It's Not Just the Movement: Experiential Information Needed for Stroke Telerehabilitation
Adegboyega Akinsiku, Ignacio Avellino, Yasmin Graham, Helena M. Mentis
IIt’s Not Just the Movement: Experiential Information Needed forStroke Telerehabilitation
Adegboyega Akinsiku
University of Maryland, Baltimore CountyBaltimore, MD, [email protected]
Ignacio Avellino
University of Maryland, Baltimore CountyBaltimore, MD, [email protected]
Yasmin Graham
University of Maryland, Baltimore CountyBaltimore, MD, [email protected]
Helena M. Mentis
University of Maryland, Baltimore CountyBaltimore, MD, [email protected]
ABSTRACT
Telerehabilitation systems for stroke survivors have been predom-inantly designed to measure and quantify movement in order toguide and encourage rehabilitation regular exercises at home. Weset out to study what aspect of the movement data was essential, tobetter inform sensor design. We investigated face-to-face stroke re-habilitation sessions through a series of interviews and observationsinvolving 16 stroke rehabilitation specialists including physiatrists,physical therapists, and occupational therapists. We found thatspecialists are not solely interested in movement data, and thatexperiential information about stroke survivors’ lived experienceplays an essential role in specialists interpreting movement data andcreating a rehabilitation plan. We argue for a reconceptualizationin stroke telerehabilitation that is more inclusive of non-movementdata, and present design implications to better account for experi-ential information in telerehabilitation systems.
CCS CONCEPTS • Human-centered computing → Empirical studies in HCI . KEYWORDS
Telerehabilitation; Stroke; Physical medicine and rehabilitation;Therapy
ACM Reference Format:
Adegboyega Akinsiku, Ignacio Avellino, Yasmin Graham, and Helena M.Mentis. 2021. It’s Not Just the Movement: Experiential Information Neededfor Stroke Telerehabilitation. In
CHI Conference on Human Factors in Com-puting Systems (CHI ’21), May 8–13, 2021, Yokohama, Japan.
ACM, New York,NY, USA, 12 pages. https://doi.org/10.1145/3411764.3445663
CHI ’21, May 8–13, 2021, Yokohama, Japan © 2021 Copyright held by the owner/author(s). Publication rights licensed to ACM.This is the author’s version of the work. It is posted here for your personal use. Notfor redistribution. The definitive Version of Record was published in
CHI Conferenceon Human Factors in Computing Systems (CHI ’21), May 8–13, 2021, Yokohama, Japan ,https://doi.org/10.1145/3411764.3445663.
Telerehabilitation involves leveraging technologies (e.g., the In-ternet) to facilitate the communication of information between apatient and their clinician at a distance in order to provide rehabil-itation services [8]. However, the type of information that needsto be communicated is not well defined. Prior to the viability oftelerehabilitation services, researchers and developers designednon-internet home-based therapy systems with the intention ofincreasing patients’ access and longevity to rehabilitation services.These home-based therapy systems largely focused on motivat-ing exercise and creating engagement (e.g., through gamification)at home. As telerehabilitation systems’ viability and developmentincreased, they seem to have organically followed the focus ofhome-based therapy technologies, where patients autonomouslyperform exercises, and movement data was primarily captured andconveyed to rehabilitation specialists. Thus, research and devel-opment both in telerehabilitation and home-based systems havelargely aimed at sensing movement data.Our research was initially guided by this focus on collectingand sharing movement data. We set out to determine the typesand exactness of movement data needed in stroke rehabilitation,with the goal of informing the design of telerehabilitation systemsfor low-resource communities. As we began to observe face-to-face rehabilitation sessions and interview physiatrists, physicaltherapists, and occupational therapists, we instead found that theinformation that they dedicated effort to extract, understand, andintegrate into their care plans, is incongruent with the currentdesign paradigm of telerehabilitation systems.In this study, we investigate the information exchanged betweenstroke survivors, clinicians, and caregivers in co-located in-clinicstroke rehabilitation sessions, with the goal of informing the designof future stroke telerehabilitation systems, such that patients andspecialists can attain the benefits of co-located interaction. What welearned is that the information needed by rehabilitation specialistsis not really a detailed understanding of the movement data, butrather a deep understanding of the experiential information , suchas the stroke survivor’s emotions and motivations. We show thatexperiential information is information shared in stroke rehabili-tation. Therefore, we posit for a paradigm reconceptualization fortelerehabilitation system design, in which telerehabilitation has afocus on communicating the patient’s situated context. a r X i v : . [ c s . H C ] F e b HI ’21, May 8–13, 2021, Yokohama, Japan Akinsiku, Avellino, et al.
Our contributions to HCI research on stroke telerehabilitationare: (1) The definition and composition of the experiential infor-mation needed in stroke rehabilitation, (2) An explanation of ourproposed paradigm reconceptualization for future stroke telereha-bilitation research, and, (3) Implications for the design and devel-opment of stroke telerehabilitation systems.
Stroke is one of the leading causes of long-term disability in theUnited States [6]. The American Heart Association and AmericanStroke Association projected in 2013 that the costs associated withstroke will increase 129% by 2030, and concluded that more rehabil-itation and acute care services are needed to address stroke becauseof the national healthcare costs increasing yearly [31]. Stroke sur-vivors in particular, will spend directly and indirectly an averageof $103 ,
576 over their lifetime on treatment [14]. Rehabilitationafter a stroke has a high cost as it involves a wide variety of experi-enced clinicians and specialized equipment. Unfortunately, accessto specialized rehabilitation locations puts high-level care outsideof the reach of many US citizens, including those in rural and low-resource communities [6]. The ongoing COVID-19 pandemic hasexacerbated these challenges, as today remote healthcare is theonly treatment option even for people where distance and cost wasnot a barrier.To give background, we describe the stroke rehabilitation processthrough the perspective of the multiple rehabilitation specialiststhat create rehabilitation care plans for outpatient stroke survivors.These specialists coordinate through an extensive care networkthat includes: (1) the stroke survivor, (2) caregivers (i.e., the strokesurvivor’s immediate care network), (3) medical specialists (e.g.,physiatrists, cardiologists, and neurologists), and (4) allied healthspecialists (physical therapists, occupational therapists, and speech-language therapists). The network engages in an extensive amountof co-interpretation [24], a collaborative and interpretive process toassess movement and treatment efficacy, and care coordination [22],organizing the different aspects of care.
Physiatrists (PHY).
Physical Medicine and Rehabilitation(PM&R), or physiatry, is the branch of medicine that treats individ-uals with physical impairments, functional limitations, pain, anddisabilities that affect the brain, spinal cord, nerves, bones, joints,ligaments, muscles, and tendons [4, 27]. Physiatrists are the primarymedical doctors that guide stroke rehabilitation treatment, preced-ing physical and occupational therapy. They are key for acquiringa holistic perspective on functional and motor stroke rehabilita-tion, even if, to our knowledge, no other qualitative HCI paper hasincluded physiatrists in their studies.
Physical Therapist (PT) and Occupational Therapist(OT).
A PT provides care to restore and maintain a patient’s sensoryand motor abilities (e.g., improve gross motor movement), whereasan OT provides care to reduce the effects of a patient’s disabilitiesthrough adaptation (e.g., retrain self-care skills) [26, 28]. Both spe-cialists have aligned objectives in providing goal-oriented care totheir patients, to ultimately restore functional ability and mobility,and improve the quality of life through a patient-centered approach . Patient-Centered Care.
Patient-centered care takes intoaccount the individual needs, values, and expressed interest of pa-tients; and it has been identified as a gap in the US health system,with the Institute of Medicine urging the United States Congress toestablish funds for this purpose [30]. Six primary dimensions makeup patient centered care [12]: (1) respect for patients’ values, (2)coordination and integrative care, (3) information, communication,and education, (4) physical comfort, (5) emotional support to combatfear and anxiety, and (6) involvement of patients’ family and friends.Patient-centered care is enacted in face-to-face stroke rehabilita-tion sessions by specialists assessing the patient’s rehabilitationprogress, having their patients conduct interventions (exercises andactivities), and then create or modify an existing rehabilitation careplan in concert with the patient. The care plan is typically a set ofprescribed interventions that the specialists evaluate and updateperiodically to monitor progress, making it dynamic and evolving.Specialists typically document their assessment using the SOAPmethod (Subjective, Objective, Assessment, and Plan), a methodwidely used by healthcare professionals to fill out patient notesduring an appointment, and used to promote continuity in healthrecords [41].
Technology for stroke rehabilitation can be classified as home-basedtherapy systems , which are not connected online to remote spe-cialists, and telerehabilitation systems , which communicates reha-bilitation information and data through the Internet to a remotespecialist [8]. Telerehabilitation systems vary in implementation,but they focus on, asynchronously or synchronously, connectingpatients to remote specialists and transmit: (1) communication data(e.g., audio, video or text message) and/or (2) sensor-based data(e.g., movement).After reviewing systematic and scoping reviews of telerehabili-tation systems from the last 10 years [5, 17, 19, 36–40], we noticeda trend that current stroke telerehabilitation system design hascentralized the asynchronous and synchronous communication ofmovement sensor data. This trend has its own complexities outsidethe scope of this work, and deserves to be studied on its own inthe future. The following review of home-based and telerehabilita-tion design below exposes a clear trend of focusing on movementdata, typically originating from sensors, rather than consideringthe importance of other information needs and goals. Most of thisresearch is driven by generating solutions that increase exercisemotivation.
Data needs in stroke rehabilitation systems had been codified in2009 in Egglestone et al’s [11] design framework for home-basedstroke therapy systems. Through workshops with stroke survivorsand clinicians, the authors identified (1) background information,such as the stroke’s disruptive effects on patients or the contribu-tion of professionals to recovery, and (2) exercise execution datafor evaluation, that can be gathered through sensors or self-report.What has transpired since then is a litany of sensor-based systemsfocused on gathering data to support the second information need. xperiential Information Needed for Stroke Telerehabilitation CHI ’21, May 8–13, 2021, Yokohama, Japan
What we have not seen is a consideration of the first form of in-formation need. Specifically, identifying the types of “background”information and how to collect the information in way to presentthe two types of information to rehabilitation specialists.Sensor development has been an important step for telerehabili-tation and home-based systems to work—we are not denying that.However, consider the following five recently published systems: alow-cost, wireless home-based rehabilitation sensor that reliablycaptures upper-limb arm posture and movement [21];
Us’em [7], awristband-like activity monitor of arm–hand performance designedstrictly for patients to motivate the use of an impaired arm duringeveryday activities; mRes [42], a low-cost device that measures ro-tational movement, aimed at training dorsal wrist extension andfinger manipulation (both in supination and pronation), with anAPI for information exchange with telerehabilitation systems; Thecombination of Microsoft’s Kinect sensor data with machine learn-ing to automatically assess stroke rehabilitation exercises [20]; and, ArmSleeve [34], a sensor-embedded sleeve that captures objectiveupper limb data in patients’ daily life, outside rehabilitation exer-cises, creating a visualization for OTs in a dashboard. In all of these,the initial focus is on the valid collection of data towards motivat-ing correct movement. Interestingly, in the evaluation of this lastsystem, the authors found a major limitation to interpreting thismovement data: the lack of contextual information.Motivating a patient to perform, or more so “correctly” perform,a movement or exercise has been a prevailing goal in many of thesystems designed. For instance, Alankus et al. [3] studied therapeu-tic games using Nintendo Wii remote controllers and a webcamto sense movement, emphasizing the role of home-based strokerehabilitation games in keeping monotony low while providing per-formance feedback to specialists. In their followup work, Alankuset al. [2] specifically used the Wii remote to reliably sense com-pensatory movement (i.e. “incorrect” movements to achieve thesame outcome because “correct” movement is not possible or tiring)during gameplay, and then recommend feedback mechanisms toreduce such compensatory movements. Likewise, mobile phonesensors have also been used for keeping rehabilitation engaging,measuring movement in upper-limb rehabilitation while providinginstant feedback on the screen through a game [13].We do not disagree that giving immediate feedback to patientsabout the correctness of exercise execution can help improve theirperformance as well as motivate them to try harder. For instance,vibrotactile feedback [15] in an Arm Usage Coach (AUC), or usinga Virtual Reality headset to be immersed in a 3D game that adaptsits intensity to increase engagement [1]. What we are questioningwith this paper is how useful is movement information in a telere-habilitation context and how might the focus on this design trendmiss the information needs of rehabilitation specialists. In the development of telerehabilitation systems, there is addedcomplexity on determining what data/information to transmit toa care specialist, and how to present it in a meaningful way. For https://developer.microsoft.com/en-us/windows/kinect/ instance, after usability testing of TeleREHA , Perry et al. [33] foundthere are data needs in planning (configuration, game parameter-izing and scheduling), executing (measuring exercise data), andassessing (viewing data). In response, Postolache et al. [35] pro-posed intelligent telerehabilitation assistants called
RehabilitativeTeleHealthCare , where sensor signals are processed and combinedinto visualizations for caregivers, including physical movement(e.g., posture and daily walking movement), physiological data (e.g.,heart rate, oxygen saturation and respiratory rate) and localization.However, there have been indications that movement data isnot enough in the telerehabilitation context. Dekker-van Weeringet al. [10] evaluated clinician needs using a telerehabilitation systemand pointed to the system’s need to integrate patient context. Forinstance, integrating patient’s mood to allow them to skip a day ofexercise, or engagement with a therapist when choosing exercises.Lastly, in telerehabilitation co-design sessions with traumatic braininjury clinicians, How et al. [16] concluded that a successful systemneeds to adapt to a patient’s physical, cognitive and emotional state,the evolution of their rehabilitation history, and the surrounding lifecontext such as social life. Note that the authors argue that designers should take into account these aspects when building systems, butthey do not study in depth what this contextual information is orhow its role in the system.In summary, stroke home-based therapy and telerehabilitationsystems have followed a trend of collecting movement data. Thistrend leads to systems that provide little to no context of a patient’ssituated condition and excludes the collection of information likesubjective data. In contrast, our research aims to integrate all com-ponents of face-to-face stroke rehabilitation into a telerehabilitationsystem design. We conducted a field study focused on face-to-facesessions to reveal information and practices that might not be inte-grated into existing telerehabilitation tools.
As we were interested in the the complex and rich informationexchange that occurs in co-located rehabilitation, we conductedan ethnographically-informed field study involving rehabilitationsessions observations, interviews with specialists, and a reviewof stroke rehabilitation-related documents/artifacts. We held fourconsultation meetings with a neurologist and a PT, specializedin stroke rehabilitation and recovery, before starting the study toinform its design.The semi-structured interviews involved specialists at the threedifferent medical centers, and observations involved specialists whoworked in two of them. Our initial observations were conducted inone hospital located in a major city in the mid-Atlantic region ofthe United States of America, later adding a hospital in a differentsystem. All hospitals had Physical Medicine & Rehabilitation depart-ments. All of the medical centers served patients who were from (1)low socio-economic areas, (2) technologically low-resourced loca-tions, (3) and/or surrounding rural areas. Documents and artifactswere collected at both hospitals as our study progressed.
HI ’21, May 8–13, 2021, Yokohama, Japan Akinsiku, Avellino, et al.
Participants were recruited through snowball sampling within theparticipating medical centers. We interviewed four physiatrists(PHYs), five physical therapists (PTs) and seven occupational thera-pists (OTs), who work with stroke survivors in an outpatient setting,and observed a subset of those based on availability. The participantdemographics are detailed in Table 1. Participants had various spe-cialties, evident by their degrees (including M.D./Ph.D., Masters inEngineering, DPT, and MBA), as well as experience in different med-ical settings, most notably in low resource and rural communitiesoutside of the US (denoted with ** in Table 1).
Participant Experience(years) PracticeSetting DataPHY1 15 P, H, R I7, O9*,O10*PHY2 17 P, H, R I8PHY3** 16 P, H I9PHY4 6 H I13PT1** 21 P, H I4PT2 29 P, H I10,O12*PT3 3 H I14,O18PT4 5 H I15,O13*,O14*,O15*,O16,O17PT5 4 H I16PT6 7 H O3OT1 5 H I1,O4*,O5,O6,O7,O8OT2 18 P, H, R I2OT3** 2 H I3OT4** 6 P, H, R I5OT5 1y & 8m P, H I6OT6** 25 P, H I11,O11OT7 2m P I12OT8 2 H O1, O2
Table 1: Participant demographics and collected data sum-mary. ** denotes experience outside of the US. * denotes care-giver was in attendance. “I” = Interview. “O” = Observation.“P” = Private. “H” = Hospital. “R” = Research.
We observed and video/audio recorded 18stroke survivors rehabilitation sessions, to understand in-clinic re-habilitation, focusing on the information exchange between strokesurvivors, caregivers, and rehabilitation specialists as they discussrehabilitation. Each session was video recorded, from the begin-ning to the end. These sessions included specialists assessing thedexterity, spasticity, and cognitive function of stroke survivors, andsubsequently, treatments, exercises, and therapies were prescribed.The exercises and activities varied in the abilities they were attempt-ing to address: motor, strength, cognitive, robotic, and aquatic.Due to a COVID-19 state of emergency, we were unable to ob-serve further physical therapy sessions, and we were not allowedto attend virtual sessions due to H policy.
We conducted semi-structuredinterviews with 16 of the 18 participants. We were unable to com-plete interviews with PT6 and OT8 due to workflow constraintsbefore and after the observed rehabilitation sessions. For the partic-ipants we observed, interviews were conducted before and after therehabilitation session at the convenience of the specialists. Beforeobservations, interviews primarily focused on (A) gaining insightinto the current work practices and data needs of the rehabilitationspecialist and (B) eliciting their perceived needs for a telerehabili-tation system. After observations, interviews focused on validatingour interpretation of their practices and the information sharedwithin the session. For those we were unable to observe, only oneinterview was conducted.At the start of this study, the semi-structured interview protocolquestions included:(1) A: What are you trying to accomplish through the rehabili-tation evaluations?(2) A: What are you looking for when patients complete a task?(repetition, completion or form?)(3) B: What will you like to know about your patients whenthey complete activities at home?Following the first set of observations and interviews with OT1and OT2 and reflecting on our consultations, we realized the spe-cialists’ interests lay beyond information on repetition completionand form, as they provided a much broader set of information types.We thus added a new prompt, asking participants to rank four datatypes we had collected at that point according to importance.(4) B: In the order of most importance to least, what data (ac-tivity repetition, frustration, motivation, and stress) are youinterested in knowing about your patient’s health at homewhile completing rehabilitation exercises? • Activity Repetition : Completion of the prescribed repetitions,computed from sensor data. • Frustration : Level of frustration when completing a pre-scribed exercise. • Motivation : Level of motivation when completing a pre-scribed exercise. • Stress : Levels of stress due to stroke or home environment.
We collected documenta-tion provided to stroke survivors to complement our understanding,including: a patient take-home exercise packet, a brochure for pa-tients explaining stroke, a cognitive impairment assessment formcalled Montreal Cognitive Assessment (MOCA) [25], and a med-ical note template that uses the Subjective, Objective, Assessmentand Plan (SOAP) method (example from [9] in Appendix A). Wealso examined the
Nine Hole Peg Test , a quantitative assessmentthat measures finger dexterity , and a Box and Block test , that assesunilateral gross movement . xperiential Information Needed for Stroke Telerehabilitation CHI ’21, May 8–13, 2021, Yokohama, Japan We obtained IRB approval from the University of Maryland, Balti-more County Institutional Review Board, and received approval tobe on site from administrators at the partnering clinical institutions.We obtained consent from the clinician participants before the ini-tial interview. Before observations began, we gained verbal consentfrom the patient and caregiver (if present) to observe and recordthe rehabilitation session. We did not document any identifiablepatient data. The patients were not the focus of the field study;observations were primarily to get insight on the work practices ofthe rehabilitation specialists. Participants were not compensatedfor their participation.
We performed a qualitative data analysis focusing on the informa-tion used in rehabilitation and the needs for a telerehabilitationsystem. Two researchers analyzed the data, a PhD student (infor-mation scientist) with over 5 years of research and developmentexperience in HCI; and an undergraduate student in mechanicalengineering. One author also has experience as a caregiver. Wecompiled our field notes, transcribed all the observations and inter-views, then performed open-coding of all data in three iterations,refining the codes as we progressed. The two coders (first and thirdauthors) compared concepts informally as they coded, discussingdifferences in interpretation and reaching either a consensus ordocumenting their differences.After open coding, we used an inductive analysis to create cat-egories based on behaviors exhibited by the participants. Our ob-servation codes included: Personal Check In/Conversation (whenthe specialist prompts the patient), RS Describe Task(s)/Goals, StartActivity, End Activity, Rehabilitation Specialist Demonstrates Taskwith Body, Rehabilitation Specialist Asks Patient to Move a Cer-tain Body Part, and Rehabilitation Specialists Taking Notes. Af-ter an initial round of coding, we noticed that Personal CheckIn/Conversation was the most-used code. Therefore, we focused onhaving a conversation with the specialists, which turned our anal-ysis efforts towards the details on the conversation. Our final setof interview and observation codes included: Describes CaregiversInvolvement, Information Documented During Session, Interdis-ciplinary Interaction, Medication Management Conversation, andComorbidity Discussion. Lastly, the first author grouped codes andfound relations among them to create the four high-level themespresented. There was one overarching theme that described allthe data, experiential information about a stroke survivor plays animportant role in rehabilitation.
Our research study began under the assumption that the ideal tel-erehabilitation system optimally uses sensors to accurately measuremovement and present it to a rehabilitation specialist. However,we quickly challenged this assumption, as our findings revealedthat although movement data is an important component of strokerehabilitation, it is not the only information need for specialists.
Experiential information in stroke rehabilitation is information de-scribing a stroke survivor’s lived experience, collected through the patient-centered approach and used to inform the rehabilitation careplan. Note, in our field study, multiple participants (e.g., PT1, PHY2,OT6, OT7) used a synonym of patient-centered, client-centered .Ultimately, experiential information is key to rehabilitation spe-cialists’ ability to build context around the patient’s health status,and movement data. We will now detail how experiential informa-tion is built and it’s components.
Rehabilitation specialists use a patient-centered approach to sub-jectively gather experiential information to build context. The ap-proach includes strategies like conducting interviews during therehabilitation session. It is important to understand that the differ-ent specialities gather experiential information and build contextsimilarly, as they all use a patient-centered approach, but theyprescribe and perform different interventions and evaluations.We observed throughout our field study that the main job ofspecialists is not limited to objectively assessing stroke survivors,as OT2 commented during an interview: “You have to rememberthat our job is more than exercising their [patients’] arm.”
Duringrehabilitation sessions, we observed specialists used interviews tobuild context on their patients, OT3 explained: “Therapists first startwith a subjective interview. They then ask them [stroke survivor]where their life situation is currently, who is looking after them, andwhat they can do for themselves.”
The experiential information isgathered from subjective interviews and directly influence the re-habilitation plan that specialists prescribe to their patients, and alsohelps specialists navigate the nuance and complexities of the strokerehabilitation process. PHY1 said: “A number of underlying issuesaffect stroke rehabilitation. Physical, cognitive, social. A challenge ishow to distill. There isn’t always a set path, you have to play it byear, you have to get the medical history and the physical[sic]. And theapproach cannot always be solved algorithmically [sic].”
Since the process is not built algorithmically , specialists put themajority of their effort into gathering experiential information dur-ing rehabilitation sessions, when compared to gathering movementdata. Figure 1 (visualization made with Noldus ObserverXT ) showsa 14-minute snapshot of an example movement-based rehabilitationtask taken from a 45-minute session. Here, the patient was requiredto complete a Nine Hole Peg Test while the therapist observed. Ina 14-minute span, the specialist acquired experiential informationfor ∼ ∼ ∼ Subjective data (e.g., experiential information and context)and
Objective data during an appointment, to ultimately form an HI ’21, May 8–13, 2021, Yokohama, Japan Akinsiku, Avellino, et al.
Figure 1: A 14-minute visualization of how time was spentduring a rehabilitation session. The different activities areshown on the left, and bars represent their duration.
Assessment and create a
Plan . Due to patient confidentially, ourresearch team did not have access to patient medical notes, so weprovide an example patient note that was created by Susan et al. [9]using the SOAP method (Appendix A).Objective information can also become experiential informationthrough subjective contextualization. OT5 explained she puts herhealth assessment (e.g., comorbidities) under the
Subjective section,including her professional medical opinion on how this affectsphysical abilities (e.g., fatigue related to medication), as this impactsthe care plan. In an interview with PHY3, he shared with us theimportance of documenting subjective insight into his medicalnotes. PHY3 gave the example of a musculoskeletal evaluation,what the patient reports to him goes under the
Subjective section,and the results of the evaluation will go under the under
Objective . We found that factors like a stroke survivor’s motivation, stress, andfrustration levels are a top priority for specialists to help gatherexperiential information. These levels are important because theyall impact the ability for a stroke survivor to make progress on,and comply to, a care plan. We observed PHY4 explain this impactto her patient: “Things that make you feel like you’re going downhill, or not recovering as fast, is depression, stress and fatigue. All ofthose things can make your symptoms feel a lot worse.”
In fact, whenasked to rank the various types of information, 14 of the 16 par-ticipants ranked motivation, stress and fatigue as more importantthan activity repetition (Figure 2). All specialists expressed howimportant it is to gauge the patient’s motivation level, as it informsthe specialists’ approach to prescribe a motivating and satisfyingrehabilitation plan.PT1 articulated how a rehabilitation plan is not bounded toobjective assessments, but it has to be tailored to the motivationsof the patient: “If they just do it [exercise] just once a day. Well, Iam happy. What I try do is, because life is so on the go, make therehab just part of their day. Like walking around the block.”
Simply,PT1 was sharing that the rehabilitation process is about creating aplan around the situated conditions. PHY2 had similar sentimentsas PT1: “Just getting a patient to exercise is good enough for me.I do not care about the reps.”
PHY2 shared examples of situatinga care plan, such as assigning activities like walking around theneighborhood for patients that have a dog, or walking to a place
Figure 2: Heat map showing the ranking of importance of ex-periential information related to activity repetitions, stress,frustration and motivation. Visualization made using theBertifier tool [32] she knows the patient will enjoy. The importance of the rankingscan also vary because different factors dynamically change for eachpatient. PT2: “It [the order] depends on their home life, and thingslike that. Motivation is first because you don’t get that, you don’t getanything.”
PT2 shared similar sentiments, and said that her rankingwould change for one of her patient’s that has a stressful home.Stress levels are also included as experiential information be-cause they can have a direct impact on the stroke survivor’s physicalhealth, such as blood pressure. OT7 expanded on the importance ofalso understanding the source of the stress such as family stress and personal stress , thus ensuring that prescribed rehabilitation plansdo not exacerbate the patient’s stress level, blood pressure and/orheart rate levels. PT4 shared, “I tend to look at the big picture, and Imight give them a home program [rehabilitation plan] that doesn’tstress their lives too much, just so they do it. I try to prioritize qualityof life and compliance.”
She went further to expand on her role, andwhat she considers the big picture. PT4 said: “My role as a PT is toget them as independent as possible, and try to return them to funthings. For example, I tried to get him [the patient] back to golfingbecause it is a stress reliever for him and he enjoys it. I am particularlypassionate to get them back to their normal role as much as possible”.
Including stress levels as experiential information allows specialiststo gauge quality of life, and prescribe an appropriate care plan.Finally, frustration levels play an important role during pre-scribed rehabilitation exercises and tasks. While observing OT6conduct a rehabilitative exergame with her patient, she informedus about her process of managing frustration during activities: “Wehave to have an idea of the game, and then we pair it with their[patients] level of function. We give them a challenge, but not somuch that they can’t be successful, otherwise it is just frustrating.”
What we see here is the specialist paying more attention to ris-ing frustration levels than focusing on the movements themselves.Her goal is not to achieve a certain number of movements, butinstead she is looking for the right balance of effort against frus-tration. Thus, frustration levels are experiential information thatbetter inform specialists, so they prescribe attainable activities. Sim-ply, having context on what motivates, stresses, and frustrates thestroke survivor is crucial for both devising an exercise plan, as wellas assessing the survivor’s success at following the plan. It is moreimportant for the specialists to learn if there is a moderate levelof motivation, low number of stressors, and low frustration thanto gather any further information on what movement the patientactually performed. xperiential Information Needed for Stroke Telerehabilitation CHI ’21, May 8–13, 2021, Yokohama, Japan
The survivor’s acclimation after a stroke and their rehabilitation goals are considered experiential information. A survivor’s accli-mation is simply how they are coping in their home after survivinga stroke; for example what kind of accessibility or health relatedissues they are having. Goals are a set of health milestone patientswant to achieve through rehabilitation, such as returning back towork or driving for themselves. These two aspects are tightly con-nected because a survivor’s acclimation can inform the progressmade towards accomplishing a goal.To evaluate the patient’s acclimation at home, specialists askquestions to determine if their patients are successfully preformingActivities of Daily Living (ADLs). ADLs are basic daily self care tasks(e.g., managing medication, meal planning, or shopping) TypicallyOTs are the specialists that assess ADLs [18] as a data point, but PT4mentioned that she also references ADLs with her patients. PT4said, “I do blend some with OT on the ADLs. Like I work on toiletingwith patients. The OT might work on using assistive device, but I domore of the gross motor stuff. Like can you transfer yourself from thewheelchair to the toilet.”
PT4 essentially uses the patient’s experienceon ADLs as a reference to inform the type of intervention sheprescribes towards a goal.It is important to evaluate acclimation, as this helps specialistscreate or modify the care plan while staying aligned to the patient’sgoals (e.g., dressing independently). However, it is more complexthan simply asking if, when, or how exercises were completed,experiential information is required. One strategy of evaluatingand validating the acclimation is using sensor data to probe furtherabout the patients’ experience at home. PHY2 for example askssome of her patients to wear a
Fitbit , so then she can review thedata during the rehabilitation sessions. PHY4 recalled a time whena patient step count steadily declined, causing her to inquire if thechange in activity was due to a comorbidity or decreased motiva-tion, a crucial difference when evolving the care plan. Validatinghome exercises is important for specialists to determine if rehabili-tation plan goals are being met. Validation can also be performedby patients themselves through self-reporting, as OT4 explained: “Validating home exercises is a combination of objective and subjectivemeasures [..] I actually tell my patients to complete a weekly reportfor me”. Anticipating acclimation is necessary to adapt goals before it istoo late, so specialists use contextual information about a patient’sschedule. For example, during an observation, the stroke survivorand caregiver informed PT2 that the survivor was having theirfamily over later that evening for their birthday. In response, PT2reduced the exercise for her patient because she, “wanted him tosurvive tonight.”
Taking into account what the patient’s acclimationwill be later on (tired), she altered the rehabilitation plan for thatparticular session, recognizing that the goal for the evening was toenjoy time with their family. Physical and mental health are key experiential information whenprescribing a rehabilitation plan, as both impact the ability to com-plete the plan. Additionally, mental health can impact compliance.
Multiple comorbidities are common in strokesurvivors. Some examples we heard during discussions include:hypertension, diabetes, malnutrition, sleep apnea, depression, in-somnia, pseudo dementia, atrial fibrillation and vertigo. Comor-bidities require close monitoring, management, and coordinatedcare amongst the stroke survivors coordinate care amongst the carenetwork because they dictate how successfully and safely exercisesare performed. Typically, these are managed through situationalchanges, such as changing diet or minimizing stressful activities,and medication.Some of the participants, such as PHY1, PHY4 and OT1, took apatient’s vital signs at the end of an activity to check for fatigue(Figure 3), determining the situational changes for the exercise pro-gram. For example, in an observation with OT1, she noticed signs ofdiscomfort and fatigue, so she took vitals and recorded them in hermedical notes, ending the exercise early. In the followup interview,she told us that this patient’s hypertension limits their ability forcertain physical and/or cognitive interventions.
Figure 3: Specialist taking the patient’s blood pressure afteran activity.
Specialists have a particular interest in under-standing cognitive abilities and depression, when inquiring aboutmental health. To assess level of cognitive ability, we observed howPHY4 and OT1 refer to the MOCA. PHY1 and PT3 later explainedduring interviews that such assessment is important in determiningwhat the patient can do, thus impacting the care plan, but also inunderstanding the patient’s own ability to comprehend the exer-cises. In the cases where cognitive impairments hinder compliance,the specialists will often coordinate with a caregiver, so they canassist the stroke survivor with completing prescribed interventionswhile at home.
HI ’21, May 8–13, 2021, Yokohama, Japan Akinsiku, Avellino, et al.
Depression in particular can impact the patient’s motivationto comply with the prescribed rehabilitation plan while at home.In our observations, we observed physiatrists (PHY1 and PHY4)speak extensively with stroke survivors and their caregivers aboutthe survivors’ ongoing battle with depression, and then prescribedinterventions. In the observation with PHY4, she modified thepatient’s rehabilitation plan by first reaching out directly to thepatient’s neuropsychologist to discuss and coordinate a responseto the patient’s bout with depression.
Many specialists considered the rehabilitation process as a teameffort, which includes the caregiver. Caregivers regularly assess thepatient, which becomes important experiential information. Theyare deeply involved with multiple aspects of a stroke survivor’s life,they can be a family member, friend, or hired professional. Theirrole is so important in the survivor’s recovery process that PT3went as far as saying, “I think caregiver support is a major predictorinto how much someone can recover. Fortunately, they can offer sup-port, encourage [them], [overcome] cognitive deficits, [monitor] theirschedule, [help with] exercise, and [cook] meals. The biggest area iscompliance.”
Caregivers play an important role in the care plan, both facilitat-ing and validating the plan. For example, in a session with PHY4the patient self-reported that they felt they were speaking muchbetter. However, the caregiver felt differently, and provided a morein-depth assessment of the progress: the patient’s speech indeedhas improved, but slowly began to decline in the weeks leading upto the appointment. PT4 expanded on the value of this assessment: “It is pretty nice to get the caregiver’s perspective, because they’re alittle bit more honest than the stroke survivor. I think they are reallyimportant, because they encourage the patient to get better, and theyare the key for their patient live their life again. I like them to be inthe therapy session.”
Having the caregiver present during a rehabilitation session ishighly valued because they provide insight on the patient’s experi-ence at home and offer assistance. We observed caregivers provid-ing assistance to the specialists (e.g., supporting the stroke whilewalking), including insight when the specialists are completing anassessment. OT4 elaborated on how the experiential informationfrom the caregiver complements the patient information: “Some-times people [rehabilitation specialists] will give out an assessment tothe patients and caregivers. That way, we will get the patient’s insighton how they performed different ADLs, and we will get the familymembers’ insight on how they felt the patient performed the ADLs.”
In this situation, the family member is reporting not on the exercise execution , but on how the patient is faring in their daily lives—whatthey are able to do and how they are doing it. This constitutes keyexperiential information.
We found that experiential information is essential in co-locatedstroke rehabilitation for rehabilitation specialists, and that they usea large variety of information other than exercise movement data.However, the focus of home-based therapy systems is on movementdata, as they aim at motivating (e.g., Us’em [7]) and monitoringpatients (e.g., ArmSleeve [34]). This means that they (1) presentmovement data to specialists upfront and center, (2) have not en-abled capturing experiential information of the movement data, and(3) do not provide means for annotating movement data for context.This has led to a paradigm that puts emphasis on movement data(Figure 4 - left), and, as a consequence this legacy has led telerehabil-itation systems to focus on the computational work for recognizingmovement and quantifying it for remote specialists to visualize(e.g., TeleREHA [33], mRes [42], Rehabilitative TeleHealthCare [35],and systems covered by Santayayon et al. [37]). Our study showsthat this paradigm is incongruent with how specialists actuallywork in face-to-face stroke rehabilitation. This does not invalidatethe motivation in the current paradigm to track movement andcount repetitions accurately for telerehabilitation. Instead, we positthat movement data needs to exist within a more sophisticated un-derstanding of patients’ experiential information, when designingtelerehabilitation systems.
Previous work hints towards the need to capture information be-yond movement. For example, How et al. [16] stated that successfulrehabilitation systems should adapt to surrounding life context, andPloderer et al. [34] recognized that lack of contextual informationhinders the interpretation of movement data. The two studies how-ever do not go into detail defining or modeling what is “context”information beyond movement exercise data. Through our work,we studied such information needs, and this has enabled us to moveforward the paradigm through which systems are designed. We pro-pose a paradigm reconceptualization for situated telerehabilitationsystems (Figure 4 - right ), that capture the experiential informationused in rehabilitation: a stroke survivor’s (1) motivation, stress, andfrustration levels, (2) acclimation and goals, (3) their health status,and (4) the caregiver’s assessment; where exercise movement datacomplements the experiential information through a semantic layerthat gives meaning to otherwise incomplete data. In this way, thesharing of a stroke survivor’s lived experience within their care net-work is now taken into consideration. This brings telerehabilitationsystem design closer to how co-located rehabilitation takes place,by integrating the patient-centered approach [12] and giving mean-ing to movement data by incorporating experiential informationon the stroke survivor’s lived experience. xperiential Information Needed for Stroke Telerehabilitation CHI ’21, May 8–13, 2021, Yokohama, Japan
Figure 4: Current Paradigm for Telerehabilitation Systems
Below we provide design implications to capture context by lever-aging our findings that identified types of experiential information.
For telerehabilitation systems to be congruent withthe existing rehabilitation practices, they need to provide ways tocapture and share various forms of experiential information. Moti-vation, Stress, and Frustration. The insight we received fromspecialists during the interviews (e.g., Figure 2) reveal the impor-tance of motivation, stress, and frustration, and that keeping trackof the factors on a regular basis can help trace the cause for changesin exercise movement and daily activity. Capturing a patient’s mo-tivation, stress and frustration levels is essential for specialists tobuild life context for their patients. Capturing the three could looklike a system autonomously prompting questions to the patient orcaregiver to answer when deviations in any of the three factorsare detected. Capturing this information in the moment would al-low contextualization of the patient’s exercise/movement data andoverall experience. The prompts will ideally elicit responses thatinform the specialists the reason behind the deviation. Additionally,specialists could predetermine alternative exercises and activitiesthey know are engaging for a particular patient, and the systemcould automatically propose the appropriates activity to respondto the contextual reasons underlying change in movement.Acclimation & Goals. Telerehabilitation systems would ben-efit from including features that capture acclimation and goals tohelp specialists evaluate patients’ progress. How et al. [16] brieflydiscussed how there may be better ways in which telerehabilitationsystems can seek to complement rehabilitation goals of traumaticbrain injury (e.g., stroke), but did not offer concrete design sug-gestions. To benefit acclimation and goal tracking, systems canimplement a feature that allows patients to create a profile or vir-tual diary that acts as a virtual acclimation report. This report wouldcapture how a patient is getting acclimated to a new environmentand everyday tasks, that is accessible to their care networks. Forexample, data from a smart watch or
Fitbit could be used to popu-late a home acclimation report instead of being used simply as a measure of exercise. The
Fitbit movement data can be tagged eitheras performing prescribed exercises or simply ADLs in the report.In this way, an event such as a decrease in tagged ADL movementdata would signal that the survivor is probably not acclimatingwell. Additionally, the specialist would use this report to discussthe patient’s goals immediately or during a session.Physical Health. There have been examples of systems thatcollect exercise movement data for physical health, TeleREHA [33]monitors arm reach and mRes [42] monitors wrist function. How-ever, our study showed that physical health is not limited to exercisemovement data. A telerehabilitation system that acquires both phys-iological and exercise movement data brings additional benefitswhen working with survivors with comorbidities (e.g., diabetes andhypertension). By periodically checking heart rate (e.g., via smartwatch) and or blood pressure (e.g., via a built-in blood pressuremonitor), a system can alert the appropriate specialist when theselevels reach concerning thresholds in combination with movementdata during rehabilitation exercises. This would enable specialiststo intervene remotely and modify the rehabilitation plan based onaccurate information, for example suggesting a rest. In a synchro-nous telerehabilitation system, the specialist might suggest a rest.In an asynchronous system, the data could be used to start discus-sion at a later time. Rehabilitative TeleHealthCare [35] was the onetelerehabilitation system we found that aligned with our finding tocombine physiological data and movement data. TeleHealthCareuses sensors to capture physiological data (heart rate, oxygen satu-ration (SpO2) and respiration rate) and movement data for healthmonitoring. However, this system does not take into account whatare the patient comorbidites (e.g., Diabetes and Atrial Fibrillation),and thus limits how an expert can use these data combined.Mental Health. As our study showed, mental health is an-other aspect that impacts the creation of a rehabilitation plan, andits compliance given a patient’s cognitive function. A system thatleverages existing standardized assessments (e.g., MOCA) to collectdata will better match current rehabilitation we observed. To keeptrack of the response evolution of such assessments, a telerehabili-tation system can be used to require patients to complete cognitive
HI ’21, May 8–13, 2021, Yokohama, Japan Akinsiku, Avellino, et al. assessments when prescribed by the specialist throughout time.It is important that systems collect, keep track of, and transmitmental health status of the patient to specialists, as this informationis an experiential data point used to contextualize deviations inexercise movement data, along with motivation stress, and frus-tration levels. Moreover, it is particularly important to considerthe collaboration among the care network when it comes to healthdata and sticking to existing workflows. As Ng et al. [29] recentlystressed, health care providers are concerned about how to adjusttheir practices, to better guide the use of sensor data within mentaltelehealth. By keeping all the care network updated with mentalhealth information, a telehealth system can enable the care networkto collaboratively interpret the collected exercise movement andmental health data, avoiding misinterpretation, but also allowingfor shared decision making when it comes to adjusting how sensordata is used.Caregiver Assessment. As we learned, caregivers provide cru-cial support for a successful execution of a rehabilitation plan athome, and thus are the ones that hold key information about theday-to-day issues around the plan. However, no system to ourknowledge considers incorporating information from a caregiver.Telerehabilitation systems should incorporate functionalities thatcapture caregivers’ insights around the execution of exercises, sothat specialists can compare and contrast this view with both thepatient’s view on their own execution, and the movement data.More importantly, they can act as surrogates for filling reports orreporting issues, when the stroke survivor is not able to do so, suchas when the survivor’s health is negatively impacted by a comor-bidity. In this way, the rehabilitation specialist can better evaluatethe health progress of the stroke survivor.
One of our main insights is that specialists create meaningsubjectively. We observed strategies around creating meaning ontop of movement data, such as reviewing
Fitbit data together with aspecialist. Ploderer et al. [34] concluded in their study that a majorlimitation of
ArmSleeve is the lack of contextual information pre-sented across the different dashboard pages; and Mentis et al. [23]showed that clinicians require patient contextual information tomake sense of
Fitbit data—for instance if a high walking day wasdue to a vacation and a low walking day was due to a poor medica-tion reaction. Telerehabilitation systems should have a dedicatedsemantic information layer for all the information stored. It is likelythat different specialists have a unique perspective on data andwant to annotate different data (e.g., the data related to their ownprofession). Having these annotations can lead to less confusionabout what other specialists are doing, how they deal with issues,and most importantly, it will facilitate coordination.
Re-lated to our last point, we believe that telerehabilitation systemshave the potential to be the central hub of information aroundrehabilitation. We observed the level of care coordination involvedbetween those in the stroke survivor’s care network. Therefore,it is important that telerehabilitation systems provide functionali-ties that best facilitate the communication and coordination of thedifferent specialists by allowing them to access each other’s data,add their own interpretation, and thus coordinate their actions. We assert this can play a major role in the care plan, as this is the oneartifact where all the actors have an influence, making it dynamic.Accommodating for dynamic, always up-to-date care plans withtransparent decision making imprinted on the plan itself (throughannotations), is a way to keep care consolidated.
Our work is limited to function and mobility rehabilitation special-ists within the Physical Medicine & Rehabilitation specialty. Thiswork does not include the other specialists that are involved in therehabilitation processes of stroke survivors or other illnesses thatrequire rehabilitation. Now that we have working relationshipswith our partners, we will conduct observations and interviewswith more rehabilitation specialists. We will also begin elicitingfeedback from additional stakeholders (stroke survivors and care-givers) on our paradigm refocus. Ultimately, with an eye to begindeveloping a prototype of a situated telerehabilitation system.
Our research reveals that experiential information is an essentialneed in stroke rehabilitation. This paper provides a more holisticview of the practices of rehabilitation specialists within the Physi-cal Medicine & Rehabilitation medical specialty. Additionally, thispaper proposes a paradigm reconceptualization in stroke telereha-bilitation system development to address the complex and dynamicnature surrounding stroke rehabilitation. We posit that our pro-posed refocus can lead to the development of telerehabilitationsystems that will be on par with the type of interactions and evalu-ations that exist in a face-to-face stroke rehabilitation session. Wedo not suggest that sensors have no role in telerehabilitation, butwe say to our research community, that the exercises are a meansto the end, and it is important to understand what the end is.
ACKNOWLEDGMENTS
We would like to thank the patients and medical personnel thatparticipated in our study. We would like to thank Dr. Wittenbergfor their insight and assistance. This work has been supported inpart by NIH/NIGMS R25 GM055036 IMSD Meyerhoff GraduateFellows Program, NSF CAREER award IIS-1552837 and SaTC awardCNS-1714514, NIH/NIGMS MARC U*STAR T34 08663 NationalResearch Service Award, and by Dean Keith J Bowman and theUMBC Constellation Professorship. xperiential Information Needed for Stroke Telerehabilitation CHI ’21, May 8–13, 2021, Yokohama, Japan
REFERENCES [1] Imad Afyouni, Anas Einea, and Abdullah Murad. 2019. Towards an AdaptiveGaming Framework for TeleRehabilitation. In
Proceedings of the 12th ACM Inter-national Conference on PErvasive Technologies Related to Assistive Environments(PETRA ’19) . Association for Computing Machinery, New York, NY, USA, 305–306.https://doi.org/10.1145/3316782.3321545[2] G Alankus and C Kelleher. 2012. Reducing compensatory motions in videogamesfor stroke rehabilitation.
Proceedings of the 2012 ACM Annual Conference onHuman Factors in Computing Systems.
Proceedings of the SIGCHIConference on Human Factors in Computing Systems (CHI ’10)
PLoS ONE
14, 11 (2019),1–18. https://doi.org/10.1371/journal.pone.0225150[6] Emelia J Benjamin, Michael J Blaha, Stephanie E Chiuve, Mary Cushman,Sandeep R Das, Rajat Deo, J Floyd, M Fornage, C Gillespie, C R Isasi, and Others.2017. Heart disease and stroke statistics-2017 update: a report from the AmericanHeart Association.
Circulation
CHI ’11 Extended Abstracts on Human Factors inComputing Systems (CHI EA ’11) . ACM, New York, NY, USA, 1279–1284. https://doi.org/10.1145/1979742.1979761[8] David Brennan, Lyn Tindall, Deborah Theodoros, Janet Brown, Michael Campbell,Diana Christiana, David Smith, Jana Cason, and Alan Lee. 2010. A Blueprint forTelerehabilitation Guidelines.
International Journal of Telerehabilitation
2, 2 (oct2010), 31–34. https://doi.org/10.5195/IJT.2010.6063[9] Susan Cameron and Imani Turtle-Song. 2002. Learning to Write Case Notes Usingthe SOAP Format.
Journal of Counseling & Development
80, 3 (2002), 286–292.https://doi.org/10.1002/j.1556-6678.2002.tb00193.x[10] M G H Dekker-van Weering and M M R Vollenbroek-Hutten. 2015. Developmentof a Telemedicine Service That Enables Functional Training for Stroke Patientsin the Home Environment. In
Proceedings of the 3rd 2015 Workshop on ICTs forImproving Patients Rehabilitation Research Techniques (REHAB ’15) . ACM, NewYork, NY, USA, 109–112. https://doi.org/10.1145/2838944.2838971[11] S. R. Egglestone, L. Axelrod, T. Nind, R. Turk, A. Wilkinson, J. Burridge, G.Fitzpatrick, S. Mawson, Z. Robertson, A. M. Hughes, K. H. Ng, W. Pearson, N.Shublaq, P. Probert-Smith, I. Rickets, and T. Rodden. 2009. A design frameworkfor a home-based stroke rehabilitation system: Identifying the key components.In . Institute of Electrical and Electronics Engineers, Piscataway, NJ, 1–8.https://doi.org/10.4108/ICST.PERVASIVEHEALTH2009.6049[12] M. Gerteis, S. Edgman-Levitan, Picker/Commonwealth Program for Patient-Centered Care, J. Daley, and T.L. Delbanco. 1993.
Through the Patient’s Eyes:Understanding and Promoting Patient-Centered Care . Wiley, United States.[13] Jin Guo, Ted Smith, David Messing, Ziying Tang, Sonia Lawson, and Jinjuan HeidiFeng. 2015. ARMStrokes: A Mobile App for Everyday Stroke Rehabilitation. In
Proceedings of the 17th International ACM SIGACCESS Conference on Computers&; Accessibility (ASSETS ’15) . Association for Computing Machinery, New York,NY, USA, 429–430. https://doi.org/10.1145/2700648.2811337[14] Paul A Heidenreich, Justin G Trogdon, Olga A Khavjou, Javed Butler, KathleenDracup, Michael D Ezekowitz, Eric Andrew Finkelstein, Yuling Hong, S ClaiborneJohnston, Amit Khera, Donald M Lloyd-Jones, Sue A Nelson, Graham Nichol,Diane Orenstein, Peter W F Wilson, and Y Joseph Woo. 2011. Forecasting theFuture of Cardiovascular Disease in the United States.
Circulation
Frontiers in Bioengineering and Biotechnology
Disability and Rehabilitation: Assistive Technology
12, 3 (2017),244–261. https://doi.org/10.3109/17483107.2015.1129457[17] Tim Johansson and Claudia Wild. 2011. Telerehabilitation in stroke care – asystematic review.
Journal of Telemedicine and Telecare
17, 1 (jan 2011), 1–7.https://doi.org/10.1258/jtt.2010.100105[18] Sidney Katz. 1983. Assessing Self-maintenance: Activities of Daily Living,Mobility, and Instrumental Activities of Daily Living.
Journal of the American Geriatrics Society
31, 12 (1983), 721–727. https://doi.org/10.1111/j.1532-5415.1983.tb03391.x arXiv:https://onlinelibrary.wiley.com/doi/pdf/10.1111/j.1532-5415.1983.tb03391.x[19] Kate E. Laver, Daniel Schoene, Maria Crotty, Stacey George, Natasha A. Lannin,and Catherine Sherrington. 2013. Telerehabilitation services for stroke.
CochraneDatabase of Systematic Reviews
Proceedings of the 24th International Conference on IntelligentUser Interfaces (IUI ’19) . Association for Computing Machinery, New York, NY,USA, 218–228. https://doi.org/10.1145/3301275.3302273[21] Chee Kian Lim, I. Ming Chen, Zhiqiang Luo, and Song Huat Yeo. 2010. A lowcost wearable wireless sensing system for upper limb home rehabilitation, In2010 IEEE Conference on Robotics, Automation and Mechatronics.
Closing the Quality Gap: A Critical Analysis of Quality Improvement Strategies(Vol. 7: Care Coordination) . Agency for Healthcare Research and Quality U.S.Department of Health and Human Services, Rockville (MD).[23] Helena M. Mentis, Anita Komlodi, Katrina Schrader, Michael Phipps, Ann Gruber-Baldini, Karen Yarbrough, and Lisa Shulman. 2017. Crafting a view of self-tracking data in the clinical visit.
Conference on Human Factors in ComputingSystems - Proceedings
Proceedings of the 33rd AnnualACM Conference on Human Factors in Computing Systems (CHI ’15) . Associationfor Computing Machinery, New York, NY, USA, 511–520. https://doi.org/10.1145/2702123.2702342[25] Ziad S. Nasreddine, Natalie A. Phillips, Valerie Bedirian, Simon Charbonneau,Victor Whitehead, Isabelle Collin, Jeffrey L. Cummings, and Howard Chertkow.2005. The Montreal Cognitive Assessment, MoCA: A Brief Screening Tool ForMild Cognitive Impairment.
Journal of the American Geriatrics Society
53, 4 (apr2005), 695–699. https://doi.org/10.1111/j.1532-5415.2005.53221.x[26] National Institutes of Health Clinical Center. 2017. About Occupational Therapy.Retrieved January 9, 2021 from https://clinicalcenter.nih.gov/rmd/ot/index.html[27] National Institutes of Health Clinical Center. 2017. About Physical Medicine andRehabilitation. Retrieved January 9, 2021 from https://clinicalcenter.nih.gov/rmd/pmr/index.html[28] National Institutes of Health Clinical Center. 2019. About Physical Therapy.Retrieved January 9, 2021 from https://clinicalcenter.nih.gov/rmd/pt/index.html[29] Ada Ng, Rachel Kornfield, Stephen M. Schueller, Alyson K. Zalta, Michael Bren-nan, and Madhu Reddy. 2019. Provider Perspectives on Integrating Sensor-Captured Patient-Generated Data in Mental Health Care.
Proceedings of theACM on Human-Computer Interaction
3, CSCW, Article 115 (Nov. 2019), 25 pages.https://doi.org/10.1145/3359217[30] Institute of Medicine. 2001.
Shaping the Future; Crossing the quality chasm: a newhealth system for the 21th century . The National Academies Press, Washington,DC. https://doi.org/10.17226/10027[31] Bruce Ovbiagele, Larry B. Goldstein, Randall T. Higashida, Virginia J. Howard,S. Claiborne Johnston, Olga A. Khavjou, Daniel T. Lackland, Judith H. Lichtman,Stephanie Mohl, Ralph L. Sacco, Jeffrey L. Saver, and Justin G. Trogdon. 2013.Forecasting the future of stroke in the united states: A policy statement from theAmerican heart association and American stroke association.
Stroke
44, 8 (2013),2361–2375. https://doi.org/10.1161/STR.0b013e31829734f2[32] C. Perin, P. Dragicevic, and J. Fekete. 2014. Revisiting Bertin Matrices: New In-teractions for Crafting Tabular Visualizations.
IEEE Transactions on Visualizationand Computer Graphics
20, 12 (2014), 2082–2091. https://doi.org/10.1109/TVCG.2014.2346279[33] J. C. Perry, J. A. Ruiz-Ruano, and T. Keller. 2011. Telerehabilitation: Toward acost-efficient platform for post-stroke neurorehabilitation. In . Institute of Electrical and ElectronicsEngineers, Piscataway, NJ, 1–6. https://doi.org/10.1109/ICORR.2011.5975413[34] Bernd Ploderer, Justin Fong, Anusha Withana, Marlena Klaic, Siddharth Nair,Vincent Crocher, Frank Vetere, and Suranga Nanayakkara. 2016. ArmSleeve:A Patient Monitoring System to Support Occupational Therapists in StrokeRehabilitation. In
Proceedings of the 2016 ACM Conference on Designing InteractiveSystems (DIS ’16) . Association for Computing Machinery, New York, NY, USA,700–711. https://doi.org/10.1145/2901790.2901799[35] G. Postolache, C. M. Moura, P. S. Girão, and O. Postolache. 2011. RehabilitativeTeleHealthCare for post-stroke outcome assessment. In . Institute of Electrical and Electronics Engineers, Piscataway, NJ,408–413. https://doi.org/10.4108/icst.pervasivehealth.2011.246141
HI ’21, May 8–13, 2021, Yokohama, Japan Akinsiku, Avellino, et al. [36] Marco Rogante, Dahlia Kairy, Claudia Giacomozzi, and Mauro Grigioni. 2015.A quality assessment of systematic reviews on telerehabilitation: what doesthe evidence tell us?
Annali dell’Istituto superiore di sanità
51, 1 (2015), 11–18.https://doi.org/10.4415/ANN_15_01_04[37] Mullika Santayayon, Jagkapong Pipitpukdee, and Wantanee Phantachat. 2012.Wii-Habilitation in Stroke Patients: A Systematic Review. In
Proceedings of the6th International Conference on Rehabilitation Engineering &; Assistive Technology(i-CREATe ’12) . Singapore Therapeutic, Assistive & Rehabilitative Technologies(START) Centre, Midview City, SGP, Article 33, 4 pages.[38] Fred S. Sarfo, Uladzislau Ulasavets, Ohene K. Opare-Sem, and Bruce Ovbiagele.2018. Tele-Rehabilitation after Stroke: An Updated Systematic Review of theLiterature.
Journal of Stroke and Cerebrovascular Diseases
27, 9 (2018), 2306–2318.https://doi.org/10.1016/j.jstrokecerebrovasdis.2018.05.013[39] Huidi Tchero, Maturin Tabue Teguo, Annie Lannuzel, and Emmanuel Rusch. 2018.Telerehabilitation for stroke survivors: Systematic review and meta-analysis.
Journal of Medical Internet Research
20, 10 (2018), 1–10. https://doi.org/10.2196/10867[40] Mirella Veras, Dahlia Kairy, Marco Rogante, Claudia Giacomozzi, and SilviaSaraiva. 2017. Scoping review of outcome measures used in telerehabilitation andvirtual reality for post-stroke rehabilitation.
Journal of telemedicine and telecare
23, 6 (jul 2017), 567–587. https://doi.org/10.1177/1357633X16656235[41] Lawrence L Weed. 1964. Medical records, patient care, and medical education.
Irish Journal of Medical Science (1926-1967)
39, 6 (1964), 271–282. https://doi.org/10.1007/BF02945791[42] Patrick Weiss, Marcus Heldmann, Alexander Gabrecht, Achim Schweikard,Thomas M Münte, and Erik Maehle. 2014. A Low Cost Tele-rehabilitation Devicefor Training of Wrist and Finger Functions After Stroke. In
Proceedings of the 8thInternational Conference on Pervasive Computing Technologies for Healthcare (Per-vasiveHealth ’14) . ICST (Institute for Computer Sciences, Social-Informatics andTelecommunications Engineering), ICST, Brussels, Belgium, Belgium, 422–425.https://doi.org/10.4108/icst.pervasivehealth.2014.255331
A SAMPLE SOAP NOTE
In Figure 5, the medical specialist documented the patients com-ments (
Subjective ) and physical actions (
Objective ) to inform theassessment and prescribed plan.) to inform theassessment and prescribed plan.