BACKGROUND: Functional limitations are associated with poor clinical outcomes, higher mortality, and disability rates, especially in older adults. Continuous assessment of patients' functionality is important for clinical practice; however, traditional questionnaire-based assessment methods are very time-consuming and infrequently used. Mobile sensing offers a great range of sources that can assess function and disability daily. OBJECTIVE: This work aims to prove the feasibility of an interpretable machine learning pipeline for predicting function and disability based on the World Health Organization Disability Assessment Schedule (WHODAS) 2.0 outcomes of clinical outpatients, using passively collected digital biomarkers. METHODS: One-month-long behavioral time-series data consisting of physical and digital activity descriptor variables were summarized using statistical measures (minimum, maximum, mean, median, SD, and IQR), creating 64 features that were used for prediction. We then applied a sequential feature selection to each WHODAS 2.0 domain (cognition, mobility, self-care, getting along, life activities, and participation) in order to find the most descriptive features for each domain. Finally, we predicted the WHODAS 2.0 functional domain scores using linear regression using the best feature subsets. We reported the mean absolute errors and the mean absolute percentage errors over 4 folds as goodness-of-fit statistics to evaluate the model and allow for between-domain performance comparison. RESULTS: Our machine learning-based models for predicting patients' WHODAS functionality scores per domain achieved an average (across the 6 domains) mean absolute percentage error of 19.5%, varying between 14.86% (self-care domain) and 27.21% (life activities domain). We found that 5-19 features were sufficient for each domain, and the most relevant being the distance traveled, time spent at home, time spent walking, exercise time, and vehicle time. CONCLUSIONS: Our findings show the feasibility of using machine learning-based methods to assess functional health solely from passively sensed mobile data. The feature selection step provides a set of interpretable features for each domain, ensuring better explainability to the models' decisions-an important aspect in clinical practice.
BACKGROUND: Today's diabetes-oriented telemedicine systems can gather and analyze many parameters like blood glucose levels, carbohydrate intake, insulin doses, and physical activity levels (steps). Information collected can be presented to patients in a variety of graphical outputs. Despite the availability of several technical means, a large percentage of patients do not reach the goals established in their diabetes treatment. OBJECTIVE: The objective of the study was to evaluate the benefits of the Diani telemedicine system for the treatment of patients with type 1 diabetes mellitus. METHODS: Data were collected during a 24-week feasibility study. Patients responded to the World Health Organization Quality of Life - BREF (WHOQOL-BREF) questionnaire and a system evaluation questionnaire. The level of glycated hemoglobin (HbA1c) and the patient's body weight were measured, and the patient's use of the telemedicine system and their daily physical activity level were monitored. All data were sent from the patient's device to the Diani server using a real-time diabetes diary app. Wilcoxon and Friedman tests and the linear mixed effects method were used for data analysis. RESULTS: This study involved 10 patients (men: n=5; women: n=5), with a mean age of 47.7 (SD 19.3) years, a mean duration of diabetes of 10.5 (SD 8.6) years, and a mean HbA1c value of 59.5 (SD 6.7) mmol/mol. The median number of days the patients used the system was 84. After the intervention, the mean HbA1c decreased by 4.35 mmol/mol (P=.01). The patients spent 18.6 (SD 6.8) minutes on average using the app daily. After the intervention, the number of patients who measured their blood glucose level at least 3 times a day increased by 30%. The graphical visualization of the monitored parameters, automatic transmission of measured data from the glucometer, compatibility, and interconnection of individual devices when entering data were positively evaluated by patients. CONCLUSIONS: The Diani system was found to be beneficial for patients with type 1 diabetes mellitus in terms of managing their disease. Patients perceived it positively; it strengthened their knowledge of diabetes and their understanding of the influences of the measured values on the management of their disease. Its use had a positive effect on the HbA1c level.
- Klíčová slova
- compensation, diabetes, diabetes mellitus, evaluation, feasibility, intervention, mHealth, mobile health, quality of life, telehealth, telemedicine, telemedicine system, telemonitoring,
- Publikační typ
- časopisecké články MeSH
BACKGROUND: It is very difficult to find a consensus that will be accepted by most players when creating health care legislation. The Czech electronic prescription system was launched in 2011 and new functions were introduced in 2018. To ensure that these functions will not conflict with any other existing law, a process modeling tool based on the patent "Method and system for automated requirements modeling" was used successfully in the Czech Republic for the first time. OBJECTIVE: The aim of this project was to develop another successful application of process modeling to add COVID-19 vaccination records to the existing electronic prescription system. METHODS: The method employed was based on the mathematical theory of hierarchical state diagrams and process models. In the first step, sketches that record the results of informal discussions, interviews, meetings, and workshops were prepared. Subsequently, the architecture containing the main participants and their high-level interactions was drafted. Finally, detailed process diagrams were drawn. Each semiresult was discussed with all involved team members and stakeholders to incorporate all comments. By repeating this procedure, individual topics were gradually resolved and the areas of discussion were narrowed down until reaching complete agreement. RESULTS: This method proved to be faster, clearer, and significantly simpler than other methods. Owing to the use of graphic tools and symbols, the risk of errors, inaccuracies, and misunderstandings was significantly reduced. The outcome was used as an annex to the bill in the legislative process. One of the main benefits of this approach is gaining a higher level of understanding for all parties involved (ie, legislators, the medical community, patient organizations, and information technology professionals). The process architecture model in a form of a graphic scheme has proven to be a valuable communication platform and facilitated negotiation between stakeholders. Moreover, this model helped to avoid several inconsistencies that appeared during workshops and discussions. Our method worked successfully even when participants were from different knowledge areas. CONCLUSIONS: The vaccination record process model was drafted in 3 weeks and it took a total of 2 months to pass the bill. In comparison, the initial introduction of the electronic prescription system using conventional legislative methods took over 1 year, involving immediate creation of a text with legislative intent, followed by paragraph-by-section wording of the legislation that was commented on directly. These steps are repeated over and over, as any change in any part of the text has to be checked and rechecked within the entire document. Compared with conventional methods, we have shown that using our method for the process of modification of legislation related to such a complex issue as the integration of COVID-19 vaccination into an electronic prescription model significantly simplifies the preparation of a legislative standard.
- Klíčová slova
- COVID-19, communication, eHealth, electronic prescription, medical, platform, process modeling, state diagram, vaccination, vaccine,
- Publikační typ
- časopisecké články MeSH
BACKGROUND: Telemedicine solutions, especially in the face of epidemiological emergencies such as the COVID-19 pandemic, played an important role in the remote communication between patients and medical providers. However, the implementation of modern technologies should rely on patients' readiness toward new services to enable effective cooperation with the physician. Thus, successful application of patient-centric telehealth services requires an in-depth analysis of users' expectations. OBJECTIVE: This study aimed to evaluate factors determining readiness for using telehealth solutions among patients with cardiovascular diseases. METHODS: We conducted a cross-sectional study based on an investigator-designed, validated questionnaire that included 19 items (demographics, health status, medical history, previous health care experience, expected telehealth functionalities, and preferred remote communication methods). Multivariate logistic regression was applied to assess the relationship between readiness and their determinants. RESULTS: Of the 249 respondents, 83.9% (n=209) consented to the use of telemedicine to contact a cardiologist. The nonacceptance of using telemedicine was 2 times more frequent in rural dwellers (odds ratio [OR] 2.411, 95% CI 1.003-5.796) and patients without access to the internet (OR 2.432, 95% CI 1.022-5.786). In comparison to participants living in rural areas, city dwellers demonstrated a higher willingness to use telemedicine, including following solutions: issuing e-prescriptions (19/31, 61.3% vs 141/177, 79.7%; P=.02); alarming at the deterioration of health (18/31, 58.1% vs 135/177, 76.3%; P=.03); and arranging or canceling medical visits (16/31, 51.6% vs 126/176, 71.6%; P=.03). Contact by mobile phone was preferred by younger patients (OR 2.256, 95% CI 1.058-4.814), whereas older patients and individuals who had no previous difficulties in accessing physicians preferred landline phone communication. CONCLUSIONS: During a nonpandemic state, 83.9% of patients with cardiovascular diseases declared readiness to use telemedicine solutions.
- Klíčová slova
- acceptance, patient-cardiologist contact, readiness, telehealth, telemedicine,
- Publikační typ
- časopisecké články MeSH
BACKGROUND: There is increasing interest in the potential uses of mobile health (mHealth) technologies, such as wearable biosensors, as supplements for the care of people with neurological conditions. However, adherence is low, especially over long periods. If people are to benefit from these resources, we need a better long-term understanding of what influences patient engagement. Previous research suggests that engagement is moderated by several barriers and facilitators, but their relative importance is unknown. OBJECTIVE: To determine preferences and the relative importance of user-generated factors influencing engagement with mHealth technologies for 2 common neurological conditions with a relapsing-remitting course: multiple sclerosis (MS) and epilepsy. METHODS: In a discrete choice experiment, people with a diagnosis of MS (n=141) or epilepsy (n=175) were asked to select their preferred technology from a series of 8 vignettes with 4 characteristics: privacy, clinical support, established benefit, and device accuracy; each of these characteristics was greater or lower in each vignette. These characteristics had previously been emphasized by people with MS and or epilepsy as influencing engagement with technology. Mixed multinomial logistic regression models were used to establish which characteristics were most likely to affect engagement. Subgroup analyses explored the effects of demographic factors (such as age, gender, and education), acceptance of and familiarity with mobile technology, neurological diagnosis (MS or epilepsy), and symptoms that could influence motivation (such as depression). RESULTS: Analysis of the responses to the discrete choice experiment validated previous qualitative findings that a higher level of privacy, greater clinical support, increased perceived benefit, and better device accuracy are important to people with a neurological condition. Accuracy was perceived as the most important factor, followed by privacy. Clinical support was the least valued of the attributes. People were prepared to trade a modest amount of accuracy to achieve an improvement in privacy, but less likely to make this compromise for other factors. The type of neurological condition (epilepsy or MS) did not influence these preferences, nor did the age, gender, or mental health status of the participants. Those who were less accepting of technology were the most concerned about privacy and those with a lower level of education were prepared to trade accuracy for more clinical support. CONCLUSIONS: For people with neurological conditions such as epilepsy and MS, accuracy (ie, the ability to detect symptoms) is of the greatest interest. However, there are individual differences, and people who are less accepting of technology may need far greater reassurance about data privacy. People with lower levels of education value greater clinician involvement. These patient preferences should be considered when designing mHealth technologies.
- Klíčová slova
- digital health, discrete choice experiment, epilepsy, health data, health economics, mHealth, mobile technology, multiple sclerosis, neurological conditions, wearable biosensors, wearable technology,
- Publikační typ
- časopisecké články MeSH
BACKGROUND: Secondary schools are an ideal setting to identify young people experiencing mental health difficulties such as anxiety or depression. However, current methods of identification rely on cumbersome paper-based assessments, which are lengthy and time-consuming to complete and resource-intensive for schools to manage. Artemis-A is a prototype web app that uses computerized adaptive testing technology to shorten the length of the assessment and provides schools with a simple and feasible solution for mental health assessment. OBJECTIVE: The objectives of this study are to coproduce the main components of the Artemis-A app with stakeholders to enhance the user interface, to carry out usability testing and finalize the interface design and functionality, and to explore the acceptability and feasibility of using Artemis-A in schools. METHODS: This study involved 2 iterative design feedback cycles-an initial stakeholder consultation to inform the app design and user testing. Using a user-centered design approach, qualitative data were collected through focus groups and interviews with secondary school pupils, parents, school staff, and mental health professionals (N=48). All transcripts were thematically analyzed. RESULTS: Initial stakeholder consultations provided feedback on preferences for the user interface design, school administration of the assessment, and outcome reporting. The findings informed the second iteration of the app design and development. The unmoderated usability assessment indicated that young people found the app easy to use and visually appealing. However, school staff suggested that additional features should be added to the school administration panel, which would provide them with more flexibility for data visualization. The analysis identified four themes relating to the implementation of the Artemis-A in schools, including the anticipated benefits and drawbacks of the app. Actionable suggestions for designing mental health assessment apps are also provided. CONCLUSIONS: Artemis-A is a potentially useful tool for secondary schools to assess the mental health of their pupils that requires minimal staff input and training. Future research will evaluate the feasibility and effectiveness of Artemis-A in a range of UK secondary schools.
- Klíčová slova
- assessment, computerized adaptive testing, coproduction, mental health, mobile apps, qualitative study, schools, user-centered design, young people, youth,
- Publikační typ
- časopisecké články MeSH
