JAMIA Open最新文献

Background: Efficiently identifying the social risks of patients with serious illnesses (SIs) is the critical first step in providing patient-centered and value-driven care for this medically vulnerable population.

Objective: To apply and further hone an existing natural language process (NLP) algorithm that identifies patients who are homeless/at risk of homeless to a SI population.

Methods: Patients diagnosed with SI between 2019 and 2020 were identified using an adapted list of diagnosis codes from the Center for Advance Palliative Care from the Kaiser Permanente Southern California electronic health record. Clinical notes associated with medical encounters within 6 months before and after the diagnosis date were processed by a previously developed NLP algorithm to identify patients who were homeless/at risk of homelessness. To improve the generalizability to the SI population, the algorithm was refined by multiple iterations of chart review and adjudication. The updated algorithm was then applied to the SI population.

Results: Among 206 993 patients with a SI diagnosis, 1737 (0.84%) were identified as homeless/at risk of homelessness. These patients were more likely to be male (51.1%), age among 45-64 years (44.7%), and have one or more emergency visit (65.8%) within a year of their diagnosis date. Validation of the updated algorithm yielded a sensitivity of 100.0% and a positive predictive value of 93.8%.

Conclusions: The improved NLP algorithm effectively identified patients with SI who were homeless/at risk of homelessness and can be used to target interventions for this vulnerable group.

Objective: Latinx people comprise 30% of all new human immunodeficiency virus (HIV) infections in the United States and face many challenges to accessing and engaging with HIV care. To bridge these gaps in care, a Spanish-language mobile health (mHealth) intervention known as ConexionesPositivas (CP) was adapted from an established English-language platform called PositiveLinks (PL) to help improve engagement in care and reduce viral nonsuppression among its users. We aimed to determine how CP can address the challenges that Latinx people with HIV (PWH) in the United States face.

Materials and methods: We conducted a post-implementation study of the CP mHealth platform, guided by principles of user-centered design. We enrolled 20 Spanish-speaking CP users in the study, who completed the previously validated System Usability Scale (SUS) and semistructured interviews. Interviews were transcribed and translated for analysis. We performed thematic coding of interview transcripts in Dedoose.

Results: The SUS composite score was 75, which is within the range of good usability. Four categories of themes were identified in the interviews: client context, strengths of CP, barriers to use and dislikes, and suggestions to improve CP. Positive impacts included encouraging self-monitoring of medication adherence, mood and stress, connection to professional care, and development of a support system for PWH.

Discussion: While CP is an effective and easy-to-use application, participants expressed a desire for improved personalization and interactivity, which will guide further iteration.

Conclusion: This study highlights the importance of tailoring mHealth interventions to improve equity of access, especially for populations with limited English proficiency.

Objective: To analyze PeriData.Net, a clinical registry with linked maternal-infant hospital data of Milwaukee County residents, to demonstrate a predictive analytic approach to perinatal infant risk assessment.

Materials and methods: Using unsupervised learning, we identified infant birth clusters with similar multivariate health indicator patterns, measured using perinatal variables from 2008 to 2019 from n = 43 969 clinical registry records in Milwaukee County, WI, followed by supervised learning risk-propagation modeling to identify key maternal factors. To understand the relationship between socioeconomic status (SES) and birth outcome cluster assignment, we recoded zip codes in Peridata.Net according to SES level.

Results: Three self-organizing map clusters describe infant birth outcome patterns that are similar in the multivariate space. Birth outcome clusters showed higher hazard birth outcome patterns in cluster 3 than clusters 1 and 2. Cluster 3 was associated with lower Apgar scores at 1 and 5 min after birth, shorter infant length, and premature birth. Prediction profiles of birth clusters indicate the most sensitivity to pregnancy weight loss and prenatal visits. Majority of infants assigned to cluster 3 were in the 2 lowest SES levels.

Discussion: Using an extensive perinatal clinical registry, we found that the strongest predictive performance, when considering cluster membership using supervised learning, was achieved by incorporating social and behavioral risk factors. There were inequalities in infant birth outcomes based on SES.

Conclusion: Identifying infant risk hazard profiles can contribute to knowledge discovery and guide future research directions. Additionally, presenting the results to community members can build consensus for community-identified health and risk indicator prioritization for intervention development.

Objectives: Health-related chatbots have demonstrated early promise for improving self-management behaviors but have seldomly been utilized for hypertension. This research focused on the design, development, and usability evaluation of a chatbot for hypertension self-management, called "Medicagent."

Materials and methods: A user-centered design process was used to iteratively design and develop a text-based chatbot using Google Cloud's Dialogflow natural language understanding platform. Then, usability testing sessions were conducted among patients with hypertension. Each session was comprised of: (1) background questionnaires, (2) 10 representative tasks within Medicagent, (3) System Usability Scale (SUS) questionnaire, and (4) a brief semi-structured interview. Sessions were video and audio recorded using Zoom. Qualitative and quantitative analyses were used to assess effectiveness, efficiency, and satisfaction of the chatbot.

Results: Participants (n = 10) completed nearly all tasks (98%, 98/100) and spent an average of 18 min (SD = 10 min) interacting with Medicagent. Only 11 (8.6%) utterances were not successfully mapped to an intent. Medicagent achieved a mean SUS score of 78.8/100, which demonstrated acceptable usability. Several participants had difficulties navigating the conversational interface without menu and back buttons, felt additional information would be useful for redirection when utterances were not recognized, and desired a health professional persona within the chatbot.

Discussion: The text-based chatbot was viewed favorably for assisting with blood pressure and medication-related tasks and had good usability.

Conclusion: Flexibility of interaction styles, handling unrecognized utterances gracefully, and having a credible persona were highlighted as design components that may further enrich the user experience of chatbots for hypertension self-management.

Objective: To develop a standardizable, reproducible method for creating drug codelists that incorporates clinical expertise and is adaptable to other studies and databases.

Materials and methods: We developed methods to generate drug codelists and tested this using the Clinical Practice Research Datalink (CPRD) Aurum database, accounting for missing data in the database. We generated codelists for: (1) cardiovascular disease and (2) inhaled Chronic Obstructive Pulmonary Disease (COPD) therapies, applying them to a sample cohort of 335 931 COPD patients. We compared searching all drug dictionary variables (A) against searching only (B) chemical or (C) ontological variables.

Results: In Search A, we identified 165 150 patients prescribed cardiovascular drugs (49.2% of cohort), and 317 963 prescribed COPD inhalers (94.7% of cohort). Evaluating output per search strategy, Search C missed numerous prescriptions, including vasodilator anti-hypertensives (A and B:19 696 prescriptions; C:1145) and SAMA inhalers (A and B:35 310; C:564).

Discussion: We recommend the full search (A) for comprehensiveness. There are special considerations when generating adaptable and generalizable drug codelists, including fluctuating status, cohort-specific drug indications, underlying hierarchical ontology, and statistical analyses.

Conclusions: Methods must have end-to-end clinical input, and be standardizable, reproducible, and understandable to all researchers across data contexts.

The Multimodal Maternal Infant Perinatal Outpatient Delivery System (MOMI PODS) was developed to facilitate the pregnancy to postpartum primary care transition, particularly for individuals at risk for severe maternal morbidity, via a unique multidisciplinary model of mother/infant dyadic primary care. Specialized clinical informatics platforms are critical to ensuring the feasibility and scalability of MOMI PODS and a smooth perinatal transition into longitudinal postpartum primary care. In this manuscript, we describe the MOMI PODS transition and management clinical informatics platforms developed to facilitate MOMI PODS referrals, scheduling, evidence-based multidisciplinary care, and program evaluation. We discuss opportunities and lessons learned associated with our applied methods, as advances in clinical informatics have considerable potential to enhance the quality and evaluation of innovative maternal health programs like MOMI PODS.

Objectives: Tertiary and quaternary (TQ) care refers to complex cases requiring highly specialized health services. Our study aimed to compare the ability of a natural language processing (NLP) model to an existing human workflow in predictively identifying TQ cases for transfer requests to an academic health center.

Materials and methods: Data on interhospital transfers were queried from the electronic health record for the 6-month period from July 1, 2020 to December 31, 2020. The NLP model was allowed to generate predictions on the same cases as the human predictive workflow during the study period. These predictions were then retrospectively compared to the true TQ outcomes.

Results: There were 1895 transfer cases labeled by both the human predictive workflow and the NLP model, all of which had retrospective confirmation of the true TQ label. The NLP model receiver operating characteristic curve had an area under the curve of 0.91. Using a model probability threshold of ≥0.3 to be considered TQ positive, accuracy was 81.5% for the NLP model versus 80.3% for the human predictions (P = .198) while sensitivity was 83.6% versus 67.7% (P<.001).

Discussion: The NLP model was as accurate as the human workflow but significantly more sensitive. This translated to 15.9% more TQ cases identified by the NLP model.

Conclusion: Integrating an NLP model into existing workflows as automated decision support could translate to more TQ cases identified at the onset of the transfer process.