The links between disability, activity limitation and participation restriction are well established. Recent and continued advancement of technology, particularly smart home and communication technologies, presents new ways in which some of the limitations and restrictions experienced by people with disabilities can be overcome. The aim of this scoping review was to explore the impact of smart home and communication technology on the outcomes of people with disabilities and complex needs.
This review involved systematic searching of four databases, hand searches and data extraction. Eligibility criteria included  participant outcomes of  technology used within the home  among adults with a disability and complex needs.
Of the 2400 studies identified, 21 met our inclusion criteria. Studies were characterized by significant diversity in relation to disability and type of technology. Overall, technology appeared to improve independence, participation and quality of life among people with a disability and complex needs. Despite this, ethical considerations were raised given the vulnerability of this population, including potential risks through social participation and privacy concerns of using monitoring technology.
Smart home and communication technology can improve outcomes for people living with disabilities and complex needs. However, a number of factors impact the successful implementation of technology, including personalization, flexibility and ongoing support to the person with a disability and their close others. Future research should utilize high-quality study designs and established measures of important outcomes for this group.
IMPLICATIONS FOR REHABILITATION
There is a broad range of smart home and communication technology devices and systems available that may support the independence and participation of people with disabilities and complex needs; however, high-quality evidence documenting the impact of technology is lacking.
Soft-technology supports, including assessment, training and evaluation of technology implementation, may play just as important a role in shaping outcomes as the technology itself.
Systematic research is required to ensure there is quality evidence to inform investment in both technologies, and the soft-technology supports that promote its successful use.
Clinical practice typically emphasizes active involvement during therapy. However, traditional approaches can offer only general guidance on the form of involvement that would be most helpful to recovery. Beyond assisting movement, robots allow comprehensive methods for measuring practice behaviors, including the energetic input of the learner. Using data from our previous study of robot-assisted therapy, we examined how separate components of mechanical work contribute to predicting training outcomes.
Stroke survivors (n = 11) completed six sessions in two-weeks of upper extremity motor exploration (self-directed movement practice) training with customized forces, while a control group (n = 11) trained without assistance. We employed multiple regression analysis to predict patient outcomes with computed mechanical work as independent variables, including separate features for elbow versus shoulder joints, positive (concentric) and negative (eccentric), flexion and extension.
Our analysis showed that increases in total mechanical work during therapy were positively correlated with our final outcome metric, velocity range. Further analysis revealed that greater amounts of negative work at the shoulder and positive work at the elbow as the most important predictors of recovery (using cross-validated regression, R2 = 52%). However, the work features were likely mutually correlated, suggesting a prediction model that first removed shared variance (using PCA, R2 = 65–85%).
These results support robotic training for stroke survivors that increases energetic activity in eccentric shoulder and concentric elbow actions.
Assistance is often provided to aid limb movement during the rehabilitation process of stroke survivors. Many clinical researchers agree that active participation enhances recovery, and the goal of therapy should be to maximize “involvement” [1, 2]. Too much assistance can actually discourage patient effort . However, measurement of the degree to which patients are actually active is often difficult. Advances in rehabilitation devices allow for the measurement of forces and motion to better monitor patient activity. Here we investigate how upper limb mechanics during training relate to recovery.
Current tools for measuring physical activity during therapy offer limited information for describing interaction with the external environment or agent. While studies have shown that the intensity of therapy influences patient improvement, researchers have relied on simple metrics related to experimental conditions (e.g. movement repetitions, time-on-task, and therapy dosage) [4, 5]. More sophisticated tools have been used to directly measure energetic contributions during therapy, such as oxygen consumption devices to measure metabolic cost  or electromyography to measure muscle activity [7, 8]. However, such measures do not account for the time-varying force-motion relationships that occur during assisted movement. Robots easily measure both kinematic and kinetic variables facilitating the computation of energetic contributions in terms of mechanical power and work.
While energetic descriptions of movement have been widely studied, it has mainly focused on cyclic  or sustained movements, such as walking. Researchers have computed work and power to characterize normal and abnormal gait patterns [10, 11], to evaluate robot-assisted locomotion , and to reduce energetic costs when using exoskeletons . Recently our work has focused on robotic augmentation of upper limb dynamics to facilitate vigorous movement during practice [14, 15]. We showed that stroke survivors increase total work output during force training . Our intervention was fundamentally different than many previous strategies in that patients trained over a broader range of movements. In contrast to reaching studies [17, 18], such self-directed exploration allows for the examination of how energetics might depend on different force and motion states.
To better evaluate the variation in patient energetics, we believe more comprehensive measures are required beyond total expenditure of power or work. Researchers have also examined compartmentalized work and power measures in normal limb behaviors, for example, associating magnitudes of mechanical energy (e.g. positive/concentric and negative/eccentric work) with movement actions (e.g. flexion and extension) at individual joints . Motor impairments due to stroke are also typically described in the context of motor actions of the limb. For example, stroke survivors exhibit abnormal flexion and extension synergies  and alterations in concentric and eccentric muscle contractions [21, 22]. As such, impairments can be associated with subcomponents of work and power. As patients interact differently in response to forces, subcomponents of work and power could reveal individual differences in involvement.
An emerging trend in rehabilitation is to identify certain factors that predict individual improvement in response to therapy. Researchers have identified patient biomarkers (impairment level, neurophysiological) correlated to patient outcomes providing better recommendations for therapy [23,24,25]. Similarly, our goal is to determine if particular types of work are more important to patient recovery. Such evaluation could inform decisions on design strategies and optimize assistance to each individual. In contrast to previous studies which have relied on independent analyses of many individual predictors, our analysis goal necessitates more rigorous statistical methods to deal with potentially related work features. One possible solution is to employ multiple regression analysis which can identify features most important for prediction.
In this paper, we investigate how the energetic contributions of stroke survivors during robot-assisted training relate to upper limb recovery. We employ well-established methods of inverse dynamics to estimate the torques generated by each patient during self-directed motor exploration training with customized forces. These methods conveniently allow us to quantify the energetic involvement of each individual joint in terms of mechanical work. We then use multiple regression analysis to identify which components of work are most important for predicting recovery. We hypothesize that positive work (concentric) in elbow extension is the best predictor of outcome. This study provides a key preliminary step towards evaluating energetic descriptions of patient involvement which can inform methods for upper limb robotic therapy practice.
This investigation considered data collected from a previous study that featured 22 stroke survivors . The main inclusion criteria included: 1) chronic stroke (8+ months post-stroke) 2) hemiparesis with moderate to severe arm impairment measured by the upper extremity portion of the Fugl-Meyer Assessment (UEFM score of 15–50) 3) primary cortex involvement. Each individual gave informed consent in accordance with the Northwestern University Institutional Review Board (IRB).
Experiment participants were asked to operate a two-degree of freedom robotic device with the affected arm (Fig. 1a). A custom video display system (not shown) provided visual feedback of the location of the wrist as the arm moved in the horizontal plane. During movement, the weight of the arm was supported. Movement data was collected at 200 Hz and filtered using a 5th order Butterworth low pass filter with a 12 Hz cutoff. Using anthropometric measurements recorded from each participant, we computed inverse kinematic relationships to obtain elbow and shoulder joint angles corresponding to endpoint position data. The robot control and instrumentation were mediated by a Simulink-based XPC Target computer, with a basic rate of 1 kHz. The robot controller compensated for the dynamics of the robot arm. A force sensor attached to the end-effector measured the human-robot interaction forces.
The aims of this study were to describe patterns and dose of rehabilitation received following stroke and to investigate their relationship with outcomes.
This was a prospective observational cohort study.
A total of seven public hospitals and all subsequent rehabilitation services in Queensland, Australia, participated in the study.
Participants were consecutive patients surviving acute stroke between July 2016 and January 2017.
We tracked rehabilitation for six months following stroke and obtained 90- to 180-day outcomes from the Australian Stroke Clinical Registry.
Dose of rehabilitation – time in therapy by physiotherapy, occupational therapy and speech pathology; modified Rankin Scale (mRS)- premorbid, acute care discharge and 90- to 180-day follow-up.
We recruited 504 patients, of whom 337 (median age = 73 years, 41% female) received 643 episodes of rehabilitation in 83 different services. Initial rehabilitation was predominantly inpatient (260/337, 77%) versus community-based (77/337, 21%). Therapy time was greater within inpatient services (median = 29 hours) compared to community-based (6 hours) or transition care (16 hours). Median (Quartile 1, Quartile 3) six-month cumulative therapy time was 73 hours (40, 130) when rehabilitation commenced in stroke units and continued in inpatient rehabilitation units; 43 hours (23, 78) when commenced in inpatient rehabilitation units; and 5 hours (2, 9) with only community rehabilitation. In 317 of 504 (63%) with follow-up data, improvement in mRS was most likely with inpatient rehabilitation (OR = 3.6, 95% CI = 1.7–7.7), lower with community rehabilitation (OR = 1.6, 95% CI = 0.7–3.8) compared to no rehabilitation, after adjustment for baseline factors.
Amount of therapy varied widely between rehabilitation pathways. Amount of therapy and chance of improvement in function were highest with inpatient rehabilitation.
Post-hospital residential brain injury rehabilitation outcomes research is a complicated undertaking because of the custom-tailoring of interventions needed to meet the complex and unique need of each individual. As such, there tends to be great variability across program settings, which generally limits large-scale intervention studies. Growing literature demonstrates that post-hospital residential programs are beneficial. The main criticisms of this work include the absence of randomized-controlled studies, lack of clear definition of treatment types/settings, and small sample sizes. This study is a retrospective analysis of program evaluation data for a large, multi-site, national provider of post-hospital residential brain injury rehabilitation services. Specifically, outcome of participants completing Intensive Residential Rehabilitation (IRR) were compared to participants in the Residential Supported Living (RSL) program. Results demonstrate that participants in the IRR program improve and that participants in the RSL group preserve functional ability over time, suggesting that each program is effective in achieving its intended outcome. The IRR treatment group achieved significantly better outcomes than those in the same setting not receiving the intervention. To isolate treatment effects of IRR, a subsample of participants across program types were matched on time post-injury, age, and sex. The treatment effect of IRR was strengthened in this analysis, suggesting that chronicity alone does not account for the variance between the two groups.
Study investigated the effects of an 8-week rehabilitation exercise program combined with soymilk ingestion immediately after exercise on functional outcomes in chronic stroke patients.
Twenty-two stroke patients were randomly allocated to either the soymilk or the placebo (PLA) group and received identical 8-weeks rehabilitation intervention (3 sessions per week for 120 minutes each session) with corresponding treatment beverages. The physical and functional outcomes were evaluated before, during, and after the intervention. The 8-week rehabilitation program enhanced functional outcomes of participants.
The immediate soymilk ingestion after exercise additionally improved hand grip strength, walking speed over 8 feet, walking performance per unit lean mass, and 6-Minute Walk Test performance compared with PLA after the intervention. However, the improvements in the total score for Short Physical Performance Battery and lean mass did not differ between groups.
This study demonstrated that, compared with rehabilitation alone, the 8-week rehabilitation program combined with immediate soymilk ingestion further improved walking speed, exercise endurance, grip strength, and muscle functionality in chronic stroke patients.
Accuracy in measuring function related to one’s ability to work is central to public confidence in a work disability benefits system. In the United States, national disability programs are challenged to adjudicate millions of work disability claims each year in a timely and accurate manner. The Work Disability Functional Assessment Battery (WD-FAB) was developed to provide work disability agencies and other interested parties a comprehensive and efficient approach to profiling a person’s function related to their ability to work. The WD-FAB is grounded by the International Classification of Functioning, Disability, and Health conceptual framework.
To explore the patterns of cognitive and motor recovery at four time points from admission to nine months post-discharge from IR and to investigate the association of therapeutic factors and pre- and post-discharge conditions with long-term outcomes.
Secondary analysis of traumatic brain injury-practice based evidence (TBI-PBE) dataset.
Inpatient rehabilitation (IR) in Ontario, Canada.
A total of 85 patients with TBI consecutively admitted for IR between 2008 and 2011 and had data available from admission to nine months follow-up.
Main outcome measure
Functional Independence Measure-Rasch cognitive and motor scores at admission, discharge, three, and nine months post-discharge.
Cognitive and motor recovery showed similar patterns of improvement with recovery up to three months but no significant change from three to nine months. Having fewer post-discharge health conditions was associated with better long-term cognitive scores (95% CI -13.06, -1.2) and added 9.9 % to the explanatory power of the model. More therapy time in complex occupational therapy activities (95% CI .02, .09) and fewer post-discharge health conditions (95% CI -19.5, -3.8) were significant predictors of better long-term motor function and added 14.3% and 7.2% to the explanatory power of the model, respectively.
Results of this study inform health care providers about the influence of the timing of IR on cognitive and motor recovery. In addition, it underlines the importance of making patients and families aware of residual health conditions following discharge from IR.
Embedded sensors of the smartphones offer opportunities for granular, patient-autonomous measurements of neurological dysfunctions for disease identification, management, and for drug development. We hypothesized that aggregating data from two simple smartphone tests of fine finger movements with differing contribution of specific neurological domains (i.e., strength & cerebellar functions, vision, and reaction time) will allow establishment of secondary outcomes that reflect domain-specific deficit. This hypothesis was tested by assessing correlations of smartphone-derived outcomes with relevant parts of neurological examination in multiple sclerosis (MS) patients. We developed MS test suite on Android platform, consisting of several simple functional tests. This paper compares cross-sectional and longitudinal performance of Finger tapping and Balloon popping tests by 76 MS patients and 19 healthy volunteers (HV). The primary outcomes of smartphone tests, the average number of taps (per two 10-s intervals) and the average number of pops (per two 26-s intervals) differentiated MS from HV with similar power to traditional, investigator-administered test of fine finger movements, 9-hole peg test (9HPT). Additionally, the secondary outcomes identified patients with predominant cerebellar dysfunction, motor fatigue and poor eye-hand coordination and/or reaction time, as evidenced by significant correlations between these derived outcomes and relevant parts of neurological examination. The intra-individual variance in longitudinal sampling was low. In the time necessary for performing 9HPT, smartphone tests provide much richer and reliable measurements of several distinct neurological functions. These data suggest that combing more creatively-construed smartphone apps may one day recreate the entire neurological examination.
Neurological examination measures diverse functions of the central (CNS) and peripheral nervous systems to diagnose neurological diseases and guide treatment decisions. Thorough neurological examination takes between 30 and 60 min to complete and years of training to master. This poses problem both for developing countries, which often lack neurologists, and for developed countries where cost-hikes and administrative requirements severely limit the time clinicians spend examining patients.
Additionally, clinical scales derived from traditional neurological examination are rather insensitive and prone to biases, which limits their utility in drug development. Therefore, non-clinician administered measurements of physical disability such as timed 25-foot walk (25FW) and 9-hole peg test (9HPT) or measurements of cognitive functions exemplified by paced auditory serial addition test (PASAT) and symbol digit modalities test (SDMT), are frequently used in clinical trials of neurological diseases such as multiple sclerosis (MS) (1, 2). Especially combining these “functional scales” with clinician-based disability scales such as Expanded Disability Status Scale (EDSS)(3) into EDSS-plus (4) or Combinatorial weight-adjusted disability scale (CombiWISE) (5) enhances sensitivity of clinical trial outcomes. However, these sensitive combinatorial scales are rarely, if ever acquired in clinical practice due to time and expense constrains.
Measuring neurological functions by patients via smartphones (6–8) may pose a solution for all aforementioned problems, while additionally empowering patients for greater participation in their neurological care. We have previously found comparable sensitivity and specificity of simple, smartphone-amenable measurements of finger and foot taps to 9HPT and 25FW, respectively (9). In this study, we explored iterative development/optimization of smartphone-based measurements of neurological functions by: 1. Exploring clinical utility of new features that can be extracted from finger tapping; 2. Development of “balloon popping” smartphone test that builds on finger tapping by expanding neurological functions necessary for task completion to eye movements and cognitive skills, and 3. By decoding app-collected raw data into secondary (derived) features that may better reflect deficits in specific neurological functions.
Materials and Methods
Developing the Smartphone Apps
Tapping and Balloon popping tests were written using Java in the Android Studio integrated development environment. Both tests went through iterative development and optimization following beta testing with developers and then clinical trial testing with patients and healthy volunteers. Each of the individual tests are standalone applications and can be downloaded individually to the phone using an Android Package (APK) emailed to phones or directly installed through USB connection with Android Studio. Installation and initial testing of applications were completed on a variety of personal Android phones, with no particular specifications. Testing in the clinic with patients and longitudinal testing was completed on Google Pixel XL 2017 phones. Android 8.1 Oreo operating system was used for the most recent version of the application, with the intention of keeping the operating system the app runs on up to date with the most recent version released by Android.
For the purposes of this study, we created a front-end application that can flexibly incorporate a variety of test apps. The front-end prompts for user profiles where a testing ID, birth month and year, gender, and dominant hand may be entered so data collected is associated with the user profile. Through a cloud-based spreadsheet, “prescriptions” of test app configurations are set for each user such that they may have a unique combination of tests tailored to their disability level.
The tapping test goal was similar to previously validated non-smartphone administered tapping tests (9), where users had to tap as quickly as possible over a 10 s duration and the final score is the average of two attempts. The test uses touch recognition over a rectangular area covering the bottom half of a vertically oriented phone screen (Figure 1A). Users can tap anywhere in a marked off gray area. The total number of taps for each of two trials and the calculated average is displayed immediately afterwards on the screen. In addition to total taps over the duration of the test, the app also records the duration, Android system time, and pressure for each tap as background data. Pressure for app recording is interpreted from the size of the touch area on each tap, where larger tap area corresponds to a higher pressure reading. Because the pressure function was added later and therefore the data are missing for the majority of current cohort, this function is not investigated in current study.
Figure 1. Smartphone Apps. (A) Tapping Test where user can tap repeatedly anywhere in the gray rectangle over the bottom half of the screen. (B) Popping Test where the dark blue circle will disappear and simultaneously reappear randomly across the screen as soon as the user touches it.
The balloon popping test was conceptually envisioned as an extension of tapping test that expands neurological functions necessary for test completion from pure motoric, to motoric, visual, and cognitive (attention and reaction time). The primary goal for this test is to touch as many randomly generated dark blue circles (balloons) moving across the screen in succession over the 26-s test duration as possible. During optimization of the app we tested 3 sizes of the target balloon and a 100-pixel balloon was selected as optimal based on preliminary results. The analyses of the other two circle sizes are provided as part of sensitivity analyses (Supplementary Figure 1), as conclusions from these tests support data presented in the main text of the paper. There is only one balloon to pop on the screen at a time (Figure 1B) and as soon as the user touches anywhere on the circle, another circle will appear in a random location. The random generation of balloon locations was created by random number functions in Java for both the x and y coordinates of the center of the circle, with the constraint of the entire balloon having to be visible on the screen. If the user taps on a background location, the current balloon stays in the same location and is only moved to a new random location after accurately tapping on the balloon. Following app completion, the total number of balloons popped and calculated average (from two trials) is displayed on the phone for the user. The x and y coordinates of all balloon and background hits, the system time, duration, and pressure (in the same manner as tap pressure) for each tap are also recorded as background data and stored in cloud-based data system.
Following the completion of a tapping or balloon popping test trial, an intermediate message displayed on the screen asks if the users would like to submit their results or retake the most recent trial (Supplementary Videos 1, 2). If the user selects the retake option the collected data for the trial is discarded locally on the phone and not sent to any cloud-based database. This was implemented to avoid noise associated with test interruptions or other unforeseen circumstances that affected test performance. Following selection of the submit option, the data is uploaded immediately to a cloud-based database if the smartphone is connected to WiFi. If the phone is not connected to WiFi, then the submitted test trial results are stored locally on the phone and uploaded to the database as soon as the phone is connected to WiFi.
The app development process is in continuation given user and clinician feedback in addition to integration of more tests into the front-end. User feedback, user’s ability to perform Apps in a “practice mode”, and training videos for individual tests (Supplementary Videos 1, 2) are integrated into the front-end dashboard that manages different tests.[…]
Introduction: Standardized measurement of clinical outcomes across sites and over time is critical to clinical trials. Barriers to outcome measure training include availability of standardized materials and time to train, plus wide geographic distribution of trial personnel. To address these, an online training and certification program based on NIHSS testing was developed and implemented for the Fugl Meyer Motor Assessment (FMMA) in support of multisite stroke recovery trials.
Methods: This program includes Fugl Meyer Arm (FMA) and Leg (FML) components, runs on a web host, and is based on a valid, reliable approach to FMMA testing known to decrease variance in scoring (See et al, NNR, 2013;27:732-741). The website hosts training courses, reference manuals, video patient cases for formal certification testing plus 3 rounds of recertification; each round has 2 separate patients. A passing score of 90% is required. After each course, feedback is given.
Results: This program has served as the primary training, certification, and recertification mechanism for 4 multisite recovery trials, including 1 NIH-funded US trial and 3 industry-sponsored international trials. Three trials certify on both FMA and FML, and 1 on FMA only, as primary endpoint. Evaluators are recertified every 4-6 months. The 299 clinicians from 5 countries registered include PT/OT (n=136), MD (n=37), and RN/NP/PA (n=15). For FMA training, 299 persons have registered and 197 completed. For the first round of FMA certification, 267 have registered and 171 passed (mean 1.89 attempts to pass). For the second FMA (first recertification), 78 registered and 65 passed. The passing rate increased with successive rounds of recertification. Similar numbers have been achieved for FML training, certification, and recertification.
Conclusions: The FMMA has established value for capturing treatment-related motor gains in stroke recovery trials. The current online training program is efficient and effective for training, certifying, and recertifying examiners in arm and leg FMMA. Clinical trial assessors training with this program can be expected to provide more accurate and less variable FMMA scores, which increases statistical power, reduces sample sizes, and reduces the cost of clinical trials.
The primary purpose of this scoping review was to describe the nature and extent of the published research that assesses the relationship between psychological features and patient-reported outcome following surgery or rehabilitation of upper extremity disease or injury.
Twenty-two included studies were examined for quantitative study design, outcome measure, inclusion/exclusion criteria, follow-up and recruitment strategy. Patient population and psychological assessment tools were examined for validity.
Twenty-two studies met the inclusion criteria for this study. Only 7 of the 22 studies were longitudinal and the rest were cross sectional studies. Depression was the most common psychological status of interest and was included in 17 studies. Pain catastrophizing was the psychological status of interest in 5 of the studies. Four studies considered anxiety, 3 considered pain anxiety, 3 considered distress, 2 considered coping, 2 considered catastrophic thinking, and 2 considered fear avoidance beliefs.
The majority of studies in this review were cross-sectional studies. Cross-sectional studies may not provide conclusive information about cause-and-effect relationships. This review encourages clinicians to be mindful of the psychological implications found in rehabilitation of individuals with upper extremity disease or injury along with being cognizant of choosing appropriate measurement tools that best represent each patient’s characteristics and diagnoses.
The nature of the research addressing psychological factors affecting outcomes after hand injury focus on negative traits and have limited strength to suggest causation as most have used cross-sectional designs. Stronger longitudinal designs and consideration of positive traits are needed in future studies.