Preferences of Older Adult Veterans With Heart Failure for Engaging With Mobile Health Technology to Support Self-care: Qualitative Interview Study Among Patients With Heart Failure and Content Analysis.

BACKGROUND: Heart failure (HF) affects approximately 6.5 million adults in the United States, disproportionately afflicting older adults. Mobile health (mHealth) has emerged as a promising tool to empower older adults in HF self-care. However, little is known about the use of this approach among older adult veterans. OBJECTIVE: The goal of this study was to explore which features of an app were prioritized for older adult veterans with HF. METHODS: Between January and July 2021, we conducted semistructured interviews with patients with heart failure aged 65 years and older at a single facility in an integrated health care system (the Veterans Health Administration). We performed content analysis and derived themes based on the middle-range theory of chronic illness, generating findings both deductively and inductively. The qualitative questions captured data on the 3 key themes of the theory: self-care maintenance, self-care monitoring, and self-care management. Qualitative responses were analyzed using a qualitative data management platform, and descriptive statistics were used to analyze demographic data. RESULTS: Among patients interviewed (n=9), most agreed that a smartphone app for supporting HF self-care was desirable. In addition to 3 a priori themes, we identified 7 subthemes: education on daily HF care, how often to get education on HF, support of medication adherence, dietary restriction support, goal setting for exercises, stress reduction strategies, and prompts of when to call a provider. In addition, we identified 3 inductive themes related to veteran preferences for app components: simplicity, ability to share data with caregivers, and positive framing of HF language. CONCLUSIONS: We identified educational and tracking app features that can guide the development of HF self-care for an older adult veteran population. Future research needs to be done to extend these findings and assess the feasibility of and test an app with these features.

Predicting emergency department orders with multilabel machine learning techniques and simulating effects on length of stay.

OBJECTIVE: Emergency departments (EDs) continue to pursue optimal patient flow without sacrificing quality of care. The speed with which a healthcare provider receives pertinent information, such as results from clinical orders, can impact flow. We seek to determine if clinical ordering behavior can be predicted at triage during an ED visit. MATERIALS AND METHODS: Using data available during triage, we trained multilabel machine learning classifiers to predict clinical orders placed during an ED visit. We benchmarked 4 classifiers with 2 multilabel learning frameworks that predict orders independently (binary relevance) or simultaneously (random k-labelsets). We evaluated algorithm performance, calculated variable importance, and conducted a simple simulation study to examine the effects of algorithm implementation on length of stay and cost. RESULTS: Aggregate performance across orders was highest when predicting orders independently with a multilayer perceptron (median F1 score = 0.56), but prediction frameworks that simultaneously predict orders for a visit enhanced predictive performance for correlated orders. Visit acuity was the most important predictor for most orders. Simulation results indicated that direct implementation of the model would increase ordering costs (from $21 to $45 per visit) but reduce length of stay (from 158 minutes to 151 minutes) over all visits. DISCUSSION: Simulated implementations of the predictive algorithm decreased length of stay but increased ordering costs. Optimal implementation of these predictions to reduce patient length of stay without incurring additional costs requires more exploration. CONCLUSIONS: It is possible to predict common clinical orders placed during an ED visit with data available at triage.

A Tablet Based Embodied Conversational Agent to Promote Smoking Cessation among Veterans: A Feasibility Study.

Embodied Conversational Agent (ECA) offer a new means to support smokers as a virtual coach and motivate them to quit smoking. In this study we assess the feasibility and acceptability of an ECA to support quit smoking ("aka ECA-Q"). ECA-Q, a 14-days program, delivered through Tablet computers, interacts with participants with supporting messages for quit smoking and motivates them to set a quit date. Study participants (n = 6) were Veterans receiving medical care at Boston VA Healthcare System who responded to an open advertisement. Participants completed a survey at baseline and after 14 days follow-up. All participants were satisfied with the ECA program and liked the features of the agent; three out of six participants had set a quit date by the end of the 14 days. Participants reported several positive and less important features of the agent and made suggestions to improve the agent. This study shows that a conversation agent is acceptable to smoking veterans to help them in setting a quit date with an ultimate goal of quit smoking. Insights gained from this study would be useful to redesign the current version of ECA-Q program for a future randomized controlled trial to test the efficacy.

Anticoagulant use for atrial fibrillation in the elderly.

OBJECTIVES: To determine the influence of advanced age on anticoagulant use in subjects with atrial fibrillation and to explore the extent to which risk factors for stroke and contraindications to anticoagulant therapy predict subsequent use. DESIGN: Retrospective cohort study. SETTING: The Veterans Affairs Boston Healthcare System. PARTICIPANTS: A total of 2,217 subjects with nonvalvular atrial fibrillation. MEASUREMENTS: Administrative databases were use to identify subject's age, anticoagulant use, and the presence of a diagnosis of atrial fibrillation, cerebrovascular accident, hypertension, diabetes mellitus, congestive heart failure, or gastrointestinal or cerebral hemorrhage. RESULTS: Unadjusted analysis showed no difference in warfarin use between those aged 75 and older and younger subjects regardless of the presence (33.9% vs 35.7%, P=.37) or absence (33.4% vs 34.7%, P=.58) of contraindications to anticoagulant therapy. Multivariate modeling demonstrated a 14% reduction (95% confidence interval (CI)=4-22%) in anticoagulant use with each advancing decade of life. Intracranial hemorrhage was a significant deterrent (odds ratio (OR)=0.27 95% CI=0.06-0.85). History of hypertension (OR=2.90, 95% CI=2.15-3.89), congestive heart failure (OR=1.70, 95% CI=1.41-2.04), and cerebrovascular accident (OR=1.54, 95% CI=1.25-1.89) were significant independent predictors for anticoagulant use. CONCLUSION: Despite consensus guidelines to treat all atrial fibrillation patients aged 75 and older with anticoagulants, advancing age was found to be a deterrent to warfarin use. Better estimates of the risk:benefit ratio for oral anticoagulant therapy in older patients with atrial fibrillation are needed to optimize decision-making.

Generalizability of a simple approach for predicting hospital admission from an emergency department.

OBJECTIVES: The objective was to test the generalizability, across a range of hospital sizes and demographics, of a previously developed method for predicting and aggregating, in real time, the probabilities that emergency department (ED) patients will be admitted to a hospital inpatient unit. METHODS: Logistic regression models were developed that estimate inpatient admission probabilities of each patient upon entering an ED. The models were based on retrospective development (n = 4,000 to 5,000 ED visits) and validation (n = 1,000 to 2,000 ED visits) data sets from four heterogeneous hospitals. Model performance was evaluated using retrospective test data sets (n = 1,000 to 2,000 ED visits). For one hospital the developed model also was applied prospectively to a test data set (n = 910 ED visits) coded by triage nurses in real time, to compare results to those from the retrospective single investigator-coded test data set. RESULTS: The prediction models for each hospital performed reasonably well and typically involved just a few simple-to-collect variables, which differed for each hospital. Areas under receiver operating characteristic curves (AUC) ranged from 0.80 to 0.89, R(2) correlation coefficients between predicted and actual daily admissions ranged from 0.58 to 0.90, and Hosmer-Lemeshow goodness-of-fit statistics of model accuracy had p > 0.01 with one exception. Data coded prospectively by triage nurses produced comparable results. CONCLUSIONS: The accuracy of regression models to predict ED patient admission likelihood was shown to be generalizable across hospitals of different sizes, populations, and administrative structures. Each hospital used a unique combination of predictive factors that may reflect these differences. This approach performed equally well when hospital staff coded patient data in real time versus the research team retrospectively.

Predicting emergency department inpatient admissions to improve same-day patient flow.

OBJECTIVES: The objectives were to evaluate three models that use information gathered during triage to predict, in real time, the number of emergency department (ED) patients who subsequently will be admitted to a hospital inpatient unit (IU) and to introduce a new methodology for implementing these predictions in the hospital setting. METHODS: Three simple methods were compared for predicting hospital admission at ED triage: expert opinion, naïve Bayes conditional probability, and a generalized linear regression model with a logit link function (logit-linear). Two months of data were gathered from the Boston VA Healthcare System's 13-bed ED, which receives approximately 1,100 patients per month. Triage nurses were asked to estimate the likelihood that each of 767 triaged patients from that 2-month period would be admitted after their ED treatment, by placing them into one of six categories ranging from low to high likelihood. Logit-linear regression and naïve Bayes models also were developed using retrospective data and used to estimate admission probabilities for each patient who entered the ED within a 2-month time frame, during triage hours (1,160 patients). Predictors considered included patient age, primary complaint, provider, designation (ED or fast track), arrival mode, and urgency level (emergency severity index assigned at triage). RESULTS: Of the three methods considered, logit-linear regression performed the best in predicting total bed need, with a receiver operating characteristic (ROC) area under the curve (AUC) of 0.887, an R(2) of 0.58, an average estimation error of 0.19 beds per day, and on average roughly 3.5 hours before peak demand occurred. Significant predictors were patient age, primary complaint, bed type designation, and arrival mode (p < 0.0001 for all factors). The naïve Bayesian model had similar positive predictive value, with an AUC of 0.841 and an R(2) of 0.58, but with average difference in total bed need of approximately 2.08 per day. Triage nurse expert opinion also had some predictive capability, with an R(2) of 0.52 and an average difference in total bed need of 1.87 per day. CONCLUSIONS: Simple probability models can reasonably predict ED-to-IU patient volumes based on basic data gathered at triage. This predictive information could be used for improved real-time bed management, patient flow, and discharge processes. Both statistical models were reasonably accurate, using only a minimal number of readily available independent variables.

Architectural choices and challenges of integrating electronic patient questionnaires into the electronic medical record to support patient-centered care.

Strategies to deliver guideline-concordant, patient-centered care during office visits sometimes impose conflicting demands on clinicians. One way to help relieve time-constrained visits and to improve the patient-centeredness of care may be through patients electronically self-reporting data that flow automatically into an EMR note for clinician confirmation or editing, relieving physicians of some data entry and rote history-gathering tasks, thus freeing up time to allow clinicians to focus on significant issues and patient concerns while also increasing the likelihood that necessary data are gathered and available for decision-making. We developed a prototype to enable such data gathering and integration into the EMR. The lack of consistent provision of interfaces by vendors for sending data into EMRs and the idiosyncrasies of any particular EMR in the context of a particular organizations IT infrastructure and policies pose architectural choices and challenges that healthcare organizations embarking on such IT projects may need to consider.

Three complementary definitions of polypharmacy: methods, application and comparison of findings in a large prescription database.

PURPOSE: To develop and compare three operational definitions of polypharmacy using a large prescription database. METHODS: We defined Cumulative polypharmacy as all prescriptions filled during a 178 day window--which captured 95% of eventual refills as calculated from Kaplan-Meier and cumulative incidence curves. Continuous polypharmacy was all prescriptions filled in two such windows 6 months apart. Simultaneous polypharmacy was the number of prescriptions active on a particular day, as determined by fill dates and amount of medication given. We applied these definitions to the outpatient prescription files of New England veterans and compared the resulting estimates of polypharmacy using descriptive statistics. RESULTS: 118,013 patients received at least one prescription between January 1998 and July 1999. Cumulative polypharmacy averaged 3.54 (SD = 4.95) medications and continuous polypharmacy averaged 1.96 (SD = 3.23). Examination of simultaneous polypharmacy over 40 2-week intervals revealed an average of 2.63 (CI 2.61-2.65), a minimum of 1.09 (CI 1.08-1.10) and maximum of 4.94 (CI 4.92-4.96). One arbitrarily selected observation point had an average of 3.87 (SD = 3.17). CONCLUSIONS: Our definitions of cumulative and continuous polypharmacy serve to set upper and lower bounds for the estimate of polypharmacy. Our method for simultaneous polypharmacy gives numbers that diverge in some respects, but it is better at showing transient changes in medications. The methods are complementary and allow exploration of various aspects of medication use, such as cumulative medication exposure over time, the influence of chronic medical problems, and the causes of rapid changes in medications.