Daniel M. Levine

Preclinical Carotid Atherosclerosis in Patients with Rheumatoid Arthritis

Context Patients with rheumatoid arthritis are prone to premature death from coronary heart disease despite few risk factors. Researchers have wondered if chronic inflammation is a trigger. Content The authors measured inflammatory markers and risk factors for coronary heart disease in 98 matched case-patients and controls (mean age, 48 years). Carotid ultrasonography revealed that 44% of case-patients and 15% of controls had atherosclerotic plaque. Independent predictors of plaque were age, smoking, and rheumatoid arthritis. In patients with rheumatoid arthritis, inflammatory mediators did not predict plaque. Limitations This cross-sectional study cannot prove that rheumatoid arthritis accelerates atherosclerosis. Interpretation Carotid atherosclerotic plaque is much more common in patients with rheumatoid arthritis than in controls. The mechanism remains unknown. The Editors Compared with the general population, patients with rheumatoid arthritis die prematurely (1, 2), primarily because of cardiovascular disease (1-3). Women with this disease have high rates of nonfatal myocardial infarction (4-6), even in the absence of traditional risk factors for atherosclerosis (4, 5, 7). Although markers of disease severity have been linked to an increase in overall mortality rates (1), researchers have not been able to clearly identify specific aspects of rheumatoid arthritis or its treatment that might heighten the risk for cardiovascular disease. Use of corticosteroids or disease-modifying antirheumatic drugs does not appear to increase the risk for cardiovascular events (2). In fact, a large longitudinal study recently reported that death rates from myocardial infarction among North American patients with rheumatoid arthritis had declined to the level seen in the general population (thereby yielding a greater magnitude of decline) in the setting of increased methotrexate use (8). In another U.S. study, methotrexate use was associated with lower all-cause mortality rates in rheumatoid arthritis, mostly because cardiovascular mortality rates were decreased (9). Early diagnosis of atherosclerosis in this population might trigger more aggressive prophylaxis, but we have not determined the prevalence of preclinical atherosclerosis or identified markers for the disease. In this study, we examined the prevalence of atherosclerosis in patients with rheumatoid arthritis by using ultrasonogram-defined carotid artery plaque as a direct measure and proxy for generalized atherosclerosis and as a surrogate for coronary atherosclerosis; we also examined those features of rheumatoid arthritis that predict plaque presence. Previous studies reported that plaque prevalence in rheumatoid arthritis is statistically similar to that of control populations (10, 11). However, in systemic lupus erythematosus, another autoimmune disease characterized by chronic inflammation, cross-sectional studies that were conducted by us and by others showed a marked increase in plaque compared with that seen in carefully matched controls (12, 13). The increased plaque rate in systemic lupus erythematosus is associated with chronic inflammation (not with treatment or with traditional atherosclerosis risk factors), suggesting that similar factors may be at work in rheumatoid arthritis. Preclinical disease may also be identified by using ultrasonography to determine carotid intimamedia thickness, an indirect measure of atherosclerosis. Intimamedia thickness was increased in 2 studies of East Asian patients with rheumatoid arthritis (14, 15) but not in a U.S. study (11). Intimamedia thickness varied more with disease duration (14, 15), but an association with serum C-reactive protein levels and erythrocyte sedimentation rate (2 markers of inflammation) has not been established because of conflicting reports (11, 15). Intimamedia thickness does not always correlate with atherosclerosis, particularly in relatively young individuals with chronic inflammatory disease (12, 13), and it may measure other aspects of vascular disease. Discrete atherosclerotic plaque is a potent independent predictor of incident cardiovascular disease, whereas intimamedia thickness in areas free of discrete plaque has limited value as a marker after traditional risk factors for cardiovascular disease are considered (16-18). Because of conflicting data regarding premature preclinical atherosclerosis in rheumatoid arthritis, we chose to use the direct measure of plaque to examine the prevalence of carotid atherosclerosis in consecutive unselected, nonhospitalized patients with rheumatoid arthritis and matched controls. For our other primary outcome, we sought to determine those clinical and biological measures that best predict the presence of plaque. Methods Study Sample We consecutively recruited patients who met the American College of Rheumatologys classification criteria for a diagnosis (possessing at least 4 of 7 criteria) of rheumatoid arthritis (19) and who were enrolled in the Rheumatoid Arthritis Registry at the Hospital for Special Surgery in New York. Patients were recruited at regular visits with their rheumatologists during a 15-month period (participation rate, 94%). Exclusion criteria were age younger than 18 years, serum creatinine level of 270 mol/L or greater (3.0 mg/dL) or creatinine clearance of 0.50 mL/s or less (30 mL/min), or current or recent (within the past 3 months) pregnancy. We quantified extent of disease by recording extra-articular manifestations (for example, the Sjgren syndrome, leg ulcers, and evidence of vasculitis, such as nail fold infarcts, splinter hemorrhages, and motor neuropathy), active joint count (number of tender or swollen joints), number of joints irreversibly damaged (fixed deformity or surgical replacement) (20), and the patients score on the Multidimensional Health Assessment Questionnaire (21). We recorded treatment by patient interview and chart review. Because treatment is often intermittent or at varying dosages, we tabulated medication use as never, former, or current. An 83-year-old woman had a previous stroke that was documented by magnetic resonance imaging, and a 45-year-old man had had coronary artery bypass surgery; we calculated our results both including and excluding these 2 patients. Patients were matched to controls on the basis of age (within 5 years), sex, and ethnicity. Controls were normotensive and hypertensive individuals who participated in longitudinal studies funded by the National Institutes of Health (22, 23) at the Hypertension Center of The New York Hospital and who underwent similar imaging protocols. We assessed patients for traditional cardiovascular disease risk factors: family history of myocardial infarction in first-degree male relatives younger than 55 years of age or first-degree female relatives younger than 65 years of age, smoking, hypertension (defined as blood pressure of 140/90 mm Hg or higher or the prescribed use of antihypertensive medications), diabetes mellitus (self-reported diagnosis), and fasting serum cholesterol levels. Hypertensive controls were studied after antihypertensive medications had been withheld for 3 or more weeks, whereas antihypertensive medications were not systematically withheld in hypertensive patients with rheumatoid arthritis. Brachial blood pressure was obtained at the end of the ultrasonographic studies after patients had remained in the supine position for 45 to 60 minutes in a quiet, darkened room. Of 100 patients with rheumatoid arthritis, a 74-year-old man was unable to be matched to a suitable control and a woman was excluded because she met diagnostic criteria for systemic lupus erythematosus. The institutional review board approved the study protocol, and all participants gave written informed consent. Carotid Ultrasonography All study participants underwent carotid ultrasonography, which was performed by experienced research sonographers who used an identical protocol. A single cardiologist, who was blinded to the identity of the study participants, interpreted the results. In brief, as previously described (22, 23), participants were studied in the supine position with slight hyperextension of the neck. Both extracranial carotid arterial systems were extensively scanned in multiple planes to optimize identification of atherosclerosis, which was defined as discrete plaque protruding into the lumen at least 50% beyond the diameter of the surrounding wall. Doppler interrogation was performed to evaluate the presence of significant (50% diameter reduction) obstruction. Intimamedia thickness was measured from end-diastolic (minimum dimension) M-mode images of the far wall of the distal common carotid artery. Intimamedia thickness was not measured in a location containing plaque. Mean values of right and left intimamedia thicknesses are presented. Reproducibility of intimamedia thickness and detection of plaque has been well documented (24-26). Carotid ultrasonographic studies were performed in the control group before 1999, whereas studies in the patients with rheumatoid arthritis were performed between 2000 and 2002. Laboratory Assessment Laboratory assessment of the patients with rheumatoid arthritis included routine chemistries and serum rheumatoid factor level. A high-sensitivity assay to determine serum levels of C-reactive protein was analyzed with a Cobas Integra system (Roche Diagnostics, Basel, Switzerland). Serum lipoprotein(a) levels were measured with an immunoturbidometric reagent (Diasorin, Stillwater, Minnesota) on a Roche Diagnostics Cobas Fara II system. Serum interleukin-6 levels were measured by automated enzyme immunoassay (Biosource International, Camarilo, California) on a Roche Diagnostics Cobas Core II analyzer. Serum levels of soluble intracellular adhesion molecule-1 and vascular cell adhesion molecule-1 were measured by enzyme-linked immunosorbent assay (Caltag, Burlingame, California, and R&D Systems, Minneapolis, Minnesota,

Arterial Stiffness in Chronic Inflammatory Diseases

Chronic inflammatory diseases are associated with premature atherosclerosis; however, it is unknown whether arterial stiffness is increased in this setting, possibly as a manifestation of vascular disease preceding and/or independent of atherosclerosis. Carotid ultrasonography and radial applanation tonometry were performed in 101 patients with systemic lupus erythematosus, 80 patients with rheumatoid arthritis, and 105 healthy control subjects. The 3 groups were comparable in age, gender, and carotid artery absolute and relative wall thickness. Atherosclerotic plaque was more common in lupus (46%) and rheumatoid arthritis (38%) patients than in controls (23%) (P<0.003). Although control subjects had higher central and peripheral blood pressures, arterial stiffness was increased in patient groups compared with controls (lupus, rheumatoid arthritis, controls, respectively: &bgr;: 3.36 versus 3.22 versus 2.60, P<0.001; Young’s modulus: 441 versus 452 versus 366 mm Hg/cm, P=0.004; Peterson’s elastic modulus: 278 versus 273 versus 216 mm Hg, P<0.001) after adjustment for differences in mean brachial pressure. In multivariate analysis involving the entire population, arterial stiffness was independently related to age, serum glucose, and the presence of chronic inflammatory disease. In multivariate analysis restricted to the patients, arterial stiffness was independently related to age at diagnosis, disease duration, serum cholesterol, and C-reactive protein (and IL-6, when substituted for C-reactive protein). When analyses were repeated in the 186 study subjects without carotid plaque, arterial stiffness remained significantly elevated in patient groups after adjustment for differences in age and mean brachial pressure. In conclusion, arterial stiffness is increased in chronic inflammatory disorders independent of the presence of atherosclerosis and is related to disease duration, cholesterol, and the inflammatory mediator C-reactive protein and the cytokine that stimulates its production, IL-6.

Pravastatin therapy in hyperlipidemia: effects on thrombus formation and the systemic hemostatic profile.

OBJECTIVES The study sought to determine the effects of lipid-lowering with pravastatin on the systemic fibrinolytic profile and on thrombus formation under dynamic flow conditions. BACKGROUND Lowering cholesterol (C) decreases clinical events in coronary artery disease (CAD) patients, but an analysis of the effects of lipid-lowering on the entire hemostatic and thrombotic profile has not been conducted. METHODS We prospectively studied 93 stable patients with untreated low-density lipoprotein cholesterol (LDL-C) >145 mg/dl. The CAD patients received pravastatin, and non-CAD patients were randomized to pravastatin versus placebo (double-blind). Thrombus formation upon an injured vascular surface was assessed in a substudy of 40 patients with a previously validated ex vivo perfusion chamber system. Systemic hemostatic markers and thrombus formation were evaluated at baseline, three and six months. RESULTS Placebo produced no changes in either the lipid profile, any of the hemostatic markers, or the ex vivo thrombus formation. Both pravastatin groups (CAD and non-CAD) showed decreased LDL-C by 30% within 6 weeks (188 to 126 mg/dl, p < 0.001 vs. baseline), and decreased plasminogen activator inhibitor-1 at 3- and 6-month follow-up compared to baseline (15% to 18% decrease at 3 months and 21% to 23% at 6 months). For the tissue plasminogen activator antigen, CAD and non-CAD groups showed significant decreases at 6 months compared to baseline (10% and 13%, respectively). No significant changes were observed with treatment in d-dimer, fibrinopeptide A, prothrombin fragment F1.2, factor VIIa, von Willebrand factor, or C-reactive protein. Fibrinogen levels were significantly increased at 6 months compared to baseline, though still below the upper normal limit. In the perfusion chamber substudy, there was a decrease in thrombus area in non-CAD patients treated with pravastatin at both 3 and 6 months compared to baseline (by 21% and 34%, respectively). The CAD patients showed decreases in thrombus formation by 13% at 3 months, and by 16% at 6 months. The change in LDL-C- correlated modestly with the change in thrombus formation (r = 0.49; p < 0.01). CONCLUSIONS Pravastatin therapy significantly decreased thrombus formation and improved the fibrinolytic profile in patients with and without CAD. These early effects may, in part, explain the benefit rendered in primary and secondary prevention of CAD.

Low lipid concentrations in critical illness : Implications for preventing and treating endotoxemia

OBJECTIVES To determine the prevalence and clinical significance of hypolipidemia found in critically ill patients, and whether the addition of a reconstituted lipoprotein preparation could inhibit the generation of tumor necrosis factor-alpha (TNF-alpha) in acute-phase blood taken from these patients. SETTING Surgical intensive care unit (ICU) of a large urban university hospital. DESIGN Prospective case series. PATIENTS A total of 32 patients with a variety of critical illnesses had lipid and lipoprotein concentrations determined. Six patients and six age- and gender-matched control subjects had whole blood in vitro studies of the effect of lipoprotein on lipopolysaccharide mediated TNF-alpha production. INTERVENTIONS Blood samples were drawn on admission to the ICU and over a subsequent 8-day period. MEASUREMENTS AND MAIN RESULTS Mean serum lipid and lipoprotein values obtained from patients within 24 hrs of transfer to the surgical ICU were extremely low: mean total cholesterol was 117 mg/dL (3.03 mmol/L), low-density lipoprotein cholesterol 71 mg/dL (1.84 mmol/L), and high-density lipoprotein cholesterol 25 mg/dL (0.65 mmol/L). Only the mean triglyceride concentration of 105 mg/dL (1.19 mmol/L), and the mean lipoprotein(a) concentration of 25 mg/dL (0.25 g/L) were within the normal range. During the first 8 days following surgical ICU admission, there were trends toward increasing lipid and lipoprotein concentrations that were significant for triglycerides and apolipoprotein B. Survival did not correlate with the lipid or lipoprotein concentrations, but patients with infections had significantly lower (p = .008) high-density lipoprotein cholesterol concentrations compared with noninfected patients. Lipopolysaccharide-stimulated production of TNF-alpha in patient and control blood samples was completely suppressed by the addition of 2 mg/mL of a reconstituted high-density lipoprotein preparation. CONCLUSIONS Patients who are critically ill from a variety of causes have extremely low cholesterol and lipoprotein concentrations. Correction of the hypolipidemia by a reconstituted high-density lipoprotein preparation offers a new strategy for the prevention and treatment of endotoxemia.

Randomized, Controlled Evaluation of Short- and Long-Term Benefits of Heart Failure Disease Management Within a Diverse Provider Network The SPAN-CHF Trial

Background—Several trials support the usefulness of disease management (DM) for improving clinical outcomes in heart failure (HF). Most of these studies are limited by small sample size; absence of concurrent, randomized controls; limited follow-up; restriction to urban academic centers; and low baseline use of effective medications. Methods and Results—We performed a prospective, randomized assessment of the effectiveness of HF DM delivered for 90 days across a diverse provider network in a heterogeneous population of 200 patients with high baseline use of approved HF pharmacotherapy. During a 90-day follow-up, patients randomized to DM experienced fewer hospitalizations for HF [primary end point, 0.55±0.15 per patient-year alive versus 1.14±0.22 per patient-year alive in control subjects; relative risk (RR), 0.48, P=0.027]. Intervention patients experienced reductions in hospital days related to a primary diagnosis of HF (4.3±0.4 versus 7.8±0.6 days hospitalized per patient-year; RR, 0.54; P<0.001), cardiovascular hospitalizations (0.81±0.19 versus 1.43±0.24 per patient-year alive; RR, 0.57; P=0.043), and days in hospital per patient-year alive for cardiovascular cause (RR, 0.64; P<0.001). Intervention patients showed a trend toward reduced all-cause hospitalizations and total hospital days. On long-term (mean, 283 days) follow-up, there was substantial attrition of the 3-month gain in outcomes, with sustained significant reduction only in days in hospital for cardiac cause. Conclusions—In a population with high background use of standard HF therapy, a DM intervention, uniformly delivered across varied clinical sites, produced significant short-term improvement in HF-related clinical outcomes. Longer-term benefit likely requires more active chronic intervention, even among patients who appear clinically stable.

Relationship of hypolipidemia to cytokine concentrations and outcomes in critically ill surgical patients.

ObjectiveTo determine the relationship of hypolipidemia to cytokine concentrations and clinical outcomes in critically ill surgical patients. DesignConsecutive, prospective case series. SettingSurgical intensive care unit of an urban university hospital. PatientsSubjects were 111 patients with a variety of critical illnesses, for whom serum lipid, lipoprotein, and cytokine concentrations were determined within 24 hrs of admission to a surgical intensive care unit. Controls were 32 healthy men and women for whom serum lipid, lipoprotein, and cytokine concentrations were determined. InterventionsBlood samples were drawn on admission to the intensive care unit. Predetermined clinical outcomes including death, infection subsequent to intensive care unit admission, length of intensive care unit stay, and magnitude of organ dysfunction were monitored prospectively. Measurements and Main Results Measurements included total cholesterol, low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, apolipoproteins A-I and B, phospholipid, triglyceride, interleukin-6, interleukin-10, soluble interleukin-2 receptor, tumor necrosis factor-&agr;, and soluble tumor necrosis factor receptors p55 and p75. Mean serum lipid concentrations were extremely low: total cholesterol, 127 ± 52 mg/dL; low-density lipoprotein cholesterol, 75 ± 41 mg/dL; high-density lipoprotein cholesterol, 29 ± 15 mg/dL. Total, low-density lipoprotein, and high-density lipoprotein cholesterol concentrations and apolipoprotein concentrations inversely correlated with interleukin-6, soluble interleukin-2 receptor, and interleukin-10 concentrations, whereas the triglyceride concentration correlated positively with tumor necrosis factor soluble receptors p55 and p75. Clinical outcomes were related to whether the admission cholesterol concentration was above (n = 56) or below (n = 55) the median concentration of 120 mg/dL. Each of the clinical end points occurred between 1.9- and 3.5-fold more frequently in the very low cholesterol (<120 mg/dL) group. Nine patients (8%) died during the hospitalization. Seven of the nine patients who died had total cholesterol concentrations below the median concentration of 120 mg/dL. ConclusionsLow cholesterol and lipoprotein concentrations found in critically ill surgical patients correlate with interleukin-6, soluble interleukin-2 receptor, and interleukin-10 concentrations and predict clinical outcomes.

Independent association of rheumatoid arthritis with increased left ventricular mass but not with reduced ejection fraction.

OBJECTIVE Rheumatoid arthritis (RA) is a chronic inflammatory disease associated with premature atherosclerosis, vascular stiffening, and heart failure. This study was undertaken to investigate whether RA is associated with underlying structural and functional abnormalities of the left ventricle (LV). METHODS Eighty-nine RA patients without clinical cardiovascular disease and 89 healthy matched controls underwent echocardiography, carotid ultrasonography, and radial tonometry to measure arterial stiffness. RA patients and controls were similar in body size, hypertension and diabetes status, and cholesterol level. RESULTS LV diastolic diameter (4.92 cm versus 4.64 cm; P<0.001), mass (136.9 gm versus 121.7 gm; P=0.004 or 36.5 versus 32.9 gm/m2.7; P=0.01), ejection fraction (71% versus 67%; P<0.001), and prevalence of LV hypertrophy (18% versus 6.7%; P=0.023) were all higher among RA patients versus controls. In multivariate analysis, presence of RA was an independent correlate of LV mass (P=0.004). Furthermore, RA was independently associated with presence of LV hypertrophy (odds ratio 4.14 [95% confidence interval 1.24, 13.80], P=0.021). Among RA patients, age at diagnosis and disease duration were independently related to LV mass. RA patients with LV hypertrophy were older and had higher systolic pressure, damage index scores, C-reactive protein levels, homocysteine levels, and arterial stiffness compared with those without LV hypertrophy. CONCLUSION The present results demonstrate that RA is associated with increased LV mass. Disease duration is independently related to increased LV mass, suggesting a pathophysiologic link between chronic inflammation and LV hypertrophy. In contrast, LV systolic function is preserved in RA patients, indicating that systolic dysfunction is not an intrinsic feature of RA.

The Relationships of Hypocholesterolemia to Cytokine Concentrations and Mortality in Critically Ill Patients with Systemic Inflammatory Response Syndrome

BACKGROUND Decreased concentrations of total cholesterol, lipoproteins, and lipoprotein cholesterols occur early in the course of critical illness. Low cholesterol concentrations correlate with high concentrations of cytokines such as interleukin (IL)-6 and IL-10, and may be due to decreased synthesis or increased catabolism of cholesterol. Low cholesterol concentrations have been associated clinically with several adverse outcomes, including the development of nosocomial infections. The study was performed to test the hypothesis that a low cholesterol concentration predicts mortality and secondarily predicts the development of organ dysfunction in critical surgical illness. METHODS A prospective study was undertaken of 215 patients admitted to a university surgical ICU with systemic inflammatory response syndrome (SIRS). Serial blood samples were collected within 24 h of admission, as well as on the morning of days 2, 4, and 7 of the ICU stay for as long as the patients were in the ICU. Demographic data and predetermined outcomes were noted. RESULTS One hundred nine patients had at least two samples drawn and form the population for analysis. Sixty-two of the patients had three samples obtained, whereas 42 patients had four samples obtained. By univariate analysis, non-survivors were more severely ill on admission (APACHE III), more likely to have been admitted to the ICU as an emergency, more likely to develop a nosocomial infection, and more likely to develop severe organ dysfunction (MODS) (all, p < 0.05). Death was associated on day 1 with increased concentrations of sIL2R, IL-6, IL-10, and sTNFR-p75 (all, p < 0.01), but there were initially no differences in serum lipid concentrations. However, by day 2, concentrations of IL-6, IL-10, and cholesterol had decreased significantly (all, p < 0.05) from day 1 in non-survivors but not in survivors; the difference in serum cholesterol concentration persisted to day 7 (p < 0.05). Persistently elevated concentrations of IL-6 and IL-10 were observed in patients who developed severe MODS. By logistic regression, increased APACHE III score, development of a nosocomial infection, and decreased cholesterol concentration were independently associated with mortality. CONCLUSIONS Decreased serum cholesterol concentration is an independent predictor of mortality in critically ill surgical patients. Repletion of serum lipids is a feasible therapeutic approach for the management of critical illness.

Systemic Lupus Erythematosus Predicts Increased Left Ventricular Mass

Background— Systemic lupus erythematosus (SLE) is associated with premature atherosclerosis and vascular stiffening. Whether SLE alters left ventricular (LV) structure and function in the absence of valvular and clinical coronary artery disease is unknown. Methods and Results— SLE patients without clinical or echocardiographic evidence of valvular or coronary disease were age and gender matched to a reference group (n=173 in both groups). Subjects underwent echocardiography to quantify LV structure and function and carotid ultrasonography to detect atherosclerosis. Disease characteristics and radial applanation tonometry to measure arterial stiffness were evaluated in SLE patients. The 2 groups were similar in subjects’ body size, hypertension and diabetes status, smoking status, and cholesterol levels. LV mass (38.3 versus 32.8 g/m2.7), ejection fraction (71% versus 67%), and prevalence of LV hypertrophy (17.9% versus 6.4%) were higher in SLE patients than in referent subjects (all P<0.001). The combination of SLE and hypertension further increased LV mass. In multivariable analysis, LV mass was associated with SLE (P<0.001) in addition to body mass index, diabetes mellitus, and hypertension. Among SLE patients, LV mass was associated with arterial stiffness (P<0.001). Carotid atherosclerosis, SLE duration, damage index, serum creatinine, and homocysteine were significantly related to LV mass in univariate but not multivariable analyses. Conclusions— SLE predicts increased LV mass, possibly because of inflammation-related arterial stiffening. Excess LV hypertrophy may contribute to the increased cardiac morbidity and mortality observed in SLE patients.

A Multicenter Randomized Controlled Evaluation of Automated Home Monitoring and Telephonic Disease Management in Patients Recently Hospitalized for Congestive Heart Failure: The SPAN-CHF II Trial

BACKGROUND We performed a prospective, randomized investigation assessing the incremental effect of automated health monitoring (AHM) technology over and above that of a previously described nurse directed heart failure (HF) disease management program. The AHM system measured and transmitted body weight, blood pressure, and heart rate data as well as subjective patient self-assessments via a standard telephone line to a central server. METHODS AND RESULTS A total of 188 consented and eligible patients were randomized between intervention and control groups in 1:1 ratio. Subjects randomized to the control arm received the Specialized Primary and Networked Care in Heart Failure (SPAN-CHF) heart failure disease management program. Subjects randomized to the intervention arm received the SPAN-CHF disease management program in conjunction with the AHM system. The primary end point was prespecified as the relative event rate of HF hospitalization between intervention and control groups at 90 days. The relative event rate of HF hospitalization for the intervention group compared with controls was 0.50 (95%CI [0.25-0.99], P = .05). CONCLUSIONS Short-term reductions in the heart failure hospitalization rate were associated with the use of automated home monitoring equipment. Long-term benefits in this model remain to be studied.