OBJECTIVE: Evaluate the impact of community tele-paramedicine (CTP) on patient experience and satisfaction relative to community-level indicators of health disparity. MATERIALS AND METHODS: This mixed-methods study evaluates patient-reported satisfaction and experience with CTP, a facilitated telehealth program combining in-home paramedic visits with video visits by emergency physicians. Anonymous post-CTP visit survey responses and themes derived from directed content analysis of in-depth interviews from participants of a randomized clinical trial of mobile integrated health and telehealth were stratified into high, moderate, and low health disparity Community Health Districts (CHD) according to the 2018 New York City (NYC) Community Health Survey. RESULTS: Among 232 CTP patients, 55% resided in high or moderate disparity CHDs but accounted for 66% of visits between April 2019 and October 2021. CHDs with the highest proportion of CTP visits were more adversely impacted by social determinants of health relative to the NYC average. Satisfaction surveys were completed in 37% of 2078 CTP visits between February 2021 and March 2023 demonstrating high patient satisfaction that did not vary by community-level health disparity. Qualitative interviews conducted with 19 patients identified differing perspectives on the value of CTP: patients in high-disparity CHDs expressed themes aligned with improved health literacy, self-efficacy, and a more engaged health system, whereas those from low-disparity CHDs focused on convenience and uniquely identified redundancies in at-home services. CONCLUSIONS: This mixed-methods analysis suggests CTP bridges the digital health divide by facilitating telehealth in communities negatively impacted by health disparities.

OBJECTIVES: To determine if different formats for conveying machine learning (ML)-derived postpartum depression risks impact patient classification of recommended actions (primary outcome) and intention to seek care, perceived risk, trust, and preferences (secondary outcomes). MATERIALS AND METHODS: We recruited English-speaking females of childbearing age (18-45 years) using an online survey platform. We created 2 exposure variables (presentation format and risk severity), each with 4 levels, manipulated within-subject. Presentation formats consisted of text only, numeric only, gradient number line, and segmented number line. For each format viewed, participants answered questions regarding each outcome. RESULTS: Five hundred four participants (mean age 31 years) completed the survey. For the risk classification question, performance was high (93%) with no significant differences between presentation formats. There were main effects of risk level (all P < .001) such that participants perceived higher risk, were more likely to agree to treatment, and more trusting in their obstetrics team as the risk level increased, but we found inconsistencies in which presentation format corresponded to the highest perceived risk, trust, or behavioral intention. The gradient number line was the most preferred format (43%). DISCUSSION AND CONCLUSION: All formats resulted high accuracy related to the classification outcome (primary), but there were nuanced differences in risk perceptions, behavioral intentions, and trust. Investigators should choose health data visualizations based on the primary goal they want lay audiences to accomplish with the ML risk score.

OBJECTIVE: We developed and externally validated a machine-learning model to predict postpartum depression (PPD) using data from electronic health records (EHRs). Effort is under way to implement the PPD prediction model within the EHR system for clinical decision support. We describe the pre-implementation evaluation process that considered model performance, fairness, and clinical appropriateness. MATERIALS AND METHODS: We used EHR data from an academic medical center (AMC) and a clinical research network database from 2014 to 2020 to evaluate the predictive performance and net benefit of the PPD risk model. We used area under the curve and sensitivity as predictive performance and conducted a decision curve analysis. In assessing model fairness, we employed metrics such as disparate impact, equal opportunity, and predictive parity with the White race being the privileged value. The model was also reviewed by multidisciplinary experts for clinical appropriateness. Lastly, we debiased the model by comparing 5 different debiasing approaches of fairness through blindness and reweighing. RESULTS: We determined the classification threshold through a performance evaluation that prioritized sensitivity and decision curve analysis. The baseline PPD model exhibited some unfairness in the AMC data but had a fair performance in the clinical research network data. We revised the model by fairness through blindness, a debiasing approach that yielded the best overall performance and fairness, while considering clinical appropriateness suggested by the expert reviewers. DISCUSSION AND CONCLUSION: The findings emphasize the need for a thorough evaluation of intervention-specific models, considering predictive performance, fairness, and appropriateness before clinical implementation.

OBJECTIVE: We aimed to evaluate the feasibility of using ChatGPT as programming support for nursing PhD students conducting analyses using the All of Us Researcher Workbench. MATERIALS AND METHODS: 9 students in a PhD-level nursing course were prospectively randomized into 2 groups who used ChatGPT for programming support on alternating assignments in the workbench. Students reported completion time, confidence, and qualitative reflections on barriers, resources used, and the learning process. RESULTS: The median completion time was shorter for novices and certain assignments using ChatGPT. In qualitative reflections, students reported ChatGPT helped generate and troubleshoot code and facilitated learning but was occasionally inaccurate. DISCUSSION: ChatGPT provided cognitive scaffolding that enabled students to move toward complex programming tasks using the All of Us Researcher Workbench but should be used in combination with other resources. CONCLUSION: Our findings support the feasibility of using ChatGPT to help PhD nursing students use the All of Us Researcher Workbench to pursue novel research directions.

OBJECTIVE: Participation in healthcare research shapes health policy and practice; however, low trust is a barrier to participation. We evaluated whether returning health information (information transparency) and disclosing intent of data use (intent transparency) impacts trust in research. MATERIALS AND METHODS: We conducted an online survey with a representative sample of 502 US adults. We assessed baseline trust and change in trust using 6 use cases representing the Social-Ecological Model. We assessed descriptive statistics and associations between trust and sociodemographic variables using logistic and multinomial regression. RESULTS: Most participants (84%) want their health research information returned. Black/African American participants were more likely to increase trust in research with individual information transparency (odds ratio (OR) 2.06 [95% confidence interval (CI): 1.06-4.34]) and with intent transparency when sharing with chosen friends and family (3.66 [1.98-6.77]), doctors and nurses (1.96 [1.10-3.65]), or health tech companies (1.87 [1.02-3.40]). Asian, Native American or Alaska Native, Native Hawaiian or Pacific Islander, Multirace, and individuals with a race not listed, were more likely to increase trust when sharing with health policy makers (1.88 [1.09-3.30]). Women were less likely to increase trust when sharing with friends and family (0.55 [0.35-0.87]) or health tech companies (0.46 [0.31-0.70]). DISCUSSION: Participants wanted their health information returned and would increase their trust in research with transparency when sharing health information. CONCLUSION: Trust in research is influenced by interrelated factors. Future research should recruit diverse samples with lower baseline trust levels to explore changes in trust, with variation on the type of information shared.

OBJECTIVES: Patients increasingly use patient-reported outcomes (PROs) to self-monitor their health status. Visualizing PROs longitudinally (over time) could help patients interpret and contextualize their PROs. The study sought to assess hospitalized patients' objective comprehension (primary outcome) of text-only, non-graph, and graph visualizations that display longitudinal PROs. MATERIALS AND METHODS: We conducted a clinical research study in 40 hospitalized patients comparing 4 visualization conditions: (1) text-only, (2) text plus visual analogy, (3) text plus number line, and (4) text plus line graph. Each participant viewed every condition, and we used counterbalancing (systematic randomization) to control for potential order effects. We assessed objective comprehension using the International Organization for Standardization protocol. Secondary outcomes included response times, preferences, risk perceptions, and behavioral intentions. RESULTS: Overall, 63% correctly comprehended the text-only condition and 60% comprehended the line graph condition, compared with 83% for the visual analogy and 70% for the number line (P = .05) conditions. Participants comprehended the visual analogy significantly better than the text-only (P = .02) and line graph (P = .02) conditions. Of participants who comprehended at least 1 condition, 14% preferred a condition that they did not comprehend. Low comprehension was associated with worse cognition (P < .001), lower education level (P = .02), and fewer financial resources (P = .03). CONCLUSIONS: The results support using visual analogies rather than text to display longitudinal PROs but caution against relying on graphs, which is consistent with the known high prevalence of inadequate graph literacy. The discrepancies between comprehension and preferences suggest factors other than comprehension influence preferences, and that future researchers should assess comprehension rather than preferences to guide presentation decisions.

Objective: This integrative review identifies convergent and divergent areas of need for collecting and using patient-generated health data (PGHD) identified by patients and providers (i.e., physicians, nurses, advanced practice nurses, physician assistants, and dietitians). Methods: A systematic search of 9 scholarly databases targeted peer-reviewed studies published after 2010 that reported patients' and/or providers' needs for incorporating PGHD in clinical care. The studies were assessed for quality and bias with the Mixed-Methods Appraisal Tool. The results section of each article was coded to themes inductively developed to categorize patient and provider needs. Distinct claims were extracted and areas of convergence and divergence identified. Results: Eleven studies met inclusion criteria. All had moderate to low risk of bias. Three themes (clinical, logistic, and technological needs), and 13 subthemes emerged. Forty-eight claims were extracted. Four were divergent and twenty were convergent. The remainder was discussed by only patients or only providers. Conclusion: As momentum gains for integrating PGHD into clinical care, this analysis of primary source data is critical to understanding the requirements of the 2 groups directly involved in collection and use of PGHD.