Progress 09/01/17 to 04/30/18
Outputs
Target Audience:There were two target audiences: The rural patients, and their physical therapist. I. Rural patients (55+) recovering from hip and knee replacements. Mobility Coach benefits them by: Increased exercise compliance Greater involvement and control in the recovery process Reduced number of trips required to PT clinic Faster reduced dependency on dangerous pain medication II. Physical therapy providers managing increasing caseloads and staying connected with patients in rural areas Mobility Coach benefits Physical therapy providers by: Early awareness of patient non-compliance Productivity: Ability to effectively manage larger case load Improved documentation to speed reimbursement claims. Changes/Problems: The team originally planned to offer just one approach for measurement of the exercises, using visual pattern recognition approach that incorporated the Microsoft Kinect. We found that some patients did not want the complexity and space requirements for the setup. Therefore, the team investigated wearable technology. None of the current commercially available wearable offered the proper combination of accuracy, resolution and real-time reporting. Therefore, the team investigated accelerometers, and found a chip set that we could combine with low power Bluetooth, to interface to a standard Android tablet. The team quickly crafted a user interface, sufficient to support the Phase I evaluations. We found that about 50% of therapists and 35% of the rural senior patients preferred the Kinect method. For a brief period of time, Microsoft announced that the Kinect would be discontinued. This decision was later reversed, and units are still available. What opportunities for training and professional development has the project provided?Continued from above: Technical Objective 3: Pilot Study Evaluation. A feasibility study of the Mobility Coach system was then conducted in the lab among 24 rural patients (45 to 93 years old) and their PT. The study measured these patient's ability to independently follow and perform exercises prescribed by the PT, know when to take their pain medications, and answer health, safety, fall risk and pain-related questions. This study compared the Mobility Coach system with their current methods such as using printed/written lists. The basic hypothesis was that rural patients would be able to perform these exercises and pain medication adherence with less assistance and with fewer errors when using the Mobility Coach system prototype as compared to their current method. Selection criteria was developed by Triad: > 60 minutes to nearest PT clinic, 55+ years old, and are within 45 days post-surgery for their LEJR. All participants had lived in rural setting for most of their lives. Population: 24 (18 women, 6 men). Subjects were each paid a $50 stipend for participation in the study. Before the experimental session started, each patient completed a survey using a 5-point Likert scale covering key areas identified in focus group (Objective 1). Participants were provided with training that included hands-on instruction and practice before the measurement of the activity began. The order of presentation of the two methods (existing compared to Mobility Coach) was randomized to control for ordering or learning effects. The field evaluation took place over a 14-week period. There were two dependent measures: 1) 'accuracy' as measured by the number of errors made during the experimental session; and 2) 'independence' as measured by the number of prompts required to complete the session. The prescribed exercises and the medication trays were the same. Each participant followed the instruction, took (pretend) medications (swallowing not required), and answered survey questions. Data collection forms were used to record errors and prompts during each experimental session. The performance of each subject was closely monitored, with verbal prompts from the instructor and assistance provided when requested or as soon as mistakes were made. In this way, individuals always achieved success at the task, even if they needed assistance to achieve it. In addition to the quantitative data collected, there was room provided on the data collection forms to record additional observations as well as statements made by subjects during the test sessions. These observations identified areas for more rigorous assessment during Phase II. The data was analyzed using SPSS, a software package for behavioral statistics. A multivariate analysis of variance procedure was used to evaluate mean differences between the two experimental conditions (Mobility Coach or their current method) for each dependent measure Average Errors and Average Prompts. Table 3 below provides descriptive statistics for each condition by dependent measure. The following four lists document the descriptive statistics by dependent measure by experimental condition: Average Errors - Experimental Condition Typical Method Mean = 4.93 SD = 3.11 SE = 0.782 Minimum = 0 Maximum = 14 Average Errors - Experimental Condition Mobility Coach Mean 1.05 SD = 1.23 SE = 0.291 Minimum = 0 Maximum = 14 Average Prompts - Experimental Condition Typical Method Mean = 1.77 SD = 1.51 SE = .383 Minimum = 0 Maximum = 4 Average Prompts - Experimental Condition Mobility Coach Mean = 6.57 SD = 4.11 SE = 1.004 Minimum = 1 Maximum = 16 The first dependent measure with a significant F statistic was accuracy, as measured by recording Average Errors (p<.001), which was a measure of subjects' ability to correctly complete the task using both methods. Therefore, post-hoc tests were performed to specifically investigate the source for the significance. When using Mobility Coach (X = 1.05, SD= 1.23) subjects made significantly fewer errors when compared to their normal method (p<.001). The second dependent variable, independence, as measured by Average Prompts (p<.001) provided to subjects while performing each of the exercises during the experimental sessions. Subjects required significantly fewer prompts when using Mobility Coach (X = 1.77, SD= 1.51) to complete the tasks as compared to when using their typical method (p>.001). After the testing was completed, each participant was again asked to rate their experience with Mobility Coach using the same statements and 5-point Likert scale that they had used before the testing began. These answers, plus capturing their qualitative responses helped CCI understand the quantitative data collected. The Mobility Coach system was also tested simultaneously with the PTs of the patients in the study. PT population: 24 (15 women, 9 men). These 24 PTs first completed a survey focused on key areas identified in the focus groups. This exercise was repeated after having used Mobility Coach's web-based Dashboard. PTs performed the following tasks: viewed the dashboard and navigated the HIPPA-compliant screens, verified the patient's exercise compliance, played the patient movement data, reviewed med usage, reviewed their fall risk assessment, prescribed new exercises, and posed new questions to the patient. The following Likert scale was used for the following 2 tables: 1 = strongly disagree, 2 = disagree, 3 = Neutral, 4 = agree, 5 = strongly agree Table 4a: Rural Patients using the typical method: Increased exercise compliance Typical Method - Low = 1; High = 3; % 4-5 = 0% Mobilty Coach - Low = 3; High = 5; % 4-5 = 94% Faster recovery Typical Method - Low = 1; High = 3; % 4-5 = 0% Mobilty Coach - Low = 3; High = 5; % 4-5 = 90% Functional improvement Typical Method - Low = 1; High = 2; % 4-5 = 0% Mobilty Coach - Low = 3; High = 5; % 4-5 = 82% Satisfaction with the approach Typical Method - Low = 1; High = 2; % 4-5 = 0% Mobilty Coach - Low = 4; High = 5; % 4-5 = 100% Table 4b: Physical Therapists using the typical method Immediate awareness of a problem Typical Method - Low = 1; High = 3; % 4-5 = 0% Mobilty Coach - Low = 4; High = 5; % 4-5 = 95% Intuitive Dashboard Typical Method - Low = 1; High = 2; % 4-5 = 0% Mobilty Coach - Low = 4; High = 5; % 4-5 = 100% Asynchronous review of patient data Typical Method - Low = 1; High = 2; % 4-5 = 0% Mobilty Coach - Low = 4; High = 5; % 4-5 = 100% Easily adjust / prescribe new exercises Typical Method - Low = 1; High = 3; % 4-5 = 0% Mobilty Coach - Low = 3; High = 5; % 4-5 = 62% Monitor pain medication usage Typical Method - Low = 1; High = 2; % 4-5 = 0% Mobilty Coach - Low = 4; High = 5; % 4-5 = 100% Improves care coordination and reporting Typical Method - Low = 1; High = 3; % 4-5 = 0% Mobilty Coach - Low = 3; High = 5; % 4-5 = 93% Samples of Qualitative data from study participants: From rural seniors: "This is really great, when can I get one?" "Very cool, I didn't know this was possible." "I really want to do my exercises like I'm supposed to, but I just forget to do them, or can't remember how to do them. I think this would really help." "I think that this would help me stay more involved and feel more in control of my rehab." "I normally hate wearing things, like my 'I've fallen and can't get up' button around my neck, but this is different, and I would use this band thing." From their Physical Therapists: "I really liked the dashboard reporting, quick identification of an outlier, and the natural, intuitive flow." "I think this will save my staff a lot of time." Technical Objective 4 is detailed below. How have the results been disseminated to communities of interest?National conferences, such as the Rehabilitation Engineering Society of North America (RESNA), and regional conferences, such as Easter Seals Crossroads. Continued from abvove: Technical Objective 4: Assessment of the Impact on Rural Areas. This proposal is in response to the USDA's SBIR Program Priority and Societal Challenge Area 8.6, Rural and Community Development. Specifically, the technology developed in this SBIR will contribute to an increase in efficiency and effectiveness of Local Government and Public and Private Institutions through improved health care delivery. The integration of new technologies and cloud services will accelerate the physical therapy and recovery of rural patients recovering from hip and knee replacements, also known as lower extremity joint replacements (LEJR). The Phase I results demonstrated that Mobility Coach offers PTs a cost-effective solution and rural patients a method to supplement their rehab with a non-intrusive option to help them stay compliant with their exercises post LEJR. Naturally, the ultimate benefactor of Mobility Coach is the rural patient. However, the primary target market for commercialization are PT providers who are aligned with LEJR orthopedic surgery physician groups. This associated orthopedic doctor may weigh in, but typically, relies on the PT to evaluate and select what services and products to use in the rehab phase. Using this model, a business case analysis of the system included a return on investment projection and impact on reimbursements. The incremental cost was found to be less that 3% of the portion allocated to rehab and generated a return on investment (ROI) of under one month (1/3 of the 90-day episode). Beyond these projections (see full Commercialization Plan), this assessment addressed how Mobility Coach may change the lives of current and future residents of the community, such as: employment opportunities that may shift as the diffusion of technology continues to extend to rural areas, encompassing technical support and training on remote patient monitoring systems and other technologies, which typically offer higher pay. Pathfinders is a potential contract manufacturer in the Huntington, Indiana area that hires the majority of its workforce from suburban and rural areas. Pathfinders Industrial services is a potential manufacturing partner for the hardware component of the Mobility Coach system. More physical therapists can live in rural areas due to the tele-rehab aspects of Mobility Coach. With the Mobility Coach system, physical therapists can serve many more clients and live and work wherever they wish. There were also multiple opportunities for traing and professional development: Two computer scientists on the project grew in their skill sets related to interfacing to verification techniques (sensor and visual pattern recognition). The PI grew in his understanding of the complex Center for Medicaid Medicare Systems (CMS) bunbled payments and other reimbursement regulations, as well as working with various physical therapy and orthopedic surgeon groups. What do you plan to do during the next reporting period to accomplish the goals?
Nothing Reported
Impacts
What was accomplished under these goals?
Technical Objective 1: Requirements Development. Triad Exercise Technologies, a subcontractor on the project, conducted face-to-face focus groups, emails and phone interviews with 10 physical therapists and 12 rural patients, post lower extremity joint replacement (LEJR). Participants were given a survey instrument designed to measure the key areas of importance for a technology such as Mobility Coach as well as the potential impact on their quality of life. The team quickly determined that Mobility Coach benefitted all adult rural patients evenly (not just seniors). There were additional issues that need to be considered when addressing rural seniors, but all can be easily addressed to accommodate senior patients as needed. Those additional accommodations are listed in Table 1. Table 1: Additional Considerations for Rural Seniors Over General Rural Adult Patients Less likely to have Internet Increased risk of falling post LEJR Increased importance on a simplicity in setup and use Therefore, in response to these findings and with the permission of its USDA SBIR Program Manager, CreateAbility Concepts, Inc. has expanded the scope of this project to include all rural adult patients and will assure that the extra considerations for rural seniors are addressed by the system. Primary requirements gathered included the amount of information the system must supply to help the user: a) become oriented to the system, and b) navigate the user interface. Patients were also asked how the system could increase engagement and help them understand what and how to perform the exercises, as well as remind them of pain medications, and answer questions from their PT. PTs were asked how the system could help them asynchronously review their patient's exercises and prescribe new exercises. The survey also asked if the patient would benefit from tracking pain medication usage. Additional noteworthy conclusions are summarized in Table 2 (a full list of will be provided in the Phase I final report): Table 2a: Benefits to Physical Therapists Saves time and money through faster recovery Reduces risk of expensive interventions High level dashboard showing all patients Match the appropriate method to the patient (Kinect or Band) Fast Identification of problem Visual and text alerts Predictive analytics Quickly zoom in to get Details Rehab stage Asynchronous review of exercise data Pain medication usage Push new exercise protocol to patient Confirm battery status / correct setup Initiate/answer video visit session Table 2b: Benefits to Rural Patients Accelerated recovery through exercise compliance Reduces trips to therapy and risk of painful interventions Increased engagement, clarity on what to do and when Reminders Guided instruction Immediate feedback on progress Works with TV remote or tapping on tablet Assurance they set it up correctly Pain medication reminders and dispensers Reduced Fall risk Works with land line Automatic updates of new exercises No hassle equipment returns Request and initiate a video visit session from app. Technical Objective 2: Prototype Development. Six prototypes of the Mobility Coach system were designed and developed by CCI for use in the pilot study evaluation. This first release of the software, hardware and cloud service components concentrated on supporting the essential features and capabilities that determined the technical feasibility, as well as the usability with study participants. The physical in-home components of the prototypes are illustrated in a YouTube video, and consist of: 1) The two methods are available for tracking the patient's knee or hip exercises: a) a special 3D motion tracking camera (via the Microsoft Kinect) or b) using a special exercise band with a low power Bluetooth (BLE) movement sensor (an enhanced version of the accelerometer technology found in fitness bands and smart watches). The 3D motion tracking camera is preferred by PTs when the patient has the ability to setup and use the equipment, and when higher accuracy on joint rotation, or when the overall body movement is important. 2) The patient gets detailed instruction on what to do and how via tapping on a mobile device, such as the table shown here, or using their TV remote and their TV. 3) For tracking pain medication usage, a disc-style medication tray with a rotating slot includes an integrated wireless sensor to detect when pain medication was dispensed. YouTube videos of each approach can be viewed at: https://youtu.be/rtz8LXehtLk, and https://www.youtube.com/watch?time_continue=1&v=QKY3iiWt44g A dashboard enables PTs with the proper permissions to review all exercise data and history to make more informed decisions on what steps to take in the physical rehab of the rural patient. Typically, PT's only need any device that can use a standard browser to view and navigate the HIPAA-compliant dashboard. On the dashboard, a red/yellow/green status indicator is displayed next to each patient, indicating if this patient is on track with their therapy (green, slightly behind (yellow0, or this patient should be contacted immediately (red). The PT simply selects the patient of interest to display the top three parameters over time: number of repetitions, joint range of motion, and pain medication usage. The PT also can select and play any recorded day and exercise type. PTs can then review answers to health and pain-related questions, set alerts for the PT or designated individual when thresholds are exceeded, pose new questions, and prescribe new exercises. Alternatively, the PT may also use the 3D motion tracking camera (via Kinect) if they wish to record a custom exercise for the patient. These are then auto-sent to the patient's system and viewed by the patient on their television or tablet. The main page on the PT Dashboard also allows the PT to drill down to specific details, such as date and time stamped excise events, number of reps per set, number of sets per day, range of motion and pain medication usage.
Publications
|