofeHealth interventions is not limited to Internet interventions deliveredon personal computers because it also applies to mHealth interventionsdelivered on mobile devices (Whittaker et al., 2009). Little is known,however, about, what distinguishes effective from less effective inter-ventions (Webb et al., 2010).

The burgeoning field of eHealth interventions has focused more onoutcomes than on underlying factors and mechanisms — a Black Boxapproach (Brendryen et al., 2010; Strecher, 2008). Researchers haveproposed several possible remedies to shed more light into the BlackBox, including the use of more detailed, standardized reporting ofbehavior change strategies (Abraham and Michie, 2008), providingcomprehensive reporting of the complete intervention rationale alongwith a description of specific techniques (Bartholomew et al., 2011;Brendryen et al., 2013; Schaalma and Kok, 2009), and testing new theo-ries of health behavior change (Riley et al., 2011).

There is a growing evidence of the efficacy of mHealth programs toencourage a wide variety of behavior changes, but considerably lessresearch to help inform the intervention designer in choosing the tech-nological tool(s) and devices that will engage participants and help themachieve their desired therapeutic outcomes. It is premature to try to syn-thesize findings regarding the optimal designs and benefits of mHealthinterventions because the field is so new, and the interventions arebeing used to address so many diverse behaviors/disorders over diversepopulations. Instead, in this paper we hope to informmHealth interven-tion development by shedding light into the mHealth Black Box byoutlining: (1) mobile device functionality — the technological toolboxavailable to intervention developers; and (2) the pervasive informationarchitecture ofmHealth interventions— theway that an integrated inter-vention can be delivered concurrently using mobile phones, personalcomputers, and other devices.

2. Defining the domain

mHealth interventions include health behavior change interventionsthat are ostensibly deliveredusing “…computer devices that are intendedto be always on and carried on the person throughout the day” (Rileyet al., 2011). mHealth interventions are intended to “…travel throughtime and spacewith the participant [whereas] the traditional desktop ac-cessmethod implies [that] participants are tethered to a particular deviceand are therefore more sedentary (p. 314)” (Turner-McGrievy and Tate,2014). We also agree with the distinction recommended by Riley et al.(2011) to exclude iPads and other tablets from primary considerationin this paper because they are not typically carried by person throughoutthe day. Finally, our paper was informed by our adaptation of Ritterbandand Thorndike's (2006) distinction between internet interventions andpatient information websites in order to distinguish mHealth inter-ventions from myriad mHealth programs: mHealth interventionsare “typically behaviorally or cognitive-behaviorally-based treatmentsthat have been operationalized and transformed for delivery via”mobiledevices. We also exclude using mobile devices for ecologically momen-tary assessments except when they are used to inform behavioralinterventions.

mHealth interventions have emerged in large part in response to thenearly ubiquitous use of mobile phones: more than 90% of Americansare mobile phone users (Fox and Raine, 2014), with few differences intheir gender and race/ethnicity (Lenhart et al., 2010). Trend data indi-cate that by 2018 almost all Americans will be using smartphones(Smith, 2013). Worldwide use is also very large and rapidly growing,

with 2014 estimates of 4.55 billion mobile phone users and 1.75 billionsmartphone users (eMarketer, 2014).

mHealth interventions can leverage the fact thatmobile phone userstypically carry their phones with them throughout the day and evenkeep them nearbywhen asleep, making it possible to deliver helpful be-havior change content to – andeven have interactionswith – individualsas they go about their normal everyday lives (Heron and Smyth, 2010;Lenhart, 2010; Patel et al., 2006). mHealth intervention componentscan be proactive in that they reach out to users to deliver content,prompt interchange, and candeliver persuasive content that encouragesbehavior change (Fogg, 2007). Heron and Smyth (2010) have describedthese as ecologically momentary interventions that occur at specificallyidentified moments in everyday life providing real-time support in thereal world.

mHealth interventions can be designed to provide just in timesupport and guidance when most needed (Beale, 2009; USDHHS, 2014;Wikipedia, 2014e). There are at least two ways that mHealth interven-tions are just-in-time. First, intervention content can change based ondata obtained during the course of the intervention, as in delivery oftextmessages that are relevant to a participant's recent success/problemsin managing eating (Intille et al., 2003) or quitting tobacco (Riley et al.,2011; Ybarra et al., 2012). A technological elaboration of this point canbe found in the just-in-time-interventions described by Kumar and hiscolleagues (Kumar, 2012; Sarker et al., 2014) in which wearable wirelesssensors can inform intervention content to enhance successful behaviorchange (e.g., quitting smoking). The second just-in-time aspect ofmHealth interventions involves their immediate accessibility. Becausemobile phones are literally within reach they can act as an “as-needed”and available resource, as when coping with a difficult smoking urgethe participant could immediately review – and obtain benefit from –

helpful content on the smartphone, which might include a personal listof reasons to quit (Ybarra et al., 2012) and/or a relaxation audio(Whittaker et al., 2008).

2.1. Taxonomy for defining smartphones

In contrast to current smartphones, early mobile phones did not havea touchscreen, a QWERTY keypad, or the benefits of an advanced operat-ing system. These phones have been described variously as mobilephones having standard features, feature phones, and/or basic phones(iHeed Institute, 2011), conventional (The Nielsen Company, 2013;Wikipedia, 2014c) or common (WHO, 2011). These older mobile phoneshave even been referred to as “dumb phones” to clearly distinguishthem from the current generation of smartphones (Wikipedia, 2014c).However, while the label “smartphone” is driven by marketing consider-ations, an important caution needs to be acknowledged because the“smartness” of today's phones inevitably will appear much less “smart”when they are compared to the next generation mobile devices. Sincesmartphones offer so much more functionality than merely makingphone calls, the label “mobile device” better captures the breadth oftheir toolset and the fact that people are able to use them as “convergeddevices that combine mobility, connectivity, and programmability”(Yuan, 2005).

2.2. Taxonomy for defining mHealth

The World Health Organization describes mHealth as a componentof the broader category of eHealth (WHO, 2011). A casual Google searchusing the term will quickly reveal that the mHealth label has beenapplied to a very considerable breadth of programs and initiatives,including using mobile computing and communications technologiesto facilitate care of medical patients (Kotz, 2011) and the use of mobilephones within developing countries to support health workers, collectpublic health data, and enable health information messaging andhelpline services (iHeed Institute, 2011). A taxonomy for mHealth isstill emerging and some have questioned whether it will endure as a

93B.G. Danaher et al. / Internet Interventions 2 (2015) 91–101

separate domain because it conceptually and empirically overlaps soconsiderably with telemedicine, telehealth, and eHealth (Bashshuret al., 2011). In this paper we have appended the term “intervention”to the mHealth label (yielding “mHealth intervention”) to highlightthose programs that are of particular interest to researchers and pro-gram implementers of Internet Interventions.

3. mHealth toolbox

The specific features available in mHealth interventions depend onthe operating system of the mobile device and type of app (if any)being used. Two first generationmHealth interventions – textmessagingand Interactive Voice Response (IVR) calls – are able to use the basicfunctionality found on early mobile phones. The toolbox of possiblemHealth intervention features available when designed for smartphoneusers still includes texting and IVR but it is vastly more varied andpowerful. For example, smartphone apps can access the Internet,which enables them to access Web-based content, use GPS to tracklocation and provide trip guidance using online maps, and playaudio and/or video content. And, rather than a loosely connectedcollection of tools that reach out to participants, smartphone apps canbe designed to offer participants a cohesive multifaceted program touse.

3.1. Short message service (SMS) text messaging

SMS messaging involves the delivery of brief text messages that areshared between/among mobile phones. Text messaging can reach allmobile phones irrespective of service provider (Aguilera and Munoz,2011) and is the most common non-voice use of mobile phones(Smith, 2011), with more than 153.3 billion text messages sent eachmonth in the U.S. (CTIA, 2014). Exchanging text messages can incurcharges from the user's cellular plan, although this cost varies by plan,and unlimited texting is becoming a more commonplace bundledoption. It is possible to avoid per-message surcharges altogether byusing the device's proprietary text message functionality availablewhen sender and receiver(s) all use the same mobile device brand

Fig. 1. SMS text messa

(e.g., the iPhone's iMessage capability). A schematic depiction of theunderlying technology required for mHealth interventions to delivertext messages is shown in Fig. 1.

Text messages use very brief messages (limited to 160 characters)displayed as an unthreaded top-downmanner that are grouped accord-ing to the sender phone number. By default, mobile phones typicallynotify users about the arrival of text messages with an audible alert, amessage displayed in the forefront of the screen, and possibly a numericbadge on the text messages app onscreen icon. As a result, textmessages can push program content to participants in a way that canbe relatively difficult to ignore. mHealth programs can be programmedto use text messages as the vehicle for sending users small chunks ofprogramcontent aswell as brief reminders that have a prompting effect(Fry andNeff, 2009). Pre-programmedmessages can be scheduled to bedelivered at predefined times during the day and in different amounts(numbers of messages) over the course of an intervention. The processcan be unidirectional (message arrive with program content) as well asbi-directional (text messages ask questions of users whose simple textmessage replies can be used by the program to tailor the delivery ofprogram content). There is a growing research track record showingbeneficial effects from using automated text messaging for healthbehavior interventions (Fjeldsoe et al., 2009, 2010; Patrick et al.,2008). Meta-analyses (Fjeldsoe et al., 2009, 2010; Krishna et al.,2009) have documented text message interventions that have beenused for diabetes self-management (Franklin et al., 2006; Yoon andKim, 2008), eating disorders (Robinson et al., 2006; Shapiro et al.,2010), physical activity (Prestwich et al., 2009), and tobacco cessation(Brendryen et al., 2008; Brendryen and Kraft, 2008; Free et al., 2009;Rodgers et al., 2005).

The therapeutic benefits associated with text messages will likelyvary based upon their content and tone, their bi-directional sharing ofcontent, and the schedule and density of their delivery. For example, de-livering too many messages may well be intrusive, annoying, and thusunhelpful. But length limitations can cause text messages to read likefortune cookie messages that have extraordinarily limited opportunityfor empathy, nuance, and engagement. All of these features representimportant areas for research inquiry.

ge infrastructure.

94 B.G. Danaher et al. / Internet Interventions 2 (2015) 91–101

3.2. IVR automated calls

IVR programs delivered onmobile phones essentially involve sendinga recorded phonemessage in an automated calls. Because of this simplic-ity, these IVR programs function quite well on standard feature mobilephones, and, of course, also on land-line phones. Similar to text messag-ing, calls programmed for delivery by the IVR system arrivewith an audi-ble alert (an incomingphone call ringtone). Users are then able to listen toprogram content (unidirectional audio content delivery). Some IVRsystems enable users to reply to questions presented in a call they havereceived (bi-directional content). Many mobile phone users may find itdifficult to use their phone's keypad for data entry while they remainengaged in the phone call. More sophisticated IVR systems are able tointerpret simple verbal responses thereby avoiding this usability barrier.

Just as with text messages, IVR calls proactively push content toparticipants. Many health behavior change interventions and medicalsupport programs have successfully used IVR systems (Abu-Hasaballahet al., 2007; Bartholomew et al., 2011), including programs for physicalactivity (Pinto et al., 2002), healthy eating (Delichatsios et al., 2001;Estabrooks et al., 2009; Estabrooks and Smith-Ray, 2008; Piette, 2000,2002), medication refills (Reidel et al., 2008), caregiver support(Mahoney et al., 2003), depressive symptoms (Osgood-Hynes et al.,1998), delivery of ambulatory care (Oake et al., 2009), diabetes self-management (Piette, 2000), and smoking cessation (Ramelson et al.,1999; Reid et al., 2007). When IVR calls are programmed for receipton amobile phone they become part of the possible toolset for mHealthinterventions (Brendryen et al., 2008; Brendryen and Kraft, 2008).

3.3. Smartphone apps

While there is increasing discussion regarding the taxonomy of theterm apps vs. applications (Lewis et al., 2014) and app classificationschemes (Wang et al., 2014), we consider three general types ofsmartphone applications: native (operating-system-based) apps, Webapps, and hybrid apps that combine features found in both native andWeb apps (Nielsen Norman Group, 2014).

Native apps use the sophisticated features and functionality madeavailable through the mobile phone's operating system (e.g., iOS foriPhones as well as Android). For example, they can use GPS-derivedlocation, the system calendar, system alarms, and other notifications.Some native apps can function effectively without persistent or live In-ternet access. Because native apps use data available through themobilephone's operating system, they generally must adhere to various designand review requirements of the company overseeing the operatingsystem (Apple, 2013; Google, 2014), and be downloaded via app storeshosted by the smartphone's manufacturer. One exception to this ruleinvolves “enterprise deployment” which permits the provisioning ofin-house apps by large organizations directly to end-users whilebypassing the app store altogether (Apple, 2014). For example, enterprisedeployment can be used to deliver custom apps by corporations to theiremployees, health care organizations to their patients, and even researchorganizations to distribute mHealth interventions to a designated audi-ence of users according to custom rules (e.g., treatment allocation,research assignment), something not as easily accomplished using appstores.

Mobile browser orWeb apps are essentially websites that are deliveredusing the smartphone's browser. The selection of content and themannerinwhich it is displayed are controlled by the logic contained in a programhosted on a remote server (server-side). As with desktop-based Webinterventions, mHealth interventions using mobile Web apps are able toincorporate sophisticated levels of interactivity, tailoring, and engage-ment tracking. Compared to a native app, mobile browser apps don't re-quire review and distribution by a smartphone manufacturer. Becausethey do not require different programming in order to fit the functionalityof different operating systems,Web apps can be easier and less costlyto develop than native apps when mHealth interventions are to be

implemented on multiple smartphone devices. Because their con-tent and program rules are controlled by a remote server, mobilebrowser apps require persistent access to the Internet. Moreover,mobile browser apps tend to be somewhat slower and less responsivethan native apps.

Hybrid apps incorporate features and functionality found in nativeapps with the versatility and efficiency associated with using mobilebrowser apps. They are able to display program content using a browserthat is embedded within the native app itself rather than simply usingthe smartphone's browser. As a result, hybrid apps can offer a moretightly integrated environment (envelope) than mobile browser apps.They can providemultiple tools some ofwhich drawupon the intelligenceand data that are only available from other native (built-in) smartphonefeatures.

The choice about what type of app is best for mHealth interventionsdepends upon analysis of available programming/development re-sources, the need to access data from the native operating system, theimportance of delivering a very tightly integrated intervention, and easeof distribution. Interventions that use websites – whether deliveredwithin Web apps or hybrid apps – tend to be reactive; they wait forusers to visit (Brendryen et al., 2010; Heron and Smyth, 2010; Rileyet al., 2011). This passivity can be balancedwith othermHealth interven-tion components (e.g., text messages) that push program content toparticipants in salient ways that the user is more likely to notice and use.

3.4. Email

Email fits within the broader category of mobile phone text notifica-tion tools (Mohr et al., 2014). Similar to text messaging and IVR calls,email proactively pushes content to mHealth intervention participants.By default, the arrival of email is far less salient because it may not havean audible or visible signal. In these circumstances, email may requireparticipants to seek it out at their own initiative. It is possible to increasesalience by asking programparticipants to change their default settings inorder to assign audible sounds or visible alerts that announce its arrival.Automated email has been found to be a helpful feature that is frequentlyincluded in eHealth interventions to increase participant engagementand program efficacy across a wide variety of problems (e.g., Civljaket al., 2013).

3.5. Program content display using onscreen text

mHealth smartphone apps display some portion of their content asparagraphs of onscreen text. Browser-based mHealth interventionscan be designed to automatically rearrange and reduce the size of pro-gram text for different screen sizes (personal computer, mobile phone,large-scale mobile phone or phablet, and tablet) and display orientation(portrait vs. landscape). These responsive (Wikipedia, 2014g) or adap-tive (Wikipedia, 2014a) scripting approaches enable Web applicationsto automatically identify the user's device or the attributes of the user'sdevice (e.g., browser viewport size and screen resolution). This informa-tion is then used to tailor the types of content displayed as well as themanner in which it is displayed— thereby resulting in a distinct user in-terface for each device (Nielsen and Budiu, 2013). Although conceptual-ly attractive, this one-size-fits-all browser-based approach can be quitecomplicated to program, and many times underlying program logicneeds to be redesigned and not just the display of content. The deliveryand display of mHealth intervention content will inevitably require thedesigner to create rules to define which content to display as well ashow it should be formatted.

3.6. Text notifications

mHealth apps can also proactively push intervention content toparticipants by displaying salient text notificationswith alerts;mHealthintervention apps can use a broader set of tools to notify participants in

95B.G. Danaher et al. / Internet Interventions 2 (2015) 91–101

variousways. For example, they can display text onscreen which occurswith a corresponding audible alert sound (see Fig. 2). Rather than beingconstrained to unthreaded top-down “conversation” grouped withinthe sender's phone number, custom text notifications can be displayedand catalogued in ways that are easier for the user to find and review—

thus potentially increasing their impact. Moreover, when tapped,custom text notifications can enable users to access specific contentavailable within a native app — a degree of integration that cannot beaccomplished using text messages.

3.7. Audio and video notifications

Smartphone-basedmHealth interventions can be designed to includeaudio content from “simulated places and people” (Mohr et al., 2013), av-atars or even mobile health counseling agents (Bickmore et al., 2009).Asking participants to have human-like audio interchanges with theirmobile phones seems increasingly acceptable – and possibly feasible –

based upon the rapid widespread use of intelligent personal assistants(conversational agents) available on current smartphones includingSiri (iOS) (Wikipedia, 2014h), Cortana (Microsoft) (Warren, 2014),GoogleNow (on multiple smartphone OS) (Wikipedia, 2014d), as well as assis-tants from software companies (e.g., Nina from Nuance (Nuance,2014)). Users are increasingly using voice commands to interact withfeatures of their smartphones using these digital personal assistants,and in some cases, to be guided by verbal and/or text advice from theirsmartphone's digital assistant.

Similarly, mHealth intervention apps can deliver video content toparticipants. Examples include the use of video testimonials (Whittakeret al., 2008, 2012), and animated presentations.

3.8. Recording pictures, audio, and video

Designers of mHealth interventions can draw upon the smartphone'sbuilt-in tools to provide key data. For example, participants in a“photovoice” intervention (PhotoVoice, 2014; Wikipedia, 2014f) usedtheir smartphone camera to accomplish an assigned task of takingpictures of surroundings and experiences that reflected their weight-related concerns (Woolford et al., 2012). OnemHealth interventionmea-sured participantmeal portion sizes using images captured by participantsmartphone camera (Six et al., 2010). Other studies have asked partici-pants in an Internet cessation study to use their computer's video capa-bility to show a reading from a Carbon Monoxide meter as a way tovalidate their self-report — an approach that could be easily adaptedfor a smartphone intervention (Dallery et al., 2007, 2013) (See Fig. 3).Similarly, the smartphone's microphone can be used to record conver-sations with coaches/counselors for subsequent review (Luxton et al.,2011). Finally, the tempo of smartphone-delivered music has beentailored to encourage participants to increase their physical activity (Liuet al., 2008).

Fig. 2. Example of text notification feature of the National Cancer Institute's QuitPaliPhone-based smoking cessation intervention app (NCI, 2014).

3.9. Sensor functionality

The use of sensors in mHealth interventions is in its infancy. Thereare a growing number of published descriptions and initial evaluationsof innovative sensor-enhanced interventions already appearing(e.g., Clifton et al., 2013; Cowie et al., 2013; Kumar et al., 2013; Mooreet al., 2007; Scully et al., 2012; Warmerdam et al., 2012) — although anumber of the more sophisticated current generation sensors tend tobe bulky contraptions attached to the user's body rather than builtinto a smartphone. Nonetheless, sensors offer the promise of beingable to provide the unobtrusive capture of personally relevant informa-tion that may be able to detect when a particular intervention strategymight be most helpful (Mohr et al., 2013). In the U.S., Kumar and hiscolleagues (Kumar, 2012; Sarker et al., 2014) and his NIH-funded center(MD2K, 2014) are in the early phases of developing just-in-time-interventions that use behavioral and situational data derived fromwearable wireless sensors to inform intervention content as a way tosignificantly enhance behavior change (e.g., quitting smoking). A parallelinitiative involving European researchers (Guiry et al., 2014;Warmerdamet al., 2012) has examined the use of attached sensors within the contextof an eHealth depression intervention that also included a smartphoneadjunct (ict4depression, 2014).

Mobile phone sensors can capture data using devices external to themobile phone such as a blood pressure monitor, wearable devices likesmartwatches and wristbands, as well as Internet-based data availableto mHealth participants like relevant weather and traffic. mHealth inter-ventions could consider using data from sensors embedded in themobilephone (e.g., data on the calendar and time of day from the calendar, loca-tion/proximity from GPS, movement via the accelerometer, step counter,sound from the microphone (Kumar, 2014; Mohr et al., 2014)). For ex-ample, a recent mHealth intervention (Addiction-Comprehensive HealthEnhancement Support System or A-CHESS) that used the smartphoneGPS system to detect when a participant was within a certain distanceof a high-risk location (e.g., a bar visited in the past) in order to send atext message alert (Gustafson et al., 2014). Other examples using GPSto assist alcohol treatment also have emerged (Dulin et al., 2014). Thedistinction regarding the source of sensor data (i.e., data derived fromsensors within the mobile phone vs. from sources external to the mobilephone) blurs as device-based sensors feed as well as read from externaldevices, including cloud-based data sources and wearable devicesand possibly even devices placed strategically in our environment(geofencing). Although the accuracy, usability, and battery consump-tion of current sensors need to be improved, the use of sensors bymobile devices to trackmyriad user data can be expected to dramaticallyimprove over time — and increase in value to mHealth interventions.

3.10. Discussion forums and blogs

mHealth applications sometimes use discussion forums (web blogsor forums) designed to enable program participants to interact witheach other, sharing their stories and support (Gustafson et al., 2014).This feature can be particularly helpful if sources of support are difficultto find and if anonymity and privacy are key barriers to otherwise seek-ing support (as in cases that involve considerable social stigma). Aspeer-to-peer social media tools like Facebook and Twitter have becomeubiquitous, itmay be superfluous formHealth interventions to try to de-sign their own tools that attempt to mimic this functionality. However,there may still be reasons to wrap a discussion forum within the enve-lope of anmHealth app, aswhen there are extreme concerns about partic-ipant privacy and/or when associated with higher risk situations(e.g., suicidal behavior) that might well require the ongoing review offorumactivities by a trainedmoderator.Moreover, it is difficult to careful-lymeasure the extent towhich participants use Facebook and Twitter be-cause unobtrusive measures of participant engagement are typicallylimited to use of features directly provided by the mHealth app.

Fig. 3. Example of how video of a carbon monoxide meter can be used to confirm self-reported smoking abstinence (used with permission from R. Dallery, 2014).

96 B.G. Danaher et al. / Internet Interventions 2 (2015) 91–101

3.11. App-specific design and engagement activities

The manner in which themHealth intervention app developer takesadvantage of the mobile device's functionality can encourage participantengagement. At a relatively simple level, this can include following inter-face guidelines unique to the smartphone so that the user can build onstandard appearance and functionality of what is presented on-screen(Apple, 2013; Google, 2014). Examples include ensuring that apps donot include an exit button, that they start back up where the user leftoff, and they respond to changes in device orientation (landscape or

Table 1Engagement activities for consideration in mHealth intervention apps.

Activity Function Ex

Lists Add personal content using lists or typing in own content Tosit

Expand-collapsecontent

Explore additional detail on topics of interest. To

Wizards tool Multi-step interaction that builds towards a strategy Toav

Practice changeactivities

Homework tasks to be accomplished in normal routine To

Behaviortracking

Capture and display participant data over time designed toencourage self-monitoring, show patterns and progress

To

Videos To provide content and encourage use of recommended strategies Towa

Animatedtutorials

Explain underlying models for change Usbe

portrait). In addition, mHealth intervention app activities will probablyembody variations of the same participant engagement activities usedin eHealth interventions designed for personal computer-based interven-tions — see Table 1 adapted from Danaher et al. (2013). A sample of theLists activity is illustrated in Fig. 4.

Another just-emerging category of activities not included in thetable but nonetheless deserving consideration involves serious gamesthat are designed to use smartphone tools to be engaging while encour-aging “players” to change their behavior. For example, a commercialAndroid app named “Dance! Don't Fall” asks users to wear their

amples

list pleasant activities, supporters, reasons for wanting to change, high-tensionuations, warning signsexplore FAQs, Myths & Facts, etc.

encourage goal selection or identify lessons learned (e.g., a lapse/relapse in order tooid slips in the future)track use of relaxation methods to manage stress, or to anticipate and savor activities

track and chart smoking status, mood ratings, pleasant activities

deliver content from program host, testimonials from others describing experiences,ys to overcome barriers, revise strategies, plan for the futuree animation to show downward spirals for mood and urges as well as how they caninterrupted at critical choice points

Fig. 4.Draft screenshots (my health reasons for quitting) illustrating an interactive Lists activity (see Table 1) excerpted from ourmHealth smoking cessation browser app. Tapping field inleft screen triggers popup to appear (center screen), which enables user to choose from fixed list or type in text, which then causes popup to close and reveals personally-relevant healthreason for quitting (right screen).

97B.G. Danaher et al. / Internet Interventions 2 (2015) 91–101

smartphone on their lower back to enable the device's accelerometer totrack dance steps designed to prevent falls and promote exercise athome (Kerwin et al., 2012; Silva et al., 2013). Another Android gameapp called “SmartCAT” is a smartphone app for childhood anxiety treat-ment designed to work with a therapist portal (Pramana et al., 2014;University of Pittsburgh, 2013).

4. Pervasive information architecture

4.1. Information architecture of interventions delivered on personalcomputers

In our original discussion of information architecture and healthbehavior change websites (Danaher et al., 2005) we focused on thedesign of individual websites that we assumed would be available ondesktop computers. In that report we delineated a number of informa-tion architecture designs that differed in the way that users were ableto view content:

• a tunnel design guides users through a step-by-step process to allowprogram webpages to be accessed in a particular order to improvethe chances of achieving a goal that is measurable and consistent.Upon entering a tunnel the user accepts a lowered degree of autonomy,as if themHealth programbecomes a demanding coachwho pushes theparticipant to realize his/her success (Fogg, 2003).

• a matrix design enables users to explore available content without anyprogrammatic constraints.

• a hierarchical design organizes information in a top-down manner sothat users can choose to drill down in order to access increasinglydetailed content.

• a hybrid design combines elements of the tunnel, matrix, and hierar-chical designs to enable users to access program content according

to workflow rules (Mohr et al., 2014) that govern, for example, theamount, order (timing), and detail of program content.

4.2. Combining mHealth with interventions delivered on personal computers

At a conceptual level, the four information architecture designscould simply be extrapolated to fit mHealth interventions. But theparadigm of an intervention delivered exclusively on a personalcomputer fails to adequately capture the push and pull features ofmHealth interventions. For example, the emerging mHealth paradigmis shaped by what usability experts have described as mobile phones'“impoverished user experience”with “…tiny screens, slow connectivity,higher interaction cost (especially when typing, but also due to users'inability to double-click or hover), and less precision in pointing due tothe “fat finger” problem” (p. 34; Nielsen and Budiu, 2013). Constrainedscreen size and interactivity point to the need to design mHealthinterventions that have less complicated content and require lessinteraction from users. One way this might be accomplished is to in-telligently ration howmuch and what kind of content such interven-tions should include. For example, it might be best for them to focuson content that has the greatest potential impact in terms of chang-ing certain behaviors (e.g., motivational messages, behavioralprompting, self-monitoring and charting, and/or other proactiveaction-focused program content) using the toolset features and func-tionality described in this report (e.g., text messages, IVR call prompts,notifications).

Secondly, simply porting a personal computer design to thesmartphone would potentially ignore the push features available whenusing text messaging, IVR calls, or text notification in mHealth interven-tions. These features can encourage greater participant use of the inter-vention, and provide a sense of program vitality and responsiveness.

98 B.G. Danaher et al. / Internet Interventions 2 (2015) 91–101

Third, hybrid eHealth interventions combine mHealth programcomponents with Web-based interventions accessed on a personalcomputer. In these scenarios, mobile phones could push the deliveryof just-in-time content designed to promote interaction, increase moti-vation, challenge dysfunctional beliefs, and provide cues to action(Klasnja and Pratt, 2012; Webb et al., 2010). With its greater screenspace and a more usable interface, the personal computer componentcould provide rich content with enhanced interactivity and multimediafeatures, charting, long-term access to resource descriptions, etc. Forexample, a recent mHealth depression intervention was deliveredusing this type of hybrid approach: “…monthly text messages directedparticipants to a mobile website. This website (available only to recog-nized participant mobile phone numbers) provided a summary of thekey messages, information on how to get more help, and a download-able relaxation audio. New videos were posted on thewebsite monthly.Ringtones, wallpaper images, and music downloads were linked to themobile website” (Whittaker et al., 2012).

Just as with adjunctive eHealth interventions (Danaher and Seeley,2009), a variation on the hybrid model involves mHealth interventionsthat are combined with non-technology treatment adjuncts including,for example, face-to-face treatment, telephone-based interventions(Glasgow, 2007; Glasgow et al., 2009), and pharmacotherapy (Brathet al., 2013; McGillicuddy et al., 2013). Moreover, a therapist portal ona personal computer could be used to increase the integration of anmHealth intervention with face-to-face treatment (e.g., Pramana et al.,2014) possibly by enhancing the supportive accountability associatedwith therapist/coach contact/feedback (Mohr et al., 2011).

4.3. Pervasive information architecture of mHealth interventions

The emergence of hybrid eHealth interventions delivered on bothpersonal computers and mobile devices requires a more sophisticatedview of information architecture — one that coordinates the interplaybetween different devices. This expanded view ubiquitous computing(“ubicomp”) underscores the need for “pervasive information architec-ture” that takes into consideration the broader interplay of devices andthe way that they share data/information (Greenfield, 2006; Resminiand Rosati, 2011). The simpler information architecture that informsthe top-down and left-right flow of websites on personal computersneeds to be considered within the added context that designs the inter-play of content delivered on other devices (see Fig. 5). For example, it isessential that the hybrid design enables the participant to benefit froma single coherent, tightly-integrated experience: “…the lack of coordina-tion between communicating or mutually-supporting channels is boundto affect the whole process. When multiple interactions are designed asunstructured and unrelated, but are in fact perceived as one single expe-rience by the user…structural gaps and behavioral inconsistencies are

Fig. 5. Schematic of data shared acrossmultiple devices in a hybrid eHealth/mHealth inter-vention that warrants consideration of pervasive information architecture.

common and unavoidable, and the sheer cognitive load and awkward-ness of switching back and forth between noncommunicating and appar-ently diverse touch points hampers the final user experience.” (p. 43,Resmini and Rosati, 2011).

To achieve the single user experience, the framing of the messagesshould be similar and thus familiar across devices. And the data theuser provides to the intervention as well as the information providedby the program to the participant must be immediately populatedacross all related devices. Similarly, although content may be deliveredon each device using a distinct schedule or calendar – probably relatedto the steps involved in making health behavior changes – the usershould perceive the program as having a single schedule tailored to fittheir needs. Seamlessness requires extremely tight integration. Similarly,Resmini and Rosati (2011) offer five heuristics to inform the design ofunified, integrated interventions:

1. Place-making—helps users find their way across complementarydigital, cross-channel, and even physical environments.

2. Consistency—provides amodel that works for the program's purposeand benefits end-users across different media, channels, and time.

3. Resilience—adapts program content to fit user needs, preferences.4. Reduction—presents program content to users in a way that is

simpler andmore usable than the underlying complexity of programdesign.

5. Discussion

This report presents a current snapshot of mHealth interventionissues. Some of these issues focused on the technology available in themost current generation of mobile phones that can be brought to bearon mHealth interventions. Also reviewed was how device functionalitymaps onto the essential strategies that can help encourage behaviorchange: the push/pull impact, the use of behavioral prompts via email,IVR, and notifications (text, pictures, audio/video), the possible inte-grating of data available from internal and external sensors and fromwithin the native operating system data, etc. Next we examinedthe increasingly important contributions of the field of pervasive in-formation architecture that informs the way to take best advantageof the fact that people tend to have – and selectively use–multiple tech-nology tools.

The use of technology to encourage behavior change using mHealthinterventions is in its infancy, and it is ripe for both creativity and em-pirical examination. For example:

• How practical (conceptually, behaviorally, financially) is it to develop,deliver, and maintain intervention content in one medium/channelversus others?

• Under what circumstance might an intervention use only a singlechannel (e.g., a mobile device) to encourage health behavior change?

• What barriers to efficacy might point to the need to deliver interven-tion content on a personal computer versus a mobile device? Howwould a staged-model apply to choosing one approach first followedby the second approach if unsuccessful?

• Is it important to assess the extent towhich individuals assigned to anmHealth intervention actually use a mobile device to access thatcontent? (See Turner-McGrievy and Tate (2014) and The NielsenCompany (2013)).

• Should participants be prevented (or actively discouraged) from usingmHealth content on their personal computers?

• To what extent are tablets used like personal computers or mobiledevices? Their screen size is larger than smartphones yet their interac-tivity still similarly constrained.Moreover, tablets tend not to be used ina portable manner during everyday routines.

• Is there a benefit to scheduling the delivery of a series of brief, interre-lated text messages – e.g., text message adaptations of sequentialBurma-Shave messaging (Wikipedia, 2014b) – rather than as indepen-dent chunks of content?

99B.G. Danaher et al. / Internet Interventions 2 (2015) 91–101

• In an effort to be sensitive to respondent burden, should longer andmore complex assessments associated with research trials be availableonly via personal computers rather than on mobile devices? Becauseof their brevity, could screening assessments be on both?

• Given that the boundaries between/among devices are becomingmorepermeable (e.g., Apple's Yosemite and iOS 8 operating systems nowdeliver text messages and phone calls to personal computers as wellas iPhones), does this foreshadow broader platform reach for textmessaging andother notifications in designinghybrid eHealth interven-tions?

• Are there certain mHealth tools (e.g., video, digital assistants, tailoredtext messages, etc.) that are particularly helpful in terms of strengthen-ing the working/therapeutic alliance with the intervention, whichresults in enhanced engagement and improved outcomes?

• When should interventions bedesigned/deliveredonlyon smartphonesversus delivered with other devices (e.g., personal computers) using ahybrid model? How does therapist/coach contact enhance participantengagement in mHealth interventions and affect outcome?

• How can data and results from mHealth interventions populate elec-tronic medical records (EMRs)?

Clearly, additional research is required to determine the proper rolesfor mHealth and personal-computer-based intervention components.Because the trend towards pervasive smartphone usage seems inexorableand the attraction to use smartphones and other mobile devices for be-havior change interventions is growing in parallel, intervention designerswill need to take care to avoid simply porting over intervention designsintended for personal computer users while using multiple channelsand devices. By leveraging the ever-increasing toolbox functionality ofmobile devices and addressing the context of pervasive informationarchitecture, we believe that mHealth intervention developers will bebetter able to achieve the promise of this burgeoning arena by engagingparticipants and encouraging meaningful behavior change.

Acknowledgments

The authors thank Edward Lichtenstein and ColeenHudkins for theirhelpful review of earlier drafts of this report. Thisworkwas supportedby grants from the National Cancer Society (R01-CA172205 andR01-CA140310).

References

Abraham, C., Michie, S., 2008. A taxonomy of behavior change techniques used in inter-ventions. Health Psychol. 27 (3), 379–387. http://dx.doi.org/10.1037/0278-6133.27.3.379.

Abu-Hasaballah, K., James, A., Aseltine Jr., R.H., 2007. Lessons and pitfalls of interactive voiceresponse in medical research. Contemp. Clin. Trials 28 (5), 593–602. http://dx.doi.org/10.1016/j.cct.2007.02.007.

Aguilera, A., Munoz, R.F., 2011. Text messaging as an adjunct to CBT in low-incomepopulations: a usability and feasibility pilot study. Prof. Psychol. Res. Pract. 42 (6),472–478.

Apple, 2013. iOS Human Interface Guidelines. Retrieved 1/6/2014, from. https://developer.apple.com/library/ios/documentation/userexperience/conceptual/mobilehig/MobileHIG.pdf.

Apple, 2014. iOS Deployment Technical Reference. Retrieved from. http://www.apple.com/ipad/business/it/deployment.html.

Bartholomew, L.K., Parcel, G.S., Kok, G., Gottlieb, N.H., Fernandez, M.E., 2011. Planning HealthPromotion Programs: An Intervention Mapping Approach. 3rd ed. Jossey-Bass, SanFrancisco.

Bashshur, R., Shannon, G., Krupinski, E., Grigsby, J., 2011. The taxonomy of telemedicine.Telemed. J. E Health 17 (6), 484–494. http://dx.doi.org/10.1089/tmj.2011.0103.

Beale, R., 2009. What does mobile mean? Int. J. Hum. Comput. Interact. 1 (3), 1–8.Bickmore, T.W., Mauer, D., Brown, T., 2009. Context awareness in a handheld exercise

agent. Pervasive Mob. Comput. 5 (3), 226–235. http://dx.doi.org/10.1016/j.pmcj.2008.05.004.

Brath, H., Morak, J., Kastenbauer, T., Modre-Osprian, R., Strohner-Kastenbauer, H.,Schwarz, M., Kort, W., Schreier, G., 2013. Mobile health (mHealth) based medicationadherence measurement — a pilot trial using electronic blisters in diabetes patients.Br. J. Clin. Pharmacol. 76 (Suppl. 1), 47–55. http://dx.doi.org/10.1111/bcp.12184.

Brendryen, H., Kraft, P., 2008. Happy ending: a randomized controlled trial of a digitalmulti-media smoking cessation intervention. Addiction 103 (3), 478–484. http://dx.doi.org/10.1111/j.1360-0443.2007.02119.x (discussion 485–476).

Brendryen, H., Drozd, F., Kraft, P., 2008. A digital smoking cessation program deliveredthrough internet and cell phone without nicotine replacement (happy ending):randomized controlled trial. J. Med. Internet Res. 10 (5), e51. http://dx.doi.org/10.2196/jmir.1005.

Brendryen, H., Kraft, P., Schaalma, H., 2010. Looking inside the black box: using interven-tionmapping to describe the development of the automated smoking cessation inter-vention ‘happy ending’. J. Smok. Cessat. 5 (1), 29–56. http://dx.doi.org/10.1375/jsc.5.1.29.

Brendryen, H., Johansen, A., Nesvag, S., Kok, G., Duckert, F., 2013. Constructing a theory- andevidence-based treatment rationale for complex eHealth interventions: developmentof an online alcohol intervention using an intervention mapping approach. JMIR Res.Protocol. 2 (1), e6. http://dx.doi.org/10.2196/resprot.2371.

Civljak, M., Stead, L.F., Hartmann-Boyce, J., Sheikh, A., Car, J., 2013. Internet-based interven-tions for smoking cessation. Cochrane Database Syst. Rev. 7, CD007078. http://dx.doi.org/10.1002/14651858.CD007078.pub4.

Clifton, L., Clifton, D.A., Pimentel, M.A., Watkinson, P.J., Tarassenko, L., 2013. Gaussianprocesses for personalized e-health monitoring with wearable sensors. IEEE Trans.Biomed. Eng. 60 (1), 193–197. http://dx.doi.org/10.1109/TBME.2012.2208459.

Cowie, M.R., Chronaki, C.E., Vardas, P., 2013. e-Health innovation: time for engagement withthe cardiology community. Eur. Heart J. 34 (25), 1864–1868. http://dx.doi.org/10.1093/eurheartj/ehs153.

CTIA, 2014. Annual Wireless Industry Summary Report, Year-end 2013 Results. Retrieved11/19/2014, from. http://www.ctia.org/your-wireless-life/how-wireless-works/annual-wireless-industry-survey.

Dallery, J., Glenn, I.M., Raiff, B.R., 2007. An Internet-based abstinence reinforcement treat-ment for cigarette smoking. Drug Alcohol Depend. 86 (2–3), 230–238. http://dx.doi.org/10.1016/j.drugalcdep.2006.06.013.

Dallery, J., Raiff, B.R., Grabinski, M.J., 2013. Internet-based contingency management topromote smoking cessation: a randomized controlled study. J. Appl. Behav. Anal. 46(4), 750–764. http://dx.doi.org/10.1002/jaba.89.

Danaher, B.G., Seeley, J.R., 2009. Methodological issues in research on web-based behavioralinterventions. Ann. Behav. Med. 38 (1), 28–39. http://dx.doi.org/10.1007/s12160-009-9129-0.

Danaher, B.G., McKay, H.G., Seeley, J.R., 2005. The information architecture of behaviorchange websites. J. Med. Internet Res. 7 (2), e12. http://dx.doi.org/10.2196/jmir.7.2.e12.

Danaher, B.G., Milgrom, J., Seeley, J.R., Stuart, S., Schembri, C., Tyler, M.S., Ericksen, J.,Lester, W., Gemmill, A.W., Kosty, D.B., Lewinsohn, P., 2013. MomMoodBooster web-based intervention for postpartum depression: feasibility trial results. J. Med. InternetRes. 15 (11), e242. http://dx.doi.org/10.2196/jmir.2876.

Delichatsios, H.K., Friedman, R.H., Glanz, K., Tennstedt, S., Smigelski, C., Pinto, B.M., Kelley,H., Gillman, M.W., 2001. Randomized trial of a “talking computer” to improve adults'eating habits. Am. J. Health Promot. 15 (4), 215–224.

Dulin, P.L., Gonzalez, V.M., Campbell, K., 2014. Results of a pilot test of a self-administeredsmartphone-based treatment system for alcohol use disorders: usability and earlyoutcomes. Subst. Abus. 35 (2), 168–175. http://dx.doi.org/10.1080/08897077.2013.821437.

eMarketer, 2014. Smartphone UsersWorldwide will Total 1.75 Billion in 2014. Retrieved 11/20/2014, from. http://www.emarketer.com/Article/Smartphone-Users-Worldwide-Will-Total-175-Billion-2014/1010536.

Estabrooks, P.A., Smith-Ray, R.L., 2008. Piloting a behavioral intervention deliveredthrough interactive voice response telephone messages to promote weight loss in apre-diabetic population. Patient Educ. Couns. 72 (1), 34–41. http://dx.doi.org/10.1016/j.pec.2008.01.007.

Estabrooks, P.A., Shoup, J.A., Gattshall, M., Dandamudi, P., Shetterly, S., Xu, S., 2009.Automated telephone counseling for parents of overweight children: a randomizedcontrolled trial. Am. J. Prev. Med. 36 (1), 35–42. http://dx.doi.org/10.1016/j.amepre.2008.09.024.

Fjeldsoe, B.S., Marshall, A.L., Miller, Y.D., 2009. Behavior change interventions delivered bymobile telephone short-message service. Am. J. Prev. Med. 36 (2), 165–173. http://dx.doi.org/10.1016/j.amepre.2008.09.040.

Fjeldsoe, B.S., Miller, Y.D., Marshall, A.L., 2010. MobileMums: a randomized controlled trialof an SMS-based physical activity intervention. Ann. Behav. Med. 39 (2), 101–111.http://dx.doi.org/10.1007/s12160-010-9170-z.

Fogg, B.J., 2003. Persuasive Technology: Using Computers to Change What We Think andDo. Morgan Kaufmann Publishers, San Francisco.

Fogg, B.J., 2007. The future of persuasion is mobile. In: Fogg, B.J., Eckles, D. (Eds.), Mobilepersuasion: 20 perspectives on the future of behavior change. Persuasive TechnologyLab, Stanford University, Captology Media, Stanford, CA, pp. 5–11.

Fox, S., Raine, L., 2014. The Web at 25 in the U.S.. Retrieved 7/3/2014, from. http://www.pewinternet.org/2014/02/27/the-web-at-25-in-the-u-s/.

Franklin, V.L., Waller, A., Pagliari, C., Greene, S.A., 2006. A randomized controlled trialof Sweet Talk, a text-messaging system to support young people with diabetes.Diabet. Med. 23 (12), 1332–1338. http://dx.doi.org/10.1111/j.1464-5491.2006.01989.x.

Free, C., Whittaker, R., Knight, R., Abramsky, T., Rodgers, A., Roberts, I.G., 2009. Txt2stop: apilot randomised controlled trial of mobile phone-based smoking cessation support.Tob. Control. 18 (2), 88–91. http://dx.doi.org/10.1136/tc.2008.026146.

Fry, J.P., Neff, R.A., 2009. Periodic prompts and reminders in health promotion and healthbehavior interventions: systematic review. J. Med. Internet Res. 11 (2), e16. http://dx.doi.org/10.2196/jmir.1138.

Glasgow, R.E., 2007. eHealth evaluation and dissemination research. Am. J. Prev. Med. 32(5 Suppl.), S119–S126. http://dx.doi.org/10.1016/j.amepre.2007.01.023.

100 B.G. Danaher et al. / Internet Interventions 2 (2015) 91–101

Glasgow, R.E., Christiansen, S., Smith, K.S., Stevens, V.J., Toobert, D.J., 2009. Developmentand implementation of an integrated, multi-modality, user-centered interactivedietary change program. Health Educ. Res. 24 (3), 461–471. http://dx.doi.org/10.1093/her/cyn042.

Google, 2014. Android Design. Retrieved 6/2/2014, from. https://developer.android.com/design/index.html.

Graham, A.L., Cobb, N.K., Papandonatos, G.D., Moreno, J.L., Kang, H., Tinkelman, D.G., Bock,B.C., Niaura, R.S., Abrams, D.B., 2011. A randomized trial of Internet and telephonetreatment for smoking cessation. Arch. Intern. Med. 171 (1), 46–53. http://dx.doi.org/10.1001/archinternmed.2010.451.

Greenfield, A., 2006. Everyware: The Dawning Age of Ubiquitous Computing. New Riders,Berkeley, CA.

Griffiths, K.M., Farrer, L., Christensen, H., 2010. The efficacy of internet interventions fordepression and anxiety disorders: a review of randomised controlled trials. Med.J. Aust. 192 (11 Suppl.), S4–S11.

Guiry, J.J., van de Ven, P., Nelson, J., Warmerdam, L., Riper, H., 2014. Activity recognitionwith smartphone support. Med. Eng. Phys. 36 (6), 670–675. http://dx.doi.org/10.1016/j.medengphy.2014.02.009.

Gustafson, D.H., McTavish, F.M., Chih, M.Y., Atwood, A.K., Johnson, R.A., Boyle, M.G., Levy,M.S., Driscoll, H., Chisholm, S.M., Dillenburg, L., Isham, A., Shah, D., 2014. Asmartphone application to support recovery from alcoholism: a randomized clinicaltrial. JAMA Psychiatry 71 (5), 566–572. http://dx.doi.org/10.1001/jamapsychiatry.2013.4642.

Heron, K.E., Smyth, J.M., 2010. Ecological momentary interventions: incorporating mobiletechnology into psychosocial and health behaviour treatments. Br. J. Health Psychol.15 (Pt 1), 1–39. http://dx.doi.org/10.1348/135910709X466063.

ict4depression, 2014. Moodbuster Components: Mobile Phone App, PhysiologicalSensors, Adherence Monitor, Website. Retrieved 10/22/2014, from. http://www.ict4depression.eu/moodbuster/.

iHeed Institute, 2011. mHealth Education: Harnessing the Mobile Revolution to Bridgethe Health Education and Training Gap in Developing Countries. Retrieved from.http://www.iheed.org/reports/iheedreport_2011.pdf.

Intille, S.S., Kukla, C., Farzanfar, R., Bakr, W., 2003. Just-in-time technology to encourageincremental, dietary behavior change. AMIA Annu Symp Proc, p. 874.

Kerwin, M., Nunes, F., Silva, P.A., 2012. Dance! Don't fall— preventing falls and promotingexercise at home. Stud. Health Technol. Inform 177, 254–259.

Klasnja, P., Pratt, W., 2012. Healthcare in the pocket: mapping the space of mobile-phonehealth interventions. J. Biomed. Inform. 45 (1), 184–198. http://dx.doi.org/10.1016/j.jbi.2011.08.017.

Kotz, D., 2011. A threat taxonomy for mHealth privacy. Paper presented at the Third Inter-national Conference on Communication Systems and Networks (COMSNETS),Bangalore.

Krishna, S., Boren, S.A., Balas, E.A., 2009. Healthcare via cell phones: a systematic review.Telemed. J. E Health 15 (3), 231–240. http://dx.doi.org/10.1089/tmj.2008.0099.

Kumar, S., 2012. Predicting Smoking Abstinence via MobileMentoring of Stress and SocialContext (R01DA035502). NIH Research Portfolio Online Reporting Tools. Retrieved 10/24/2014, from. http://projectreporter.nih.gov/project_info_description.cfm?aid=8720744&icde=22255235&ddparam=&ddvalue=&ddsub=&cr=2&csb=default&cs=ASC.

Kumar, S., 2014. Behavioral Assessment withMobile Sensors. Retrieved 10/26/2014, from.http://nih.gov/news/videos/2014/0501-mobilesensors.htm.

Kumar, S., Nilsen, W.J., Abernethy, A., Atienza, A., Patrick, K., Pavel, M., Riley, W.T., Shar, A.,Spring, B., Spruijt-Metz, D., Hedeker, D., Honavar, V., Kravitz, R., Lefebvre, R.C., Mohr,D.C., Murphy, S.A., Quinn, C., Shusterman, V., Swendeman, D., 2013. Mobile healthtechnology evaluation: the mHealth evidence workshop. Am. J. Prev. Med. 45 (2),228–236. http://dx.doi.org/10.1016/j.amepre.2013.03.017.

Lenhart, A., 2010. Cell phones and American adults. Pew Internet & American Life Project(http://www.pewinternet.org/~/media/Files/Reports/2010/PIP_Adults_Cellphones_Report_2010.pdf).

Lenhart, A., Purcell, K., Smith, A., Zickuhr, K., 2010. Social media & mobile internet useamong teens and young adults. Pew Internet & American Life Project (Retrieved12/1/2011, 2011, from http://www.pewinternet.org/Reports/2010/Social-Media-and-Young-Adults.aspx).

Lewis, T.L., Boissaud-Cooke, M.A., Aungst, T.D., Eysenbach, G., 2014. Consensus on use ofthe term “app” versus “application” for reporting of mHealth research. J. Med. Inter-net Res. 16 (7), e174. http://dx.doi.org/10.2196/jmir.3460.

Liu, W.T., Wang, C.H., Lin, H.C., Lin, S.M., Lee, K.Y., Lo, Y.L., Hung, S.H., Chang, Y.M., Chung,K.F., Kuo, H.P., 2008. Efficacy of a cell phone-based exercise programme for COPD. Eur.Respir. J. 32 (3), 651–659. http://dx.doi.org/10.1183/09031936.00104407.

Luxton, D.D., McCann, R.A., Bush, N.E., Mishkind, M.C., Reger, G.M., 2011. mHealth formental health: integrating smartphone technology in behavioral healthcare. Prof.Psychol. Res. Pract. 42 (6), 505–512. http://dx.doi.org/10.1037/a0024485.

Mahoney, D.F., Tarlow, B.J., Jones, R.N., 2003. Effects of an automated telephone supportsystem on caregiver burden and anxiety: findings from the REACH for TLC interven-tion study. Gerontologist 43 (4), 556–567.

McGillicuddy, J.W., Gregoski, M.J., Weiland, A.K., Rock, R.A., Brunner-Jackson, B.M., Patel,S.K., Thomas, B.S., Taber, D.J., Chavin, K.D., Baliga, P.K., Treiber, F.A., 2013. Mobile healthmedication adherence and blood pressure control in renal transplant recipients:a proof-of-concept randomized controlled trial. JMIR Res. Protocol. 2 (2), e32.http://dx.doi.org/10.2196/resprot.2633.

MD2K, 2014. NIH Center of Excellence on Mobile Sensor Data-to-Knowledge: AdvancingBiomedical Discovery and Improving Health Through Mobile Sensor Big Data.Retrieved 10/26/2014, from. http://md2k.org.

Mohr, D.C., Cuijpers, P., Lehman, K., 2011. Supportive accountability: a model for providinghuman support to enhance adherence to eHealth interventions. J. Med. Internet Res. 13(1), e30. http://dx.doi.org/10.2196/jmir.1602.

Mohr, D.C., Burns, M.N., Schueller, S.M., Clarke, G., Klinkman, M., 2013. Behavioral inter-vention technologies: evidence review and recommendations for future research inmental health. Gen. Hosp. Psychiatry 35 (4), 332–338. http://dx.doi.org/10.1016/j.genhosppsych.2013.03.008.

Mohr, D.C., Schueller, S.M., Montague, E., Burns, M.N., Rashidi, P., 2014. The behavioral in-tervention technologymodel: an integrated conceptual and technological frameworkfor eHealth and mHealth interventions. J. Med. Internet Res. 16 (6), e146. http://dx.doi.org/10.2196/jmir.3077.

Moore, H.K., Hughes, C.W., Mundt, J.C., Rush, A.J., Macleod, L., Emslie, G.J., Jain, S., Geralts,D.S., Bernstein, I.H., Horrigan, J.P., Trivedi, M.H., Greist, J.H., 2007. A pilot study of anelectronic, adolescent version of the quick inventory of depressive symptomatology.J. Clin. Psychiatry 68 (9), 1436–1440.

Myung, S.K., McDonnell, D.D., Kazinets, G., Seo, H.G., Moskowitz, J.M., 2009. Effects ofweb- and computer-based smoking cessation programs: meta-analysis of random-ized controlled trials. Arch. Intern. Med. 169 (10), 929–937. http://dx.doi.org/10.1001/archinternmed.2009.109.

NCI, 2014. NCI QuitPal Quit Smoking App. Retrieved 11/5/2014, from. http://www.cancer.gov/cancertopics/tobacco/smoking/quitting/nciquitpal-app.

Nielsen Norman Group, 2014. Mobile: Native Apps, Web Apps, and Hybrid Apps.Retrieved 6/2/2014, from. http://www.nngroup.com/articles/mobile-native-apps/.

Nielsen, J., Budiu, R., 2013. Mobile Usability. New Riders (Pearson Education), Berkeley, CA.Nuance, 2014. Self-service with an Intelligent Touch of Humanity. Retrieved 5/20/2014,

from. http://www.nuance.com/landing-pages/products/nina/default.asp.Oake, N., Jennings, A., vanWalraven, C., Forster, A.J., 2009. Interactive voice response systems

for improving delivery of ambulatory care. Am. J. Manag. Care 15 (6), 383–391.Osgood-Hynes, D.J., Greist, J.H., Marks, I.M., Baer, L., Heneman, S.W., Wenzel, K.W., Manzo,

P.A., Parkin, J.R., Spierings, C.J., Dottl, S.L., Vitse, H.M., 1998. Self-administered psycho-therapy for depression using a telephone-accessed computer system plus booklets:an open U.S.–U.K. study. J. Clin. Psychiatry 59 (7), 358–365.

Patel, S.N., Kientz, J.A., Hayes, G.R., Bhat, S., Abowd, G.D., 2006. Farther than you maythink: an empirical investigation of the proximity of users to their mobile phones.UbiComp 2006 Ubiquitous Computingpp. 123–140.

Patrick, K., Griswold, W.G., Raab, F., Intille, S.S., 2008. Health and the mobile phone. Am.J. Prev. Med. 35 (2), 177–181. http://dx.doi.org/10.1016/j.amepre.2008.05.001.

PhotoVoice, 2014. PhotoVoice: Participatory Photography for Social Change. Retrieved 6/3/2014, from. http://www.photovoice.org.

Piette, J.D., 2000. Interactive voice response systems in the diagnosis and management ofchronic disease. Am. J. Manag. Care 6 (7), 817–827.

Piette, J.D., 2002. Enhancing support via interactive technologies. Curr. Diab. Rep. 2 (2),160–165.

Pinto, B.M., Friedman, R., Marcus, B.H., Kelley, H., Tennstedt, S., Gillman, M.W., 2002.Effects of a computer-based, telephone-counseling system on physical activity. Am.J. Prev. Med. 23 (2), 113–120.

Pramana, G., Parmanto, B., Kendall, P.C., Silk, J.S., 2014. The SmartCAT: an m-health platformfor ecological momentary intervention in child anxiety treatment. Telemed. J. E Health20 (5), 419–427. http://dx.doi.org/10.1089/tmj.2013.0214.

Prestwich, A., Perugini, M., Hurling, R., 2009. Can the effects of implementation intentionson exercise be enhanced using text messages? Psychol. Health 24 (6), 677–687.http://dx.doi.org/10.1080/08870440802040715.

Ramelson, H.Z., Friedman, R.H., Ockene, J.K., 1999. An automated telephone-basedsmoking cessation education and counseling system. Patient Educ. Couns. 36 (2),131–144.

Reid, R.D., Pipe, A.L., Quinlan, B., Oda, J., 2007. Interactive voice response telephony to pro-mote smoking cessation in patients with heart disease: a pilot study. Patient Educ.Couns. 66 (3), 319–326. http://dx.doi.org/10.1016/j.pec.2007.01.005.

Reidel, K., Tamblyn, R., Patel, V., Huang, A., 2008. Pilot study of an interactive voice re-sponse system to improve medication refill compliance. BMC Med. Inform. Decis.Mak. 8, 46. http://dx.doi.org/10.1186/1472-6947-8-46.

Resmini, A., Rosati, L., 2011. Pervasive Information Architecture: Designing Cross-channelUser Experiences. Morgan Kaufmann, Burlington, MA.

Riley, W.T., Rivera, D.E., Atienza, A.A., Nilsen, W., Allison, S.M., Mermelstein, R., 2011.Health behavior models in the age of mobile interventions: are our theories up tothe task? Transl. Behav. Med. 1 (1), 53–71. http://dx.doi.org/10.1007/s13142-011-0021-7.

Riper, H., Spek, V., Boon, B., Conijn, B., Kramer, J., Martin-Abello, K., Smit, F., 2011. Effec-tiveness of E-self-help interventions for curbing adult problem drinking: a meta-analysis. J. Med. Internet Res. 13 (2), e42. http://dx.doi.org/10.2196/jmir.1691.

Ritterband, L.M., Thorndike, F., 2006. Internet interventions or patient education websites? J. Med. Internet Res. 8 (3), e18. http://dx.doi.org/10.2196/jmir.8.3.e18 (authorreply e19.).

Robinson, S., Perkins, S., Bauer, S., Hammond, N., Treasure, J., Schmidt, U., 2006. Aftercareintervention through text messaging in the treatment of bulimia nervosa—feasibilitypilot. Int. J. Eat. Disord. 39 (8), 633–638. http://dx.doi.org/10.1002/eat.20272.

Rodgers, A., Corbett, T., Bramley, D., Riddell, T., Wills, M., Lin, R.B., Jones, M., 2005. Do usmoke after txt? Results of a randomised trial of smoking cessation using mobilephone text messaging. Tob. Control. 14 (4), 255–261. http://dx.doi.org/10.1136/tc.2005.011577.

Sarker, H., Sharmin, M., Ali, A.A., Rahman, M.M., Bari, R., Hossain, S.M., Kumar, S., 2014.Assessing the availability of users to engage in just-in-time intervention in the naturalenvironment. UBbiComp '14 http://dx.doi.org/10.1145/2632048.2636082.

Schaalma, H., Kok, G., 2009. Decoding health education interventions: the times are a-changin'. Psychol. Health 24 (1), 5–9. http://dx.doi.org/10.1080/08870440903126348.

Scully, C.G., Lee, J., Meyer, J., Gorbach, A.M., Granquist-Fraser, D., Mendelson, Y., Chon, K.H.,2012. Physiological parameter monitoring from optical recordings with a mobilephone. IEEE Trans. Biomed. Eng. 59 (2), 303–306. http://dx.doi.org/10.1109/TBME.2011.2163157.

101B.G. Danaher et al. / Internet Interventions 2 (2015) 91–101

Shapiro, J.R., Bauer, S., Andrews, E., Pisetsky, E., Bulik-Sullivan, B., Hamer, R.M., Bulik, C.M.,2010. Mobile therapy: use of text-messaging in the treatment of bulimia nervosa. Int.J. Eat. Disord. 43 (6), 513–519. http://dx.doi.org/10.1002/eat.20744.

Silva, P.A., Nunes, F., Vasconceloas, A., Kerqwin, M., Moutinho, R., Teixeira, P., 2013. Usingthe smartphone accelerometer to monitor fall risk while playing a game: the designand usability evaluation of Dance! Don't Fall. In: Schmorrow, D.D., Fidopiastis, C.M.(Eds.), Foundations of Augmented Cognition: 7th International Conference, HCI Interna-tional 2013 Vol. 8027. Springer-Verlag, Berlin. ISBN: 978-3-642-39453-9, pp. 754–763.

Six, B.L., Schap, T.E., Zhu, F.M., Mariappan, A., Bosch, M., Delp, E.J., Ebert, D.S., Kerr, D.A.,Boushey, C.J., 2010. Evidence-based development of a mobile telephone food record.J. Am. Diet. Assoc. 110 (1), 74–79. http://dx.doi.org/10.1016/j.jada.2009.10.010.

Smith, A., 2011. Americans and text messaging. Pew Internet & American Life Project(Retrieved 9/3/2014, from http://www.pewinternet.org/~/media//Files/Reports/2011/Americans and Text Messaging.pdf).

Smith, A., 2013. Smartphone Ownership — 2013 Update. http://www.pewinternet.org/~/media//Files/Reports/2012/Smartphone ownership 2012.pdf.

Spek, V., Cuijpers, P., Nyklicek, I., Riper, H., Keyzer, J., Pop, V., 2007. Internet-based cogni-tive behaviour therapy for symptoms of depression and anxiety: a meta-analysis.Psychol. Med. 37 (3), 319–328. http://dx.doi.org/10.1017/S0033291706008944.

Strecher, V.J., 2007. Internet methods for delivering behavioral and health-related inter-ventions (eHealth). Annu. Rev. Clin. Psychol. 3, 53–76. http://dx.doi.org/10.1146/annurev.clinpsy.3.022806.091428.

Strecher, V.J., 2008. The internet: just another smoking cessation tool? Addiction 103 (3),485–486.

The Nielsen Company, 2013. The Mobile Consumer: A Global Snapshot. from.http://www.nielsen.com/content/dam/corporate/uk/en/documents/Mobile-Consumer-Report-2013.pdf.

Titov, N., 2011. Internet-delivered psychotherapy for depression in adults. Curr. Opin.Psychiatry 24 (1), 18–23. http://dx.doi.org/10.1097/YCO.0b013e32833ed18f.

Turner-McGrievy, G.M., Tate, D.F., 2014. Are we sure that Mobile Health is really mobile?An examination of mobile device use during two remotely-delivered weight loss in-terventions. Int. J. Med. Inform. 83 (5), 313–319. http://dx.doi.org/10.1016/j.ijmedinf.2014.01.002.

University of Pittsburgh, 2013. SmartCAT (Smartphone-enhanced Child Anxiety Treatment).Retrieved 11/25/2014, from. https://www.youtube.com/watch?v=b1w2Zq1BJPI.

USDHHS, 2014. Using health text messages to improve consumer health knowledge,behaviors, and outcomes: an environmental scan. Retrieved 9/5/2014, from.http://www.hrsa.gov/healthit/txt4tots/environmentalscan.pdf.

Wang, A., An, N., Lu, X., Chen, H., Li, C., Levkoff, S., 2014. A classification scheme for ana-lyzing mobile apps used to prevent and manage disease in late life. 21(1 e6). JMIRmHealth and uHealth. http://dx.doi.org/10.2196/mhealth.2877.

Wantland, D.J., Portillo, C.J., Holzemer, W.L., Slaughter, R., McGhee, E.M., 2004. The effec-tiveness of Web-based vs. non-Web-based interventions: a meta-analysis of behav-ioral change outcomes. J. Med. Internet Res. 6 (4), e40. http://dx.doi.org/10.2196/jmir.6.4.e40.

Warmerdam, L., Riper, H., Klein, M., van den Ven, P., Rocha, A., Ricardo Henriques, M.,Tousset, E., Silva, H., Andersson, G., Cuijpers, P., 2012. Innovative ICT solutions to im-prove treatment outcomes for depression: the ICT4Depression project. Stud. HealthTechnol. Inform. 181, 339–343.

Warren, T., 2014. The Story of Cortana, Microsoft's Siri Killer. Retrieved 5/20/2014, from.http://www.theverge.com/2014/4/2/5570866/cortana-windows-phone-8-1-digital-assistant.

Webb, T.L., Joseph, J., Yardley, L., Michie, S., 2010. Using the internet to promote healthbehavior change: a systematic review and meta-analysis of the impact of theoreticalbasis, use of behavior change techniques, and mode of delivery on efficacy. J. Med. In-ternet Res. 12 (1), e4. http://dx.doi.org/10.2196/jmir.1376.

Whittaker, R., Maddison, R., McRobbie, H., Bullen, C., Denny, S., Dorey, E., Ellis-Pegler, M.,van Rooyen, J., Rodgers, A., 2008. A multimedia mobile phone-based youth smokingcessation intervention: findings from content development and piloting studies.J. Med. Internet Res. 10 (5), e49. http://dx.doi.org/10.2196/jmir.1007.

Whittaker, R., Borland, R., Bullen, C., Lin, R.B., McRobbie, H., Rodgers, A., 2009. Mobilephone-based interventions for smoking cessation. Cochrane Database Syst. Rev. 7(4), CD006611. http://dx.doi.org/10.1002/14651858.CD006611.pub2 (Oct).

Whittaker, R., Merry, S., Stasiak, K., McDowell, H., Doherty, I., Shepherd, M., Dorey, E.,Parag, V., Ameratunga, S., Rodgers, A., 2012. MEMO—a mobile phone depression pre-vention intervention for adolescents: development process and postprogramfindingson acceptability from a randomized controlled trial. J. Med. Internet Res. 14 (1), e13.http://dx.doi.org/10.2196/jmir.1857.

WHO, 2011. mHealth: New Horizons for Health Through Mobile Technologies Global Obser-vatory for eHealth. Vol. 3. Retrieved from. http://www.who.int/goe/publications/goe_mhealth_web.pdf.

Wikipedia, 2014a. Responsive Web Design. Retrieved 10/28/2014, from. http://en.wikipedia.org/wiki/Responsive_web_design.

Wikipedia, 2014b. Siri. Retrieved 10/20/2014, from. http://en.wikipedia.org/wiki/Siri.Wikipedia, 2014c. Adaptive Web Design. Retrieved 10/25/2014, from. http://en.wikipedia.

org/wiki/Adaptive_web_design.Wikipedia, 2014d. Burma-Shave. Retrieved 11/26/2014, from. http://en.wikipedia.org/

wiki/Burma-Shave.Wikipedia, 2014e. Feature Phone. Retrieved 9/3/2014, from. http://en.wikipedia.org/wiki/

Feature_phone.Wikipedia, 2014f. Google Now. Retrieved 5/20/2014, from. http://en.wikipedia.org/wiki/

Google_Now.Wikipedia, 2014g. . mHealth. Retrieved 9/8/2014, from. http://en.wikipedia.org/wiki/

MHealth.Wikipedia, 2014h. . Retrieved 6/3/2014, from. http://en.wikipedia.org/wiki/Photovoice-

Photovoice_in_international_development.Woolford, S.J., Khan, S., Barr, K.L., Clark, S.J., Strecher, V.J., Resnicow, K., 2012. A picture

may be worth a thousand texts: obese adolescents' perspectives on a modifiedphotovoice activity to aid weight loss. Child Obes. 8 (3), 230–236. http://dx.doi.org/10.1089/chi.2011.0095.

Ybarra, M.L., Holtrop, J.S., Bagci Bosi, A.T., Emri, S., 2012. Design considerations in developinga text messaging program aimed at smoking cessation. J. Med. Internet Res. 14 (4),e103. http://dx.doi.org/10.2196/jmir.2061.

Yoon, K.H., Kim, H.S., 2008. A short message service by cellular phone in type 2 diabeticpatients for 12 months. Diabetes Res. Clin. Pract. 79 (2), 256–261. http://dx.doi.org/10.1016/j.diabres.2007.09.007.

Yuan, M.J., 2005. What is a Smartphone?. Retrieved 8/17/2014, from. http://www.oreillynet.com/pub/a/wireless/2005/08/23/whatissmartphone.html.