Sleep Disordered Breathing in Heart Failure with Preserved Ejection Fraction

The 17th Annual Scientific Meeting



HFSA

S79

Table 1. Rates of Healthcare Resource Utilization According to Baseline Anxiety Score Outcome N, % All-cause admission rate, per 100 person-years Heart failure admission rate, per 100 person-years Cardiovascular admission rate, per 100 person-years All-cause in-hospital days, per 100 person-years HF in-hospital days, per 100 person-years Emergency department visits, per 100 person-years

Minimal (0-4)

Mild(5-9)

Moderate or Severe(10-27)

203 (69%) 92

66 (20%) 82

54 (17%) 106*

34

33

37

30

28

40*

426

474y

421

248

330y

225

51

50

54

*P. P.

y

models for healthcare resource utilization rates. Results: Mean age was 57612 years; 64% of patients were male; 50% were white and 47% black; 42% had HF of ischemic etiology. Mean left ventricular ejection fraction was 39615%. At baseline, 203 patients (63%) had minimal anxiety; 66 (20%) had mild anxiety; and 54 (17%) had moderate or severe anxiety. Baseline GAD-7 score was not associated with major clinical events or rate of HF-related admissions. However, patients with moderateto-severe anxiety had (1) 18.5% higher rates of all-cause hospitalizations (P50.049) and (2) 36.1% higher rates of cardiovascular hospitalizations (P50.031) compared to patients with minimal or mild anxiety (Table 1). Patients with mild anxiety spent 11.5% more all-cause days in the hospital compared to those with minimal anxiety (p50.005) as well as 33.0% more days in hospital for HF (P!0.001), Table 1. Conclusion: Anxiety is not associated with major clinical events or HF admissions in stable outpatients with HF. However, patients with moderate-to-severe anxiety have higher rates of all-cause and cardiovascular admissions compared to patients with minimal or mild anxiety. Mild anxiety was associated with more days spent in the hospital compared to minimal anxiety.

229

Month

Total Heart Failure Patients

Number of Eligible Pharmacy Consults

Number of Pharmacy Consults Completed

Discharge Instruction Errors Discovered and Corrected

Total Number of Indigent Patients

Indigent Cost

Sep-11 Oct-11 Nov-11 Dec-11 Jan-12 Feb-12 Mar-12 Apr-12 May-12 Jun-12 Total

36 33 44 46 46 60 64 80 81 73 563

33 25 29 36 32 49 55 60 59 57 435

14 10 20 19 14 27 39 34 42 40 259

3 5 4 4 2 5 7 10 8 12 60

6 3 3 2 0 0 0 6 4 3 27

$87.07 $157.03 $34.42 $4.45 $0.00 $0.00 $0.00 $32.31 $109.98 $22.34 $447.60

228 Anxiety Symptoms, Major Clinical Events, and Healthcare Resource Utilization in Outpatients with Heart Failure in the Atlanta Cardiomyopathy Consortium (TACC) Study Anjan Deka, Song Li, Lampros Papadimitriou, Catherine Marti, Divya Gupta, Robert Cole, Sonjoy Laskar, Andrew L. Smith, Javed Butler, Andreas Kalogeropoulos, Sandra B. Dunbar, Vasiliki Georgiopoulou; Emory University, Atlanta, GA Introduction: Heart failure (HF), a leading cause of hospitalization, is accompanied by increased socioeconomic cost. Identifying risk factors for hospitalization plays a key role in efforts to reduce healthcare costs. Depression has been reported to increase mortality and healthcare resource utilization in patients with HF. Although both anxiety and depression are common comorbidities in these patients, the reports on the role of anxiety have not been conclusive yet. The Generalized Anxiety Disorder (GAD)-7 is a brief validated tool to measure symptoms of anxiety. Hypothesis: We hypothesized that high levels of anxiety are associated with increased mortality and healthcare resource utilization in stable patients with HF. Methods: We evaluated the severity of anxiety by the GAD-7 at baseline and its association with major clinical events (death, heart transplant, left ventricular assist devise) and healthcare resource utilization in 323 outpatients enrolled in a prospective HF cohort study (The Atlanta Cardiomyopathy Consortium). Anxiety symptoms were stratified according to GAD-7 score as minimal (0-4), mild (5-9), or moderate/severe (10-21). We used Cox proportional hazards models for major clinical events and Poisson

Sleep Disordered Breathing in Heart Failure with Preserved Ejection Fraction Kamal O. Shemisa1, Claire Sullivan1, Anupam Basuray2, Neal Sawlani3, Reena Mehra4, Colleen Lance4, Mahazarin Ginwalla2, James C. Fang2; 1University Hospitals Case Medical Center, Cleveland, OH; 2University Hospitals Case Medical Center, Cleveland, OH; 3University of Illinois, Chicago, IL; 4University Hospitals Case Medical Center, Cleveland, OH Background: The relationship between sleep-disordered breathing (SDB) and heart failure has garnered significant interest. SDB appears to be prevalent in over 50% of diastolic heart failure patients and portends a worse prognosis with increasing SDB severity. However, despite the coexistence of SDB with heart failure, there are a limited studies investigating the relationship between SDB and heart failure with preserved ejection fraction (HFpEF). Methods: This is a pilot study of stable ambulatory outpatients with HFpEF. Upon arrival to the sleep laboratory, patients were administered the STOP-Bang questionnaire and underwent overnight polysomnography and PWA/PWV measurements. The following morning we performed repeat PWA/PWV as well as echocardiography. We sought to investigate the relationship between SDB and HFpEF by analyzing data from polysomnography, vascular indices, and echocardiography with strain. Results: A total of 10 patients have been studied thus far in this prospective study. Patients have demonstrated high STOP-BANG questionnaire scores (6 s 2.4) and AHI (39.5 s 21.9). Vascular indices measured during the evening (PM) as compared to morning (AM) showed a trend towards higher mean SBP in the AM (132 s 16.5 in the PM and 138.9 s 22 in the AM). Pulse wave velocities were also higher in the AM (9.7 s 3.5 and 11.2 s 2.4) . Three dimensional echocardiography demonstrated the mean ejection fraction to be 54% (51-58%). LV size was relatively preserved with a normal LV mass index by 3D echo (81.26 g/m2, 49.14-113.37), and the LV end diastolic diameter was 4.7 cm (SD 4.29-5.11). 3D global systolic area strain was mild-moderately reduced at -32.14 (-41.00 to -23.28), suggestive of early systolic dysfunction. Diastolic parameters were abnormal with elevations seen in the estimated left atrial volume index (LAVI) 40.79 (30.38-51.20) as well as E/E’, an estimate of left atrial pressure, 11.25 (6.99-15.52). Conclusions: The aim of this study was to elucidate a pathophysiological relationship between SDB and HFpEF. This particular patient sample (n510) was at a risk for having sleep apnea. Overall findings were consistent with severe sleep apnea (AASM criteria AHI O 30). Vascular diurnal patterns were seen with increased systolic blood pressure and vascular stiffness (by PWV) in the morning relative to the evening and may reflect the consequences of overnight SDB-related physiologic stresses. Although two-tailed t-test did not approach significance, this was likely due to small sample size. Intra-cardiac pressure overload and reductions in global systolic strain seem to be consistent and may be related with the

S80 Journal of Cardiac Failure Vol. 19 No. 8S August 2013 severity of sleep disordered breathing. The results of our study may help determine a link between SDB and HFpEF.

230 Does Pain Influence Readmission Rates in Heart Failure? Meshe Chonde1, Biing-Jiun Shen2, Cam T. Nguyen1, Anita Phan1, Parag Bharadwarj3, Ernst Schwarz4; 1Cedars Sinai Medical Center, Los Angeles, CA; 2 Ohio University, Athens, OH; 3Cedars Sinai Medical Center, Los Angeles, CA; 4 Heart Institute of Southern California, Beverly Hills, CA Introduction: Hospitalizations for acute decompensated heart failure represent a progression of worsening symptoms and quality of life, yet is little known about the relationship of perceived pain and hospital readmission rates. Hypothesis: The manifestation of pain in patients with heart failure is associated with an increased rate of hospital readmissions. Methods: We examined the records of 591 patients admitted to a tertiary care center with a diagnosis of heart failure who had standardized pain scale assessment (0, no pain, to 10, worst pain ever) completed upon admission and evaluated the number of readmissions over one year. Results: The frequency of readmissions was 0.3760.85 for heart failure and 1.1661.59 for any cause, respectively. There was no association between pain and readmission in the total population, however there was significant association in the 211 patients who endorsed pain $1 and hospital length of stay #30 days. Logistic regression analysis demonstrated that pain predicted an increase in the likelihood of readmissions for heart failure within one year, with an Odds Ratio (OR) of 1.23 (p 50.006, 95% CI51.060 - 1.416) independent of coronary artery disease, diabetes, pulmonary disease, renal insufficiency, depression, anemia and depressed left ventricular function. This association also appeared stronger in men, OR 1.26 (p50.026, 95%CI51.0281.553) compared to women, OR 1.20 (p50.09, 95%CI50.973-1.471). Conclusion: The manifestation of pain in heart failure patients is associated with increased rates of readmissions independent of coronary artery disease diabetes, pulmonary disease, renal insufficiency, depression, anemia and depressed left ventricular function. Hospital Readmission for Heart Failure within 1 year in Patients with Perceived Pain HR (95% CI) Pain Age Male gender LVEF CAD DMII Anemia Depression CKD COPD/asthma

1.23 0.99 1.26 1.00 1.05 0.91 0.77 0.52 1.76 0.79

(1.06-1.44) (0.97-1.02) (0.59-2.69) (0.97-1.02) (0.49-2.26) (0.42-1.97) (0.36-1.68) (0.5-5.4) (0.82-3.77) (0.34-1.81)

p Value 0.008 0.545 0.554 0.773 0.902 0.804 0.514 0.590 0.145 0.570

Abbreviations: CI 5 Confidence Interval; HR 5 Hazard ratio. CKD 5 chronic renal disease, COPD 5 chronic obstructive pulmonary disease, DMII 5 Diabetes Mellitus II, LVEF 5 Left ventricular ejection fraction, CAD 5 Coronary Artery Disease.

230A Clinical Effectiveness of CRT and ICD Therapy in Men and Women with Heart Failure: Findings from IMPROVE HF J.E. Wilcox1, G.C. Fonarow2, Yan Zhang3, N.M. Albert4, A.B. Curtis5, M. Gheorghiade1, J.T. Heywood6, M.R. Mehra7, C.M. O’Connor8, M.N. Walsh9, D. Reynolds10, C.W. Yancy1; 1Northwestern, Chicago, IL; 2UCLA, LA, CA; 3 Medtronic, Mound View, MN; 4CCF, Cleveland, OH; 5U of Buffalo, Buffalo, NY; 6 Scripps, La Jolla, CA; 7U of MD, Baltimore, MD; 8DCRI, Durham, NC; 9The Care Group, Indianapolis, IN; 10U of OK, Oklahoma City, OK Background: Clinical trials have demonstrated a clinical benefit for cardiac resynchronization-defibrillator (CRT-D) and implantable cardioverter-defibrillator (ICD) therapies in patients with heart failure and reduced ejection fraction (HFrEF). Recently questions have been raised with regard to the benefit of ICDs for women. The purpose of this study was to determine the clinical effectiveness of CRT and ICD therapy as a function of sex in outpatients with HFrEF (#35%). Methods: IMPROVE HF was a prospective evaluation of a practice-based performance improvement intervention implemented at outpatient practices to increase the use of guideline-recommended care for eligible patients. Data were analyzed by device status and sex among guideline-eligible patients where vital status at 24 months was the outcome of interest. Multivariate GEE analyses of device therapy and sex were conducted adjusting for baseline patient and practice characteristics. Results: In the ICD/ CRT-D eligible cohort (n57748), there were 5,485 (71%) men and 2,261 (29%) women. In the CRT-P/CRT-D eligible cohort (n51188) there were 824 (69%) men and 364 (31%) women. Patients with ICD/CRT-D were less likely to die at 2 years compared to those without (20.4% vs. 27.8%, OR 0.66, 95% CI 0.58-0.74, p! 0.0001). CRT-P/CRT-D therapy showed a similar survival benefit (28.8% vs. 38.3%, OR 0.63, 95% CI 0.48-0.84, p50.0017). The clinical benefit associated with ICD/CRT-D therapy was similar in men and women (men OR 0.64, 95% CI

0.55-0.75, p!0.0001; women OR 0.65, 95% CI 0.53-0.79, p!0.0001). For CRTP/CRT-D, both men and women derived associated benefit (men OR 0.67, 95% CI 0.49-0.92, p50.0133; women OR 0.53, 95% CI 0.31-0.91, p50.0227). Multivariate analysis confirmed clinical benefit of ICD/CRT-D (OR 0.68 95% CI 0.56-0.81, p50.0002) and CRT-P/CRT-D (OR 0.51, 95% CI 0.3-0.87, p50.0096) therapies. Additionally, device by sex interactions were not significant. Conclusions: The use of guideline-directed CRT and ICD therapy was associated with substantially reduced 24 month mortality in eligible men and women with HFrEF. Our data did not show any meaningful differences in clinical effectiveness as a function of sex for either ICD or CRT therapy. These findings have important clinical implications and indicate that device therapies should be offered to all eligible HF patients, without restriction based on sex. Univariate and Multivarate Analysis of Device Therapy with 24 Month Mortality for Men and Women

ICD/CRT-D (n56694) Men (n54941) Women (n52051) CRT-P/CRT-D (n51023) Men (n5705) Women (n5318)

Unadjusted OR (95% CI)

p value

Adjusted OR (95% CI)

p value

Interaction (device*sex) p value

0.66 (0.58-0.74)

!0.0001

0.68(0.56, 0.81)

0.0002

0.5966

0.64 (0.55-0.75) 0.65 (0.53-0.79)

!0.0001 !0.0001

0.71(0.57-0.87) 0.65(0.49-0.85)

0.0012 0.0019

0.63 (0.48-0.84)

0.0017

0.51(0.3-0.87)

0.0096

0.67 (0.49-0.92) 0.53 (0.31-0.91)

0.0133 0.0227

0.59(0.33-1.06) 0.44(0.22-0.90)

0.0793 0.0243

0.4441

231 Is 30-Day Readmission a Valid Measure for Quality Care or Simply an Indicator of Severe Disease in Patients Discharged after Heart Failure Hospitalization? Jennifer L. Nixon, Jose Benuzillo, Raymond O. McCubrey, Deborah Budge, Kismet Rasmusson, Sally Brush, Rami Alharethi, Robert Dyer, Donald Lappe, Colleen Roberts, Abdallah Kfoury; Intermountain Heart Institute, SLC, UT Introduction: “In the interest of promoting high-quality care” CMS requires hospitals to report 30-day all cause readmission and mortality for patients discharged from the hospital with a primary diagnosis of heart failure (HF). Additionally, hospitals with readmission rates that are considered excessive will face financial penalties. High quality care for HF patients includes patient education, adequate diuresis, and care by HF specialists. Whether 30-day readmission and mortality is due to sub-benchmark quality care or is simply a reflection of illness severity is unknown. Hypothesis: Measures of quality care such as weight loss, HF teaching, and care by HF specialists are associated with a lower rate of 30-day readmission and mortality. Methods: Data were obtained for all patients discharged with a primary diagnosis of HF from Intermountain Medical Center from Jan to Dec 2012. All patients with complete information were included in the analysis. Patients were counted in the rehospitalization group if they met primary outcomes of 30-day all cause rehospitalization or mortality. Statistical tests were conducted at the 5% significance level, including chi-square test of association for categorical data and studentized t-test for continuous data. To analyze the effect of length of stay (LOS) on likelihood of readmission, a logistic regression model was used. Results: Of 610 patients, 588 met inclusion criteria. Of these, 134 were readmitted or died within 30 days of hospital discharge. The groups were similar in age and gender. Increased LOS was associated with increased readmission/mortality (Graph1). Readmission/mortality