Self-identified rurality in a nationally representative population in the US.

INTRODUCTION: In the US, health services research most often relies on Rural-Urban Commuting Area (RUCA) classification codes to measure rurality. This measure is known to misrepresent rurality and does not rely on individual experiences of rurality associated with healthcare inequities. We aimed to determine a patient-centered RUCA-based definition of rurality. METHODS: In this cross-sectional study, we conducted an online survey asking US residents, 'Do you live in a rural area?' and the rationale for their answer. We evaluated the concordance between their self-identified rurality and their ZIP code-derived RUCA designation of rurality by calculating Cohen's kappa (κ) statistic and percent agreement. RESULTS: Of the 774 participants, 456 (58.9%) and 318 (41.1%) individuals had conventional urban and rural RUCA classifications, respectively. There was only moderate agreement between perceived rurality and rural RUCA classification (κ=0.48; 95% confidence interval (CI)=0.42-0.54). Among people living within RUCA 2-3 defined urban areas (n=51), percent agreement was only 19.6%. Discordance was driven by their perception of the population density, proximity to the nearest neighbor, proximity to a metropolitan area, and the number of homes in their area. Based on our results, we reclassified RUCA 2-3 designations as rural, resulting in an increase in overall concordance (κ=0.56; 95%CI=0.50-0.62). DISCUSSION: Patient-centered rural-urban classification is required to effectively evaluate the impact of rurality on health disparities. This study presents a more patient-centric RUCA-based classification of rurality that can be easily operationalized in future research in situations in which self-reported rural status is missing or challenging to obtain. CONCLUSION: Reclassification of RUCA 2-3 as rural represents a more patient-centric definition of rurality.

Association between patient-reported onset-to-door time and mortality in patients hospitalized with COVID-19 disease.

INTRODUCTION: Timely hospital presentation and treatment are critical for recovery from coronavirus disease (COVID-19). However, the relationship between symptom onset-to-door time and key clinical outcomes, such as inpatient mortality, has been poorly understood due to the difficulty of retrospectively measuring symptom onset in observational data. This study examines the association between patient-reported symptom onset-to-door time (ODT) and mortality among patients hospitalized and treated for COVID-19 disease. METHODS: We conducted a retrospective cohort study of emergency department (ED) encounters of patients with COVID-19 disease who were hospitalized and received remdesivir and/or dexamethasone between March 1, 2020, and March 1, 2022. The exposure was patient-reported ODT in days. The outcome of interest was inpatient mortality, including referral to hospice care. We used multivariable logistic regression to examine the association between ODT and mortality while adjusting for patient characteristics, hospital sites, and seasonality. We tested whether severe illness on hospital presentation modified the association between ODT and mortality. Severe illness was defined by Emergency Severity Index triage level 1 or 2 and hypoxia (SpO2 < 94%). RESULTS: Of the 3451 ED hospitalizations included, 439 (12.7%) resulted in mortality, and 1693 (49.1%) involved patients with severe illness on hospital presentation. Greater ODT was significantly associated with lower odds of inpatient mortality (adjusted odds ratio (AOR) = 0.96, 95% CI = 0.93-1.00, P = 0.023). There was a statistically significant interaction between ODT and severe illness at hospital arrival on mortality, suggesting the negative association between ODT and mortality specifically pertained to patients who were not severely ill upon ED presentation (AOR = 0.93, 95% CI = 0.87-1.00, P = 0.035). The adjusted probability of mortality was significantly lower for non-severely ill, hospitalized patients who presented on days 8-14 (5.2%-3.3%) versus days 0-3 (9.4%-7.5%) after symptom onset. CONCLUSION: More days between symptom onset and hospital arrival were associated with lower mortality among hospitalized patients treated for COVID-19 disease, particularly if they did not have severe illness at ED presentation. However, onset-to-door time was not associated with mortality among hospitalized patients with severe illness at ED presentation. Collectively, these results suggest that non-severely ill COVID-19 patients who require hospitalization are less likely to decompensate with each passing day without severe illness. These findings may continue to guide clinical care delivery for hospitalized COVID-19 patients.

A 680,000-person megastudy of nudges to encourage vaccination in pharmacies.

Encouraging vaccination is a pressing policy problem. To assess whether text-based reminders can encourage pharmacy vaccination and what kinds of messages work best, we conducted a megastudy. We randomly assigned 689,693 Walmart pharmacy patients to receive one of 22 different text reminders using a variety of different behavioral science principles to nudge flu vaccination or to a business-as-usual control condition that received no messages. We found that the reminder texts that we tested increased pharmacy vaccination rates by an average of 2.0 percentage points, or 6.8%, over a 3-mo follow-up period. The most-effective messages reminded patients that a flu shot was waiting for them and delivered reminders on multiple days. The top-performing intervention included two texts delivered 3 d apart and communicated to patients that a vaccine was "waiting for you." Neither experts nor lay people anticipated that this would be the best-performing treatment, underscoring the value of simultaneously testing many different nudges in a highly powered megastudy.

A systematic review and meta-analysis of enrollment into ARDS and sepsis trials published between 2009 and 2019 in major journals.

BACKGROUND: Enrollment problems are common among randomized controlled trials conducted in the ICU. However, little is known about actual trial enrollment rates and influential factors. We set out to determine the overall enrollment rate in recent randomized controlled trials (RCTs) of patients with acute respiratory distress syndrome (ARDS), acute lung injury (ALI), or sepsis, and which factors influenced enrollment rate. METHODS: We conducted a systematic review by searching Pubmed using predefined terms for ARDS/ALI and sepsis to identify individually RCTs published among the seven highest impact general medicine and seven highest impact critical care journals between 2009 and 2019. Cluster randomized trials were excluded. Data were extracted by two independent reviewers using an electronic database management system. We conducted a random-effects meta-analysis of the eligible trials for the primary outcome of enrollment rate by time and site. RESULTS: Out of 457 articles identified, 94 trials met inclusion criteria. Trials most commonly evaluated pharmaceutical interventions (53%), were non-industry funded (78%), and required prospective informed consent (81%). The overall mean enrollment rate was 0.83 (95% confidence interval: 0.57-1.21) participants per month per site. Enrollment in ARDS/ALI and sepsis trials were 0.48 (95% CI 0.32-0.70) and 0.98 (95% CI 0.62-1.56) respectively. The enrollment rate was significantly higher for single-center trials (4.86; 95% CI 2.49-9.51) than multicenter trials (0.52; 95% CI 0.41-0.66). Of the 36 trials that enrolled < 95% of the target sample size, 8 (22%) reported slow enrollment as the reason. CONCLUSIONS: In this systematic review and meta-analysis, recent ARDS/ALI and sepsis clinical trials had an overall enrollment rate of less than 1 participant per site per month. Novel approaches to improve critical care trial enrollment efficiency are needed to facilitate the translation of best evidence into practice.

Content and Communication of Inpatient Family Visitation Policies During the COVID-19 Pandemic: Sequential Mixed Methods Study.

BACKGROUND: Inpatient health care facilities restricted inpatient visitation due to the COVID-19 pandemic. There is no existing evidence of how they communicated these policies to the public nor the impact of their communication choices on public perception. OBJECTIVE: This study aims to describe patterns of inpatient visitation policies during the initial peak of the COVID-19 pandemic in the United States and the communication of these policies to the general public, as well as to identify communication strategies that maximize positive impressions of the facility despite visitation restrictions. METHODS: We conducted a sequential, exploratory, mixed methods study including a qualitative analysis of COVID-19 era visitation policies published on Pennsylvania-based facility websites, as captured between April 30 and May 20, 2020 (ie, during the first peak of the COVID-19 pandemic in the United States). We also conducted a factorial survey-based experiment to test how key elements of hospitals' visitation policy communication are associated with individuals' willingness to seek care in October 2020. For analysis of the policies, we included all inpatient facilities in Pennsylvania. For the factorial experiment, US adults were drawn from internet research panels. The factorial survey-based experiment presented composite policies that varied in their justification for restricted visitation, the degree to which the facility expressed ownership of the policy, and the inclusion of family-centered care support plans. Our primary outcome was participants' willingness to recommend the hypothetical facility using a 5-point Likert scale. RESULTS: We identified 104 unique policies on inpatient visitation from 363 facilities' websites. The mean Flesch-Kincaid Grade Level for the policies was 14.2. Most policies prohibited family presence (99/104, 95.2%). Facilities justified the restricted visitation policies on the basis of community protection (59/104, 56.7%), authorities' guidance or regulations (34/104, 32.7%), or scientific rationale (23/104, 22.1%). A minority (38/104, 36.5%) addressed how restrictive visitation may impair family-centered care. Most of the policies analyzed used passive voice to communicate restrictions. A total of 1321 participants completed the web-based survey. Visitation policy elements significantly associated with willingness to recommend the facility included justifications based on community protection (OR 1.44, 95% CI 1.24-1.68) or scientific rationale (OR 1.30, 95% CI 1.12-1.51), rather than those based on a governing authority. The facility expressed a high degree of ownership over the decision (OR 1.16, 95% CI 1.04-1.29), rather than a low degree of ownership; and inclusion of family-centered care support plans (OR 2.80, 95% CI 2.51-3.12), rather than no such support. CONCLUSIONS: Health systems can immediately improve public receptiveness of restrictive visitation policies by emphasizing community protection, ownership over the facility's policy, and promoting family-centered care.

Preferences for Predictive Model Characteristics among People Living with Chronic Lung Disease: A Discrete Choice Experiment.

Background. Patients may find clinical prediction models more useful if those models accounted for preferences for false-positive and false-negative predictive errors and for other model characteristics. Methods. We conducted a discrete choice experiment to compare preferences for characteristics of a hypothetical mortality prediction model among community-dwelling patients with chronic lung disease recruited from 3 clinics in Philadelphia. This design was chosen to allow us to quantify "exchange rates" between different characteristics of a prediction model. We provided previously validated educational modules to explain model attributes of sensitivity, specificity, confidence intervals (CI), and time horizons. Patients reported their interest in using prediction models themselves or having their physicians use them. Patients then chose between 2 hypothetical prediction models each containing varying levels of the 4 attributes across 12 tasks. Results. We completed interviews with 200 patients, among whom 95% correctly chose a strictly dominant model in an internal validity check. Patients' interest in predictive information was high for use by themselves (n = 169, 85%) and by their physicians (n = 184, 92%). Interest in maximizing sensitivity and specificity were similar (0.88 percentage points of specificity equivalent to 1 point of sensitivity, 95% CI 0.72 to 1.05). Patients were willing to accept a reduction of 6.10 months (95% CI 3.66 to 8.54) in the predictive time horizon for a 1% increase in specificity. Discussion. Patients with chronic lung disease can articulate their preferences for the characteristics of hypothetical mortality prediction models and are highly interested in using such models as part of their care. Just as clinical care should become more patient centered, so should the characteristics of predictive models used to guide that care.

A pilot randomized trial of five financial incentive strategies to increase study enrollment and retention rates.

BACKGROUND: Enrollment and retention difficulties remain major barriers to conducting clinical trials. Financial incentives may promote clinical trial enrollment, however delivery methods to maximize enrollment, maximize retention, and minimize cost remains uncertain. METHODS: We conducted a single-blind, web-based randomized controlled trial of five financial incentive strategies on enrollment and retention rates in a longitudinal study of advance directives among community-dwelling older adults. Participants were eligible to receive a fixed total financial incentive, but the disbursement amounts at each study timepoint (baseline, 2-weeks, 4-weeks, and 6-weeks) differed between study arms. At each timepoint, participants completed a different advance directive. We conducted an intention-to-treat analysis for the primary and secondary outcomes of enrollment and retention. RESULTS: 1803 adults were randomized to one of five incentive strategies: constant n = 361; increasing n = 357; U-shaped n = 361; surprise n = 360; self-select n = 364. Overall, 989 (54.9%) participants elected to enroll in the advance directive study. There were no differences in enrollment rates between the control (constant 53.5%) and any of the four intervention study arms (increasing 54.3%, p = 0.81; U-shaped 57.3%, p = 0.30; surprise 56.9%, p = 0.35; and self-select 52.2%, p = 0.73). There were no differences in retention rates between the control (constant 2.1%) and any of the four intervention study arms (increasing 5.2%, p = 0.09; U-shaped 3.9%, p = 0.23; surprise 2.4%, p = 0.54; self-select 2.1%, p = 0.63). CONCLUSIONS: Financial incentive programs for trial enrollment informed by behavioral economic insights were no more effective than a constant-payment approach in this web-based pilot study.

What's behind the white coat: Potential mechanisms of physician-attributable variation in critical care.

BACKGROUND: Critical care intensity is known to vary across regions and centers, yet the mechanisms remain unidentified. Physician behaviors have been implicated in the variability of intensive care near the end of life, but physician characteristics that may underlie this association have not been determined. PURPOSE: We sought to identify behavioral attributes that vary among intensivists to generate hypotheses for mechanisms of intensivist-attributable variation in critical care delivery. METHODS: We administered a questionnaire to intensivists who participated in a prior cohort study in which intensivists made prognostic estimates. We evaluated the degree to which scores on six attribute measures varied across intensivists. Measures were selected for their relevance to preference-sensitive critical care: a modified End-of-Life Preferences (EOLP) scale, Life Orientation Test-Revised (LOT-R), Jefferson Scale of Empathy (JSE), Physicians' Reactions to Uncertainty (PRU) scale, Collett-Lester Fear of Death (CLFOD) scale, and a test of omission bias. We conducted regression analyses assessing relationships between intensivists' attribute scores and their prognostic accuracy, as physicians' prognostic accuracy may influence preference-sensitive decisions. RESULTS: 20 of 25 eligible intensivists (80%) completed the questionnaire. Intensivists' scores on the EOLP, LOT-R, PRU, CLFOD, and omission bias measures varied considerably, while their responses on the JSE scale did not. There were no consistent associations between attribute scores and prognostic accuracy. CONCLUSIONS: Intensivists vary in feasibly measurable attributes relevant to preference-sensitive critical care delivery. These attributes represent candidates for future research aimed at identifying mechanisms of clinician-attributable variation in critical care and developing effective interventions to reduce undue variation.

Perspectives on using decision-making nudges in physician-patient communications.

Patients engaging in shared decision making must weigh the likelihood of positive and negative outcomes and deal with uncertainty and negative emotions in the situations where desirable options might not be available. The use of "nudges," or communication techniques that influence patients' choices in a predictable direction, may assist patients in making complex decisions. However, nudging patients may be perceived as inappropriate influence on patients' choices. We sought to determine whether key stakeholders, physicians, and laypersons without clinical training consider the use of nudges to be ethical and appropriate in medical decision making. Eighty-nine resident-physicians and 336 Mechanical-Turk workers (i.e., non-clinicians) evaluated two hypothetical preference-sensitive situations, in which a patient with advanced cancer chooses between chemotherapy and hospice care. We varied the following: (1) whether or not the patient's decision was influenced by a mistaken judgment (i.e., decision-making bias) and (2) whether or not the physician used a nudge. Each participant reported the extent to which the communication was ethical, appropriate, and desirable. Both physicians and non-clinicians considered using nudges more positively than not using them, regardless of an initial decision-making bias in patients' considerations. Decomposing this effect, we found that physicians viewed the nudge that endorsed hospice care more favorably than the nudge that endorsed chemotherapy, while non-clinicians viewed the nudge that endorsed chemotherapy more favorably than the nudge that endorsed hospice care. We discuss implications and propose exploring further physicians' and patients' differences in the perception of nudges; the differences may suggest limitations for using nudges in medical decisions.

Numeracy and Understanding of Quantitative Aspects of Predictive Models: A Pilot Study.

BACKGROUND: The assessment of user preferences for performance characteristics of patient-oriented clinical prediction models is lacking. It is unknown if complex statistical aspects of prediction models are readily understandable by a general audience. OBJECTIVE: A pilot study was conducted among nonclinical audiences to determine the feasibility of interpreting statistical concepts that describe the performance of prediction models. METHODS: We conducted a cross-sectional electronic survey using the Amazon Mechanical Turk platform. The survey instrument included educational modules about predictive models, sensitivity, specificity, and confidence intervals (CIs). Follow-up questions tested participants' abilities to interpret these characteristics with both verbatim and gist knowledge. Objective and subjective numeracy were assessed using previously validated instruments. We also tested understanding of these concepts when embedded in a sample discrete choice experiment task to establish feasibility for future elicitation of preferences using a discrete choice experiment design. Multivariable linear regression was used to identify factors associated with correct interpretation of statistical concepts. RESULTS: Among 534 respondents who answered all nine questions, the mean correct responses was 95.9% (95% CI, 93.8-97.4) for sensitivity, 93.1% (95% CI, 90.5-95.0) for specificity, and 86.6% (95% CI, 83.3-89.3) for CIs. Verbatim interpretation was high for all concepts, but significantly higher than gist only for CIs (p < 0.001). Scores on each discrete choice experiment tasks were slightly lower in each category. Both objective and subjective numeracy were positively associated with an increased proportion of correct responses (p < 0.001). CONCLUSION: These results suggest that a nonclinical audience can interpret quantitative performance measures of predictive models with very high accuracy. Future development of patient-facing clinical prediction models can feasibly incorporate patient preferences for model features into their development.

Willingness to participate in pragmatic dialysis trials: the importance of physician decisional autonomy and consent approach.

BACKGROUND: Pragmatic clinical trials embedded in routine delivery of clinical care can lead to improvements in quality of care, but often have design features that raise ethical concerns. METHODS: We performed a discrete choice experiment and used conjoint analysis to assess how specific attributes of pragmatic dialysis trials influenced patients' and physicians' willingness to have their dialysis facility participate in a hypothetical trial of hypertension management. Electronic survey data were collected from 200 patients enrolled from 11 outpatient hemodialysis units and from 203 nephrologists. The three attributes studied were physicians' treatment autonomy, participants' research burden, and the approach to consent. The influence of each attribute was quantified using mixed-effects logistic regression. RESULTS: Similar proportions of patients were willing to have their facilities participate in a trial with high vs. low physician autonomy (77% vs. 79%; p = 0.13) and research burden (76% vs. 80%; p = 0.06). Opt-in, opt-out, and notification-only consent approaches were acceptable to most patients (84%, 82%, and 81%, respectively), but compared to each of these consent approaches, fewer patients (66%) were willing to have their facility participate in a trial that used no notification (p < 0.001 for each 2-way comparison). Among the physicians, similar proportions were willing to participate in trials with high and low physician autonomy (61% and 61%, respectively, p = 0.96) or with low and high burden (60 and 61%, respectively, p = 0.79). However, as for the patients, the consent approach influenced trial acceptability with 77%, 69%, and 62% willing to participate using opt-in, opt-out, and notification-only, respectively, compared to no notification (36%) (p < 0.001 for each 2-way comparison). CONCLUSIONS: Curtailing physician's treatment autonomy and increasing the burden associated with participation did not influence patients' or physicians' willingness to participate in the hypothetical research, suggesting that pragmatic dialysis trials are generally acceptable to patients and physicians. Both patients and physicians preferred consent approaches that include at least some level of patient notification, but the majority of patients were still willing to participate in trials that did not notify patients of the research.

Approximately One In Three US Adults Completes Any Type Of Advance Directive For End-Of-Life Care.

Efforts to promote the completion of advance directives implicitly assume that completion rates of these documents, which help ensure care consistent with people's preferences in the event of incapacity, are undesirably low. However, data regarding completion of advance directives in the United States are inconsistent and of variable quality. We systematically reviewed studies published in the period 2011-16 to determine the proportion of US adults with a completed living will, health care power of attorney, or both. Among the 795,909 people in the 150 studies we analyzed, 36.7 percent had completed an advance directive, including 29.3 percent with living wills. These proportions were similar across the years reviewed. Similar proportions of patients with chronic illnesses (38.2 percent) and healthy adults (32.7 percent) had completed advance directives. The findings provide benchmarks for gauging future policies and practices designed to motivate completion of advance directives, particularly among those people most likely to benefit from having these documents on record.