BMC digital health最新文献

Introduction: Older adults are at risk of medication non-adherence due to complex medication regimens and medication management challenges. Medication adherence technologies can help, but previous research demonstrates variability in usability and preferences of their features among older adults. Therefore, our objective is to examine older adults' preferences for medication adherence technology features and their trade-offs to guide the development of these technologies. This will facilitate serving older adults better by addressing their needs and preference.

Methods: Guided by the Patient-Centered Benefit-Risk Framework, we conducted a questionnaire based preference elicitation study where older adults ranked 10 medication adherence technology features identified through qualitative interviews, then identified acceptable trade-offs. Recruitment of our sample was based on convenience sampling, and the inclusion criteria was older adults above 60 years and older and able to speak and read English. The ranking was evaluated by calculating the relative importance using relative importance index (RII) and the trade-offs were assessed using win rate analysis. Statistical significance was assessed using Kruskal Wallis analysis.

Results: Thirty older adults were recruited, of which twenty-three (mean age 73 years, 47.8% males) participated. The 10 reported features were button size, screen size, device size, compartment division, setting time and alarm, alarm sound, user-friendly leaflet, battery operated, locking features, and number of steps to set up the device. Screen size was ranked highest with relative importance index (RII) of 0.75. Win-rate analysis of trade-offs revealed that a user-friendly leaflet was the most frequently selected feature with p value < 0.001 (Kruskal Wallis test).

Conclusion: This study highlights the importance of understanding the preferences of older adults to guide selecting the medication adherence technology that better meet their needs, as well as developing tools supporting medication management and adherence.

Clinical trial number: Not applicable.

Supplementary information: The online version contains supplementary material available at 10.1186/s44247-025-00222-z.

Background: Chronic Obstructive Pulmonary Disease (COPD) is a leading cause of morbidity and mortality in the United States. Frequent exacerbations result in higher use of emergency services and hospitalizations, leading to poor patient outcomes and high costs. The objective of this study is to demonstrate the feasibility of a multimodal, community-based intervention in treating acute COPD exacerbations.

Results: Over 18 months, 1,333 patients were approached and 100 (7.5%) were enrolled (mean age 66, 52% female). Ninety-six participants (96%) remained in the study for the full enrollment period. Fifty-five (55%) participated in tele-pulmonary-rehabilitation. Participants wore the smartwatch for a median of 114 days (IQR 30-210) and 18.9 hours/day (IQR16-20) resulting in a median of 1034 minutes/day (IQR 939-1133). The rate at which participants completed scheduled survey instruments ranged from 78-93%. Nearly all participants (85%) performed COPD ecological momentary assessment at least once with a median of 4.85 recordings during study participation. On average, a 2.48-point improvement (p=0.03) in COPD Assessment Test Score was observed from baseline to study completion. The adherence and symptom improvement metrics were not associated with baseline patient activation measures.

Conclusions: A multimodal intervention combining preventative care, symptom and biometric monitoring, and MIH services was feasible in adults living with COPD. Participants demonstrated high protocol fidelity and engagement and reported improved quality of life.

Background: Within long-term care (LTC) homes, telehealth use has been found to reduce unnecessary emergency department transfers, support the care needs of rural and underserved communities, and supplement in-person physician care. Despite these benefits, it is not well understood when telehealth is not an appropriate medium for providing physician care to residents with complex health needs. This knowledge gap must be addressed given the recent rise in telehealth use in LTC homes in many health systems following the COVID-19 pandemic, when virtual care use increased in many health care sectors to limit travel and in-person exposure risks, that is expected to be maintained going forward.

Methods: This analysis contributes to a broader evaluative study investigating care provider and care recipient experiences and preferences for physician telehealth in LTC homes within the Fraser Health region in British Columbia, Canada. For data collection, semi-structured interviews and focus groups were undertaken with seventy care providers (staff, physicians) and recipients (residents, family caregivers). Using a thematic approach, transcripts were analyzed to find common instances when using telehealth for physician care was seen as not appropriate across participant groups.

Results: Three types of patient care activities were identified as not appropriate to be conducted via physician visits using telehealth. First, new patient visits were thought to benefit from an interpersonal and conversational familiarity that could not be supported by telehealth. Second, difficult in-depth conversations that required conversational nuance (e.g., eye contact, supportive body language), such as palliative care planning, were thought to be inappropriate for telehealth appointments. Finally, instances where LTC staff would need to perform hands-on clinical assessments on behalf of physicians who were attending virtually via telehealth were not seen as desirable.

Conclusions: This analysis highlights perspectives surrounding when telehealth is not appropriate for providing physician services for residents in LTC based on the preferences and experiences shared by both care recipients and care providers. The findings present an opportunity to develop and implement guidelines on appropriate use of telehealth in LTC to support best care practices.

Background: Most individuals with eating disorders (EDs) do not receive treatment, and those who do receive care typically do not receive evidence-based treatment, partly due to lack of accessible provider training. This study developed a novel "all-in-one" online platform for disseminating training for mental health providers in cognitive-behavioral therapy guided self-help (CBTgsh) for EDs and supporting its implementation. The aim of the study was to obtain usability data from the online platform prior to evaluating its effects on provider training outcomes and patient ED symptom outcomes in an open pilot trial.

Methods: Nine mental health provider participants (n = 4 in Cycle 1; n = 5 in Cycle 2) and 9 patient participants (n = 4 in Cycle 1; n = 5 in Cycle 2) were enrolled over two cycles of usability testing. In Cycle 1, we recruited providers and patients separately to complete brief platform testing sessions. In Cycle 2, we recruited provider-patient dyads; providers completed training using the platform and subsequently delivered CBTgsh to a patient for three weeks. Usability was assessed using the System Usability Scale (SUS), the Usefulness, Satisfaction, and Ease of Use Questionnaire (USE), and semi-structured interviews.

Results: Interview feedback converged on two themes for providers (applicability of program for real-world use, platform structure and function) and two themes for patients (barriers and facilitators to engagement, perceived treatment effects). SUS and USE scores were in the "average" to "good" ranges across cycles.

Conclusions: Findings from this study demonstrate preliminary feasibility and acceptability of the online platform. Data collected in this study will inform further refinements to the online platform. The platform's effects on provider training outcomes and patient ED symptom outcomes will be evaluated in an open pilot trial. Given the wide treatment gap for EDs and barriers to dissemination and implementation of evidence-based treatments, the online platform represents a scalable solution that could improve access to evidence-based care for EDs.

Background: Older adults with insomnia who use benzodiazepine receptor agonists (BZAs) may be at increased risk of cognitive impairment. Cognitive testing outside of clinical settings may yield results that are more reflective of individuals' cognition in their natural environment, where they experience fluctuations in mental state (e.g. drowsiness). We assessed the feasibility of self-administered cognitive testing via smartphone apps for collecting in-moment, in-context data about a person's current state (ecological momentary assessment, EMA).

Methods: Participants (n = 20; median age 66 years; 14 females, 18 white) aged ≥ 55 years who were recruited from a BZA deprescribing trial were invited to complete (over a 28 day period) daily drowsiness assessments on an EMA app (cued by smartwatch alerts) and weekly self-administered digit span (DGS) forward/backward (2 [minimum] - 9 [maximum]), verbal paired associates (VPA; 0 [best]-24 [worst] total errors), and cued delayed recall of VPA (VPA-DR; 0 [best] - 8 [worst] errors) tests on a cognitive app. We assessed the completion of EMA (0-28 days) and cognitive sessions (# of participants per # sessions completed). We performed thematic analysis of the participant interviews.

Results: The median number of days that EMA was completed was 24.5. Twelve (60%) individuals participated in 4 sessions; 2 (10%) individuals participated in 3 sessions; 2 (10%) individuals participated in 2 sessions; and 4 (20%) individuals participated in 1 session. No drowsiness was reported 36% of the time, whereas 38% of the responses reflected feeling "a little bit" drowsy and 26% at least "somewhat" drowsy. Mean cognitive test scores were DGS-Forward = 7 (SD 1.3), DGS-Backward = 5.6 (SD 1.0), VPA total errors = 9.9 (SD 3.7), and VPA-DR = 2.2 (SD 1.9). Three themes emerged from the participant interviews: 1) concern for one's own cognitive abilities, 2) strategies employed for optimizing scores (including strategies that would invalidate results), and 3) ease of use of the applications.

Conclusions: Our findings indicate that mobile cognitive tests and EMAs are feasible in this older population. Further work is needed to understand how scores are influenced by the setting, mood, and behaviors.

Supplementary information: The online version contains supplementary material available at 10.1186/s44247-025-00158-4.

Background: Although malaria is preventable and treatable, it continues to be a significant cause of illness and death. Early diagnosis through testing is critical in reducing malaria-related morbidity and mortality. Malaria rapid diagnostic tests (mRDTs) are preferred for their ease of use, sensitivity, and rapid results, yet misadministration and misinterpretation errors persist. This study investigated whether pairing an existing application with an AI-based software could enhance interpretation accuracy among Frontline Healthcare Workers (FHWs) in Kano State, Nigeria.

Methods: A comparative analysis was conducted, examining mRDT interpretations by FHWs, trained expert mRDT reviewers (Panel Readers), and AI-based computer vision algorithms. The accuracy comparisons included: (1) AI interpretation versus Panel Read interpretation, (2) FHW interpretation versus Panel Read interpretation, (3) FHW interpretation versus AI interpretation, and (4) AI performance on faint positive lines. Accuracy was reported as a weighted F1 score, reflecting the harmonic mean of recall (sensitivity) and precision (positive predictive value).

Results: The AI algorithm demonstrated high accuracy, matching Panel Read interpretations correctly for positives 96.38% of the time and negatives 97.12%. FHW interpretations agreed with the Panel Read 96.82% on positives and 94.31% on negatives. Comparison of FHW and AI interpretations showed 97.52% agreement on positives and 93.38% on negatives. The overall accuracy was higher for AI (weighted F1 score of 96.4) compared to FHWs (95.3). Notably, the AI accurately identified 90.2% of 163 faint positive mRDTs, whereas FHWs correctly identified 76.1%.

Conclusion: AI-based computer vision algorithms performed comparably to trained and experienced FHWs and exceeded FHW performance in identifying faint positives. These findings demonstrate the potential of AI technology to enhance the accuracy of mRDT interpretation, thereby improving malaria diagnosis and reporting accuracy in malaria-endemic, resource-limited settings.

Supplementary information: The online version contains supplementary material available at 10.1186/s44247-025-00190-4.

Background: Caregivers are essential in supporting cancer patients, especially those undergoing hematopoietic cell transplantation (HCT). While traditional interventions have been beneficial for improving health-related quality of life (HRQOL), they typically require significant time and staff resources. To address this, we developed the Roadmap mHealth app, which incorporates resilience-building features based on positive psychology. The app integrates Fitbit^® devices to monitor and visualize steps, sleep, and self-reported mood in the Roadmap app. A two-arm randomized controlled trial (RCT) was conducted to evaluate the impact of Roadmap's positive activities on caregiver HRQOL, measured by PROMIS^® Global Health (physical and mental components), at day-120 post-transplant compared to assessment-only controls.

Results: From September 08, 2020 to July 31, 2023, 168 dyads enrolled in the Roadmap study (82 intervention, 87 control). At day-120 post-HCT, there were no significant differences in PROMIS^® Global Mental Health (48.18 vs. 48.16) or Global Physical Health (50.19 vs. 49.52) T-scores between the intervention and control arms. However, among caregivers who used positive activities and chat forum features in the intervention arm, daily mood significantly improved over time (β=0.0082, P<0.001, 95% CI= [0.0060, 0.0010]). Improved daily mood scores were positively associated with better PROMIS^® Global Mental Health at days 30 and 120 post-HCT (β=0.6076, P=0.03, 95% CI=[0.0494, 1.166]). The Fitbit^® device and Roadmap app also received net favorable ratings in feasibility and acceptability, indicating that they were generally easy-to-understand, set up, and use, and superior Mobile App Rating Scale (MARS) scores compared to published mental health apps, with no significant differences between study arms. Compliance with the intervention or control app declined over time.

Conclusions: While the primary outcome, PROMIS^® Global Health, did not differ between arms, caregivers who engaged in positive activities and chat forum through the app experienced a significant improvement in mood compared to those who did not. Additionally, caregivers reported the study as feasible and acceptable over the 120-day study period. Future directions for the Roadmap app are discussed based on these findings.

Trial registration: ClinicalTrials.gov: NCT04094844, Registered 09-16-2019.

Background: Step counting stands out as a highly practical and widely utilised method for assessing an individual's level of physical activity (PA). Although the progress of step counting has undergone a significant transformation in recent times, the need to validate PA applications (apps) is even more pressing to ensure their accuracy. This study aimed to compare the criterion validity of Pacer, Pedometer by ITO Technologies Inc., and Google Fit in measuring step counts in semi-structured laboratory-based conditions.

Method: This comparative experimental study involved 50 students who were fitted with Android phones running the three step counting applications (Pedometer, Pacer and Google Fit) simultaneously while they walked a 30-m walkway at a normal and fast pace during which a video of their walking was recorded with Techno Pouvoir 4 Pro running Android version 11. The steps in the recorded videos served as the criterion compared with the steps recorded by the apps and were counted only when the foot is lifted off the ground and placed in a new location. They were counted independently by two reviewers, who recounted where their level of agreement was more than 3 steps until their report was not more than 2 steps different. The Spearman's correlation was used for a relationship, while Mean Absolute Percentage Error (MAPE) and Bland plot were for validity testing at an Alpha level of 0.05.

Result: While there was no significant difference in step counts among the three apps, a significant difference was found between the steps recorded by the apps and those counted by the video criterion during normal-paced walking but not for fast-paced walking (p > 0.05). The MAPEs for the three applications were moderate, with Google Fit showing 6.6% for normal pace walking and Pedometer and Pacer showing 9.2%. For fast-paced walking, the MAPE was lower at 5.4% across all three applications.

Conclusion: Our findings suggest that a pedometer, Pacer and Google Fit could be used as outcome measures in a general population for counting steps, but Google Fit might be a better step counter when normal pace walking is assessed. However, the study's relatively short duration may have overlooked variations in the applications'performance across different conditions over a longer period; hence, future studies should consider comparing the validity of these applications for a longer duration and among diverse populations.

Background: According to the World Health Organization's Integrated Management of Childhood Illness (IMCI) guidelines, childhood pneumonia diagnosis relies on counting respiratory rate (RR). Counting RR by health workers is frequently inaccurate, leading to misdiagnosis and poor outcomes. Automated RR counters could potentially overcome these limitations. To address this gap, we introduced an automated RR counter and developed a reference video expert panel (VEP) to evaluate its performance.

Methods: We conducted a cross-sectional study involving children aged 0-59 months with suspected pneumonia in Bangladesh. The RR of children was counted using an automated counter (ChARM) and chest movements were simultaneously videotaped. These videos were interpreted by the VEP, trained to a standard procedure. We assessed ChARM's accuracy in comparison to the RR generated by the VEP and summarised the time taken to count RR by ChARM.

Results: Among 339 enrolled children, ChARM successfully counted the RR of 294 children (86.7%). The VEP reached a consensus (i.e., RR count difference within two breaths per minute (bpm) between two VEP members) in 257 of the 294 children (87.4%). ChARM and the VEP agreed on RR counts within two bpm in 68.1% of children (n = 175/257), with a mean difference of 1.7 bpm and limits of agreement ranging from -6.7 to 10.2 bpm. ChARM classified age-adjusted fast and normal breathing with a sensitivity of 95.8% (95% CI: 88.1-99.1) and a specificity of 93.5% (95% CI: 89.0-96.6), demonstrating high agreement (kappa = 0.86). The median time to count the RR by ChARM was 66 s (interquartile range: 61-73 s).

Conclusions: ChARM counted RR accurately against a VEP reference, indicating a potential role in supporting health workers to diagnose pneumonia. However, it was unsuccessful for 1 in 8 cases, typically those more clinically challenging, suggesting a possible systematic bias. Further research is needed to address these issues and confirm ChARM's reliability for broader use in real-world settings.

Trial registration: Current Controlled Trials ISRCTN14120515, registered retrospectively on 19 September 2024.

Supplementary information: The online version contains supplementary material available at 10.1186/s44247-025-00175-3.

Background: HIV disproportionately impacts Black women in the Southern United States. Interventions that incorporate social support and group learning have been proven effective, but following the onset of the COVID-19 pandemic, public health research has become increasingly interested in the usage of mobile health (mHealth) and telehealth to make interventions more accessible. Gathering feedback to ensure a culturally relevant adaptation of mHealth and virtual groups and support platforms is critical, as there is a persistent disparity in mobile device and laptop ownership, which can decrease accessibility of these platforms.

Methods: A formative phase was conducted for a trial of a virtual group and social support platform in addition to an mHealth intervention to promote pre-exposure prophylaxis (PrEP) uptake and sexual and reproductive health outcomes among young Black women in North Carolina aged 18 to 30 years. Focus group discussions were conducted with service providers (n = 12) and the study population (n = 15), followed by pretesting of the virtual platform with members of the study population (n = 7).

Results: Feedback from the focus group discussions and pretesting sessions were overall positive. Participants expressed the importance of a design that is culturally relevant, including customizable avatars that display a diversity of body types, skin tones, and hairstyles. Pretesting participants indicated that utilization of Gather.town on a mobile device was more difficult than on a laptop. Overall, participants valued the privacy mechanisms in place and supported the idea of discussing these critical topics with other young women like themselves.

Conclusions: The importance of this initial step is essential when discussing a new virtual group platform that might address accessibility. Although most feedback was positive, participants also provided important considerations for accessibility of the platform, especially when considering disparities in device ownership.

Trial registration: NCT05753683, first registered 01-06-2023.

Supplementary information: The online version contains supplementary material available at 10.1186/s44247-025-00204-1.