Development and Optimization of the Veterans Affairs' National Heart Failure Dashboard for Population Health Management.

BACKGROUND: In 2020, the Veterans Affairs (VA) health care system deployed a heart failure (HF) dashboard for use nationally. The initial version was notably imprecise and unreliable for the identification of HF subtypes. We describe the development and subsequent optimization of the VA national HF dashboard. MATERIALS AND METHODS: This study describes the stepwise process for improving the accuracy of the VA national HF dashboard, including defining the initial dashboard, improving case definitions, using natural language processing for patient identification, and incorporating an imaging-quality hierarchy model. Optimization further included evaluating whether to require concurrent ICD-codes for inclusion in the dashboard and assessing various imaging modalities for patient characterization. RESULTS: Through multiple rounds of optimization, the dashboard accuracy (defined as the proportion of true results to the total population) was improved from 54.1% to 89.2% for the identification of HF with reduced ejection fraction (HFrEF) and from 53.9% to 88.0% for the identification of HF with preserved ejection fraction (HFpEF). To align with current guidelines, HF with mildly reduced ejection fraction (HFmrEF) was added to the dashboard output with 88.0% accuracy. CONCLUSIONS: The inclusion of an imaging-quality hierarchy model and natural-language processing algorithm improved the accuracy of the VA national HF dashboard. The revised dashboard informatics algorithm has higher use rates and improved reliability for the health management of the population.

Facility variation and predictors of serum potassium monitoring after initiation of a mineralocorticoid receptor antagonist in patients with heart failure.

BACKGROUND: Mineralocorticoid receptor antagonists (MRAs) have been shown to reduce morbidity and mortality in patients with heart failure (HF) with reduced ejection fraction but are associated with hyperkalemia. We sought to evaluate the frequency, variation, and predictors associated with serum potassium monitoring in patients with HF initiated on an MRA among facilities in the Veterans Affairs (VA) Health Care System. METHODS: We performed a retrospective cohort analysis of patients with HF across 133 Veterans Affairs facilities from 2003 to 2013 who were given a new prescription of an MRA. The primary outcome was the mean percentage of patients per facility with serum potassium monitoring within 14 days of MRA dispensing. Univariate and covariate analyses were performed to determine factors associated with monitoring. RESULTS: There were 142,880 patients identified with HF initiated on an MRA who met the study inclusion and exclusion criteria. The mean (SD) percentage of patients per facility with serum potassium monitoring within 14 days was 41.6% (standard deviation 8.0%; minimum 18.9%, maximum 56.7%). Facilities with a higher frequency of monitoring were associated with membership in the Council on Teaching Hospitals (n = 70, P < .0001), had academic affiliations (n = 100, P < .0001), and a higher annual volume of patients with HF (≥200 patients, P < .0001). CONCLUSIONS: In a large multicenter national sample of patients with HF receiving a new MRA prescription, the frequency of serum potassium monitoring was below recommended guidelines. Academic facilities and those with a higher volume of patients with HF were associated with an increased frequency of monitoring.

Shared Medical Appointments in Heart Failure for Post Acute Care Follow-Up: A Randomized Controlled Trial.

BACKGROUND: Shared medical appointments (SMAs) in heart failure (HF) are medical visits where several patients with HF meet with multidisciplinary providers at the same time for efficient and comprehensive care. It is unknown whether HF-SMAs can improve overall and cardiac health status for high-risk patients with HF discharged from acute care. METHODS AND RESULTS: A 3-site, open-label, randomized-controlled-trial was conducted. Participants within 12 weeks of HF acute care (emergency-room/hospitalization) requiring intravenous diuretic therapy were randomized to receive either HF-SMA or usual HF clinical care (usual-care) on a 1:1 ratio. The HF-SMA team, which consisted of a nurse, nutritionist, psychologist, nurse practitioner and/or a clinical pharmacist, provided four 2-hour session HF-SMAs that met every other week for 8 weeks. Primary outcomes were the overall health status measured by European Quality of Life Visual Analog Scale and cardiac health status by Kansas City Cardiomyopathy Questionnaire, 180 days postrandomization. Of the 242 patients enrolled (HF-SMA n=117, usual-care n=125, mean age 69.3±9.4 years, 71.5% White patients, 94.6% male), 84% of participants completed the study (n=8 HF-SMA and n=9 usual-care patients died). After 180 days, both HF-SMA and usual-care participants had significant improvements from baseline in Kansas City Cardiomyopathy Questionnaire that were not statistically different. Only HF-SMA participants had significant improvements in European Quality of Life Visual Analog Scale (mean change = 7.2±15.8 in HF-SMA versus -0.4±19.0 points in usual-care, P < 0.001). CONCLUSIONS: Both HF-SMA and usual-care in participants with HF achieved significant improvements in cardiac health status, but only a team approach through HF-SMA achieved significant improvements in overall health status. Future larger studies are needed to evaluate hospitalization and death outcomes. REGISTRATION: URL: https://www.clinicaltrials.gov; Unique identifier: NCT02481921.

Preferences of people with advanced heart failure-a structured narrative literature review to inform decision making in the palliative care setting.

UNLABELLED: BACKGROUND AND APPROACH: There is a growing emphasis on the need for high-quality and patient-centered palliative care for patients with heart failure (HF) near end of life. Accordingly, clinicians require adequate knowledge of patient values and preferences, but this topic has been underreported in the HF literature. In response, we conducted a structured narrative review of available evidence regarding patient preferences for HF care near end of life, focusing on circumstances of death, advance care planning, and preferences for specific HF therapies. RESULTS: Patients had widely varying preferences for sudden ("unaware") death versus a death that was anticipated ("aware"), which would allow time to make arrangements and time with family; preferences influenced their choice of HF therapies. Patients and physicians rarely discussed advance care planning; physicians were rarely aware of resuscitation preferences. Advance care planning discussions rarely included preferences for limiting implantable cardioverter defibrillator use, and patients were often uninformed of the option of implantable cardioverter defibrillator deactivation. A substantial minority of patients strongly preferred improved quality of life versus extended survival, but preferences of individuals could not be easily predicted. CONCLUSIONS: Current evidence regarding preferences of patients with HF near end of life suggests substantial opportunities for improvement of end-of-life HF care. Most notably, the wide distribution of patient preferences highlights the need to tailor approach to patient wishes, avoiding assumptions of patient wishes. A research agenda and implications for health care provider training are proposed.

Reduction in body weight but worsening renal function with late ultrafiltration for treatment of acute decompensated heart failure.

OBJECTIVES: The safety, effectiveness and indications for ultrafiltration (UF) are not well established. We hypothesized that UF would not worsen renal function in patients with heart failure (HF) who were not responding to medical therapy. METHODS: Data was collected for patients who underwent UF between 2006 and 2010 (n = 72, median age 61 years, 54% males, 61% Caucasian, 54% left ventricular ejection fraction ≥ 40%). RESULTS: Baseline GFR was 38 ml/min/1.73 m2. All patients were initially treated with loop diuretics and 58% required a thiazide-like diuretic or vasoactive agent. UF resulted in total fluid removal of 11.3 liters and weight loss was 9.7 kg. The median decrease in eGFR during UF was 4.5 ml/min/m2 (IQR--13, 0; p < 0.01) and 43% of patients experienced a ≥ 20% decrease in eGFR. Ten percent of patients required dialysis and 13% died, received a ventricular assist device/cardiac transplant or were discharged to hospice. CONCLUSIONS: In a cohort of HF patients who did not respond to medical therapy, UF was associated not only with a significant reduction of body weight and fluid removal, but also acute worsening of renal function. Further research to identify the appropriate population for UF, long-term outcomes and the intensity of treatment is required if UF is to gain wide acceptance for HF management.

Healthcare providers experiences with shared medical appointments for heart failure.

Shared medical appointments (SMAs) offer a means for providing knowledge and skills needed for chronic disease management to patients. However, SMAs require a time and attention investment from health care providers, who must understand the goals and potential benefits of SMAs from the perspective of patients and providers. To better understand how to gain provider engagement and inform future SMA implementation, qualitative inquiry of provider experience based on a knowledge-attitude-practice model was explored. Semi-structured interviews were conducted with 24 health care providers leading SMAs for heart failure at three Veterans Administration Medical Centers. Rapid matrix analysis process techniques including team-based qualitative inquiry followed by stakeholder validation was employed. The interview guide followed a knowledge-attitude-practice model with a priori domains of knowledge of SMA structure and content (understanding of how SMAs were structured), SMA attitude/beliefs (general expectations about SMA use), attitudes regarding how leading SMAs affected patients, and providers. Data regarding the patient referral process (organizational processes for referring patients to SMAs) and suggested improvements were collected to further inform the development of SMA implementation best practices. Providers from all three sites reported similar knowledge, attitude and beliefs of SMAs. In general, providers reported that the multi-disciplinary structure of SMAs was an effective strategy towards improving clinical outcomes for patients. Emergent themes regarding experiences with SMAs included improved self-efficacy gained from real-time collaboration with providers from multiple disciplines, perceived decrease in patient re-hospitalizations, and promotion of self-management skills for patients with HF. Most providers reported that the SMA-setting facilitated patient learning by providing opportunities for the sharing of experiences and knowledge. This was associated with the perception of increased comradery and support among patients. Future research is needed to test suggested improvements and to develop best practices for training additional sites to implement HF SMA.

Implementation of early follow-up care after heart failure hospitalization.

OBJECTIVES: The evidence supporting early postdischarge hospital follow-up is limited. We implemented a new, multidisciplinary, multistrategy heart failure (HF) team approach that included new clinic slots, predischarge nurse visit, providing a blood pressure cuff and scale, and cardiologist supervision. STUDY DESIGN: Pre- vs postintervention evaluation of outcomes in patients hospitalized with HF between September 1, 2010, and May 30, 2013. We utilized the RE-AIM (reach, effectiveness, adoption, implementation, maintenance) framework to evaluate the intervention. METHODS: For the quantitative evaluation, we compared the proportion of patients in both groups who were scheduled for and completed a cardiology appointment within 7 days after hospitalization ("reach"). We created a Cox model to evaluate the "effectiveness" of the intervention period on a 30-day composite outcome (all-cause emergency department [ED] visit, all-cause hospitalization, or death). In qualitative evaluation, we describe the adoption, implementation, and maintenance of the intervention. RESULTS: Data for 261 patients were analyzed (preintervention, n = 142; post intervention, n = 119). The postintervention period was associated with a higher proportion of patients who were referred to (40% vs 12%; P < .001) and completed (24% vs 10%; P = .003) cardiology follow-up within 7 days of hospital discharge (reach) compared with the preintervention period. After adjustment, the postintervention period was associated with a reduced hazard of the 30-day composite end point (HR, 0.59; 95% CI, 0.37-0.96; P = .04) (effectiveness). CONCLUSIONS: The intervention succeeded in increasing referral to and completion of cardiology appointments within 7 days of discharge. In adjusted analysis, the intervention was associated with lower risk of 30-day all-cause ED visits, all-cause hospitalizations, or death.

Group medical visits after heart failure hospitalization: Study protocol for a randomized-controlled trial.

A primary goal of this research project is to better understand how shared medical appointments (SMAs) can improve the health status and decrease hospitalization and death for patients recently discharged with heart failure (HF) by providing education, disease state monitoring, medication titration, and social support to patients and their caregivers. We propose a 3-site randomized-controlled efficacy trial with mixed methods to test a SMA intervention, versus usual care. Patients within 12 weeks of a HF hospitalization will be randomized to receive either HF-SMA (intervention arm) with optional co-participation with their caregivers, versus usual care (control arm). The HF-SMA will be provided by a non-physician team composed of a nurse, a nutritionist, a health psychologist, a nurse practitioner and/or a clinical pharmacist and will consist of four sessions of 2-h duration that occur every other week for 8 weeks. Each session will start with an assessment of patient needs followed by theme-based disease self-management education, followed by patient-initiated disease management discussion, and conclude with break-out sessions of individualized disease monitoring and medication case management. The study duration will be 180 days for all patients from the day of randomization. The primary study hypothesis is that, compared with usual care, patients randomized to HF-SMA will experience better cardiac health status at 90 and 180 days follow-up. The secondary hypotheses are that, compared to usual care, patients randomized to HF-SMA will experience better overall health status, a combined endpoint of hospitalization and death, better HF self-care behavior, and lower B-type natriuretic peptide levels.

A Description of Patient and Provider Experience and Clinical Outcomes After Heart Failure Shared Medical Appointment.

BACKGROUND: Shared medical appointments (SMAs) are clinical visits in which several patients meet with 1 or more providers at the same time. OBJECTIVE: To describe the outcomes of an interdisciplinary SMA for veterans recently discharged for heart failure (HF). METHODS: A retrospective chart review for patients' readmission rates, survival, medication adherence, and medication-related problems. For qualitative outcomes, we performed semistructured interviews on 12 patients who had undergone HF SMAs and their respective caregivers focusing on care satisfaction, HF knowledge, disease self-care, medication reconciliation, and peer support. RESULTS: The cohort comprised 70 patients-49% had left ventricular function <40% and 50% were prescribed >10 medications. Medication-related problems occurred in 60% of patients. Interviews revealed overall satisfaction with HF-SMA, but patients felt overwhelmed with HF instructions, perceived lack of peer support and self-efficacy, and feelings of hopelessness related to HF. CONCLUSION: Shared medical appointments are well-perceived. Medication problems and need for medication management are prevalent along with patient's lack of self-efficacy in HF care. Multiple HF-SMA visits may be needed to reinforce concepts, reduce confusion, and garner peer support.

Barriers to Adoption of Mineralocorticoid Receptor Antagonists in Patients With Heart Failure: A Mixed-Methods Study.

BACKGROUND: Mineralocorticoid receptor antagonists (MRAs) are the most underutilized pharmacotherapy for heart failure. Minimal data are available on the barriers to MRA adoption from the perspective of prescribing clinicians. METHODS AND RESULTS: A mixed-methods study consisting of a survey (n=50), focus groups (n=39), interviews (n=6) with clinicians at a single US Department of Veterans Affairs medical center served to ascertain barriers to optimal use of MRAs. Participants were drawn from 6 groups: cardiology providers, cardiology fellows, hospitalists, clinical pharmacists, internal medicine residents, and primary care providers. Qualitative data were iteratively coded with qualitative data analysis software. The survey response rate was 17.3%. Overall, 51% of survey respondents were unfamiliar with eplerenone, and 6% were unfamiliar with spironolactone. In addition, 30% of respondents reported that they would order a laboratory test >2 weeks after a new MRA prescription, although that is beyond the guideline recommendation. Most providers correctly identified New York Heart Association class 3 and 4 patients as MRA eligible, but only 42% identified class 2 patients as MRA eligible. Through analysis of focus groups, we identified 8 barriers to MRA use in 3 categories: patient-based barriers (concerns about polypharmacy and comorbidities, adverse effects, perceived patient nonadherence), provider-based barriers (unclear roles and responsibilities, coordination and transitions of care, lack of experience or familiarity with MRAs), and system-based barriers (system overload and provider time constraints, lack of systematic follow-up procedures). CONCLUSIONS: Eight primary barriers to MRA adoption at the provider, patient, and health system levels were identified from the prescriber perspective. These barriers can inform the creation of multilevel interventions that will be required to close the gap in MRA adoption.

Temporal Trends and Hospital Variation in Mineralocorticoid Receptor Antagonist Use in Veterans Discharged With Heart Failure.

BACKGROUND: Despite concerns about mineralocorticoid receptor antagonist therapies (MRAs) underuse and misuse in patients with heart failure, temporal and institutional variations of MRA prescription have not been reported. METHODS AND RESULTS: We studied a national sample of veterans hospitalized for heart failure between 2003 and 2009 and left ventricular ejection fraction <40%. We identified ideal and non-ideal candidates for MRA therapy based on American College of Cardiology/American Heart Association guidelines. We measured temporal trends and hospital variation of MRA prescriptions within 90 days after discharge. We determined the median odds ratio (MOR), a measure of the relative odds of an MRA prescription for 2 individuals with similar characteristics discharged at 2 randomly selected hospitals. From 37 126 patients (n=131 hospitals), 9355 were ideal-MRA candidates, and 4056 were non-ideal candidates. Among ideal candidates, 36% received an MRA, but there was a decline in use (41% in 2003 to 31% in 2009, P<0.001). Of non-ideal candidates, 27% received an MRA with a decline in use (34% in 2003 to 22% in 2009, P<0.001). Hospital MRA prescription ranged from 0% to 71% for ideal candidates and 0% to 100% for non-ideal candidates. The median odds ratios of MRA prescription for ideal and non-ideal candidates were 1.44 and 1.36, respectively; a median odds ratio >1.2 indicates significant practice-level variation. CONCLUSIONS: There was decreasing MRA use between 2003 and 2009 with wide institutional variation in MRA prescription, which suggests opportunities for improvement to stimulate MRA use in ideal candidates while further reducing use in those with contraindications.

Safety and benefit of discontinuing statin therapy in the setting of advanced, life-limiting illness: a randomized clinical trial.

IMPORTANCE: For patients with limited prognosis, some medication risks may outweigh the benefits, particularly when benefits take years to accrue; statins are one example. Data are lacking regarding the risks and benefits of discontinuing statin therapy for patients with limited life expectancy. OBJECTIVE: To evaluate the safety, clinical, and cost impact of discontinuing statin medications for patients in the palliative care setting. DESIGN, SETTING, AND PARTICIPANTS: This was a multicenter, parallel-group, unblinded, pragmatic clinical trial. Eligibility included adults with an estimated life expectancy of between 1 month and 1 year, statin therapy for 3 months or more for primary or secondary prevention of cardiovascular disease, recent deterioration in functional status, and no recent active cardiovascular disease. Participants were randomized to either discontinue or continue statin therapy and were monitored monthly for up to 1 year. The study was conducted from June 3, 2011, to May 2, 2013. All analyses were performed using an intent-to-treat approach. INTERVENTIONS: Statin therapy was withdrawn from eligible patients who were randomized to the discontinuation group. Patients in the continuation group continued to receive statins. MAIN OUTCOMES AND MEASURES: Outcomes included death within 60 days (primary outcome), survival, cardiovascular events, performance status, quality of life (QOL), symptoms, number of nonstatin medications, and cost savings. RESULTS: A total of 381 patients were enrolled; 189 of these were randomized to discontinue statins, and 192 were randomized to continue therapy. Mean (SD) age was 74.1 (11.6) years, 22.0% of the participants were cognitively impaired, and 48.8% had cancer. The proportion of participants in the discontinuation vs continuation groups who died within 60 days was not significantly different (23.8% vs 20.3%; 90% CI, -3.5% to 10.5%; P=.36) and did not meet the noninferiority end point. Total QOL was better for the group discontinuing statin therapy (mean McGill QOL score, 7.11 vs 6.85; P=.04). Few participants experienced cardiovascular events (13 in the discontinuation group vs 11 in the continuation group). Mean cost savings were $3.37 per day and $716 per patient. CONCLUSIONS AND RELEVANCE: This pragmatic trial suggests that stopping statin medication therapy is safe and may be associated with benefits including improved QOL, use of fewer nonstatin medications, and a corresponding reduction in medication costs. Thoughtful patient-provider discussions regarding the uncertain benefit and potential decrement in QOL associated with statin continuation in this setting are warranted. TRIAL REGISTRATION: clinicaltrials.gov Identifier: NCT01415934.

Prevalence, correlates, and temporal trends in antiarrhythmic drug use at discharge after implantable cardioverter defibrillator placement (from the National Cardiovascular Data Registry [NCDR]).

Patients with implantable cardioverter defibrillators (ICDs) can require antiarrhythmic drugs to manage arrhythmias and prevent device shocks. We sought to determine the prevalence, clinical correlates, and institutional variation in the use of antiarrhythmic drugs over time after ICD implantation. From the ICD Registry (2006 to 2011), we analyzed the trends in the use of antiarrhythmic agents prescribed at hospital discharge for patients undergoing first-time ICD placement. The patient, provider, and facility level variables associated with antiarrhythmic use were determined using multivariate logistic regression models. A median odds ratio was calculated to assess the hospital-level variation in the use of antiarrhythmic drugs. Of the cohort (n = 500,995), 15% had received an antiarrhythmic drug at discharge. The use of class III agents increased modestly (13.9% to 14.9%, p <0.01). Amiodarone was the most commonly prescribed drug (82%) followed by sotalol (10%). Among the subgroups, the greatest increase in prescribing was for patients who had received a secondary prevention ICD (26% in 2006% and 30% in 2011, p <0.01) or with a history of ventricular tachycardia (23% to 27%, p <0.01). The median odds ratio for antiarrhythmic prescription was 1.45, indicating that 2 randomly selected hospitals would have had a 45% difference in the odds of treating identical patients with an antiarrhythmic drug. In conclusion, antiarrhythmic drug use, particularly class III antiarrhythmic drugs, is common among ICD recipients at hospital discharge and varies by hospital, suggesting an influence from local treatment patterns. The observed hospital variation suggests a role for augmentation of clinical guidelines regarding the use of antiarrhythmic drugs for patients undergoing implantation of an ICD.

"Not the 'grim reaper service'": an assessment of provider knowledge, attitudes, and perceptions regarding palliative care referral barriers in heart failure.

BACKGROUND: Although similar to cancer patients regarding symptom burden and prognosis, patients with heart failure (HF) tend to receive palliative care far less frequently. We sought to explore factors perceived by cardiology, primary care, and palliative care providers to impede palliative care referral for HF patients. METHODS AND RESULTS: We conducted semistructured interviews regarding (1) perceived needs of patients with advanced HF; (2) knowledge, attitudes, and experiences with specialist palliative care; (3) perceived indications for and optimal timing of palliative care referral in HF; and (4) perceived barriers to palliative care referral. Two investigators analyzed data using template analysis, a qualitative technique. We interviewed 18 physician, nurse practitioner, and physician assistant providers from 3 specialties: cardiology, primary care, and palliative care. Providers had limited knowledge regarding what palliative care is, and how it can complement traditional HF therapy to decrease HF-related suffering. Interviews identified several potential barriers: the unpredictable course of HF; lack of clear referral triggers across the HF trajectory; and ambiguity regarding what differentiates standard HF therapy from palliative care. Nevertheless, providers expressed interest for integrating palliative care into traditional HF care, but were unsure of how to initiate collaboration. CONCLUSIONS: Palliative care referral for HF patients may be suboptimal due to limited provider knowledge and misperceptions of palliative care as a service reserved for those near death. These factors represent potentially modifiable targets for provider education, which may help to improve palliative care referral for HF patients with unresolved disease-related burden.

Comparing unmet needs between community-based palliative care patients with heart failure and patients with cancer.

BACKGROUND: As the role of palliative care (PC) has yet to be clearly defined in patients with heart failure (HF), such patients may face barriers regarding PC referral. In order to maximally meet the needs of HF patients, it is necessary to understand how they compare to the classic PC population: patients with cancer. OBJECTIVE: To characterize the unresolved symptom and treatment needs with which patients with HF and those with cancer present to PC. METHODS: We used data from the Palliative Care Research Registry (PCRR), a repository of quality improvement data from three community-based PC organizations. We abstracted first PC visit data from the PCRR for patients with primary diagnoses of HF or cancer seen between 2008 and 2012. We assessed the association of primary diagnosis (i.e., HF or cancer) on three outcomes: unresolved symptoms, treatment gaps, and a composite indicator of symptom control and quality of life. Analyses included descriptive statistics and multivariate Poisson regression. RESULTS: Our analytic sample comprised 334 patients with HF and 697 patients with cancer, the majority of whom were white and male. Compared to patients with cancer, patients with HF presented with fewer unresolved symptoms, both overall and at moderate/severe distress levels. Patients with HF more commonly reported moderately/severely distressful dyspnea (25% versus 18%, p=0.02), and more commonly experienced dyspnea-related treatment gaps (17% versus 8%, p<0.001). CONCLUSIONS: Patients with HF possess care needs that are squarely within the purview of PC. Future work is needed to delineate how PC referral policies should be refined to optimize PC access for patients with HF.

Link between decisions regarding resuscitation and preferences for quality over length of life with heart failure.

AIMS: Survival preferences, ascertained from time-trade-off utilities, have not been studied in heart failure patients who designate a 'do not resuscitate' (DNR) status. Therefore, the aim of this study was to determine the association of heart failure patients' resuscitation preferences with survival preferences and mortality in the ESCAPE trial. METHODS AND RESULTS: We analysed the association of resuscitation orders at 1 month with time-trade-off utilities and 6-month mortality. There were 26 and 349 patients with a DNR order and Full Code order, respectively. DNR patients were older, had more coronary artery disease, hypertension, renal impairment, and poorer exercise capacity than Full Code patients. DNR patients also experienced longer hospitalization and higher 6-month mortality. In multivariate analysis, DNR preference was associated with 10-fold higher odds of willingness to trade survival time (lower time-trade-off utility) in favour of improved quality of life [odds ratio 10.33, 95% confidence interval (CI) 1.65-64.80]. DNR preference was the best predictor of mortality (χ(2) 26.12, P < 0.0001, hazard ratio 6.88, 95% CI 3.28-14.41), despite adjustment for known predictors including brain natriuretic peptide. CONCLUSIONS: Heart failure patients' requests to forego resuscitation may signify more than simply 'what-if' directives for emergency care. DNR decisions may reflect preferences for intervention to enhance quality rather than prolong survival, which is particularly important as these patients have high early mortality.