Aten Primaria. 2020;52(7):477---487 Atención Primaria www.elsevier.es/ap ORIGINAL ARTICLE Short-term mortality in end-stage heart failure patients Jose Maria Verdú-Rotellara,b,c , Esther Calerob,e,g , Rosa Abellanab,d , José Verdú-Sorianoe , Ernest Vinyolesa,b,d , José Luis del Val-Garcíaa,b , Mar Domingof , Miguel-Angel Muñoza,b,c,∗ a Gerencia Atención Primaria de Barcelona, Institut Català de la Salut, Barcelona, Spain b Primary Health Care University Research Institute Jordi Gol (IDIAPJGol), Barcelona, Spain c School of Medicine, Universitat Autonoma de Barcelona, Bellaterra, Spain d School of Medicine, Universitat de Barcelona, Barcelona, Spain e Community Nursing, Preventive Medicine, Public Health and History of Science Department, Faculty of Health Sciences, University of Alicante, Spain f Heart Failure Unit, Hospital Universitari Germans Trias i Pujol, Institut Català de la Salut, Badalona, Spain g Hospital Universitari de Bellvitge, Institut Català de la Salut, L’Hospitalet de Llobregat, Barcelona, Spain Available online 10 January 2020 KEYWORDS Abstract Heart failure; Objectives: This study is aimed at analyzing the impact of the main factors contributing to End of life; short and long-term mortality in patients at final stages of heart failure (HF). Epidemiology; Setting: Patients attended at any of the 279 primary health care centers belonging to the Prognostic factors Institut Català de la Salut, in Catalonia (Spain). Participants: Patients with Advanced HF. Design: Multicenter cohort study including 1148 HF patients followed for one-year after reach- ing New York Heart Association (NYHA) IV. Main measurements: The primary outcome was all-cause mortality. Multivariate logistic regres- sion models were performed to assess the outcomes at 1, 3, 6, and 12 months. Results: Mean age of patients was 82 (SD 9) years and women represented 61.7%. A total of 135 (11.8%) and 397 (34.6%) patients died three months and one year after inclusion, respectively. Male gender, age, and decreased body mass index were associated with higher mortality at three, six and twelve months. In addition, low systolic blood pressure levels, severe reduction in glomerular filtration, malignancy, and higher doses of loop diuretics were related to higher mortality from 6 to 12 months. The most important risk factor over the whole period was presenting a body mass index lower than 20 kg/m2 (three months OR 3.06, 95% CI: 1.58---5.92; six months OR 4.42, 95% CI: 2.08---9.38; and 12 months OR 3.68, 95% CI: 1.76---7.69). ∗ Corresponding author. E-mail address: mamunoz.bcn.ics@gencat.cat (M.-A. Muñoz). https://doi.org/10.1016/j.aprim.2019.07.019 0212-6567/© 2019 The Authors. Published by Elsevier España, S.L.U. This is an open access article under the CC BY-NC-ND license (http:// creativecommons.org/licenses/by-nc-nd/4.0/). 478 J.M. Verdú-Rotellar et al. Conclusions: We may conclude that male, age, and decreased body mass index determined higher short-term mortality in NYHA IV. In addition, low systolic blood pressure, reduced glomerular filtration, malignancy, and higher doses of loop diuretics contribute to increasing the risk of mortality at medium and long-term. Such variables are easily measurable and can help to decide the best way to face the most advances stages of the disease. © 2019 The Authors. Published by Elsevier España, S.L.U. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). PALABRAS CLAVE Mortalidad a corto plazo en los pacientes con insuficiencia cardiaca en estadios Insuficiencia finales cardíaca; Resumen Fin de vida; Objetivos: Analizar los factores que contribuyen a la mortalidad de pacientes en las etapas Epidemiología; finales de la insuficiencia cardiaca (IC). Factores pronósticos Ámbito: Centros de atención primaria del Institut Català de la Salut, Cataluña, España. Participantes: Pacientes con IC avanzada. Diseño: Estudio de cohortes multicéntrico. Incluyó 1.148 pacientes de IC seguidos durante un año tras el registro de estadio funcional NYHA IV. Mediciones principales: El resultado principal fue la mortalidad por todas las causas. Se realizaron modelos de regresión logística multivariada (1, 3, 6 y 12 meses). Resultados: Edad media 82 años (DE 9), las mujeres representaron el 61,7%. Un total de 135 (11,8%) y 397 (34,6%) pacientes murieron 3 meses y un año después de su inclusión. El sexo masculino, la edad y el índice de masa corporal (IMC) < 20 kg/m2 se asociaron con una mayor mortalidad a los 3, 6 y 12 meses. Bajos niveles de presión arterial sistólica, reducción severa en el filtrado glomerular, malignidad y dosis altas de diuréticos fueron relacionadas con una mortalidad más alta de 6 a 12 meses. El factor de riesgo más importante fue un IMC < 20 kg/m2 (3 meses OR: 3,06; IC 95%: 1,58-5,92; 6 meses OR: 4,42; IC 95%: 2,08-9,38 y 12 meses OR: 3,68; IC 95%: 1,76-7,69). Conclusiones: Los varones, la edad avanzada y un IMC disminuido determinaron una mortalidad a corto plazo más alta en pacientes NYHA IV. La baja presión arterial sistólica, la reducción del filtrado glomerular, la malignidad y las dosis altas de diuréticos aumentan el riesgo de mortalidad a medio y largo plazo. Estas variables son fáciles de obtener, y pueden ayudar a decidir las mejores estrategias para afrontar los estadios más avanzados de la enfermedad. © 2019 Los Autores. Publicado por Elsevier España, S.L.U. Este es un artı́culo Open Access bajo la licencia CC BY-NC-ND (http://creativecommons.org/licenses/by-nc-nd/4.0/). Introduction fluctuate rapidly as the patient’s condition becomes increas- ingly deteriorated. Heart failure (HF) affects more than 10% of those aged over There are no specifically designed models to describe 70 years, and in individuals over 65 it is the major reason for the prognosis of terminal HF patients, consequently, some hospitalization and the third leading cause of death.1,2 The of them underestimate the risk of dying, and often include clinical course of HF is progressive and eventually results in a number of variables not easily obtained in general prac- advanced stages leading to transplantation or death.3 tice. Such models do not thus properly represent most of Although HF symptoms, burden, and mortality are similar the advanced HF population attended in primary care.6---9 to the most prevalent cancers,2,3 little has been published The aim of our study was to identify variables accessible regarding the final stages of advanced HF patients man- in general practice which can determine the impact of the aged mainly in primary care setting. Due to their particular main prognostic factors contributing to mortality at short characteristics, such as old age and comorbidities, most and long-term in patients at final stages of HF. of these patients are not candidates for transplantation or implantable devices, and palliative measures are commonly Methods needed. Accurate knowledge regarding the factors related to a curtailed life expectancy would guide health profession- We carried out a multicenter cohort study based on infor- als, together with patients and families, in deciding where mation from the SIDIAP database (Information System for and when to provide treatment, as well as how to carry the Development of Research in Primary Care System). This out the appropriate palliative and comfort measures.4,5 database contains the daily clinical activities registered At these final stages, the relevance of such issues might by primary care nurses and family physicians. It includes Short-term mortality in end-stage heart failure patients 479 information from over five million patients’ electronic Vienna, Austria). The multiple imputations were carried out records (EMR).10 with the aregImpute function in the Hmisc package. Ethics Patient population All ethical aspects were taken into consideration. The study Study population were patients older than 44 years regis- protocol was approved by the local Clinical Ethic Committee tered as having HF (International Classification Diseases 10th (reference number P13/052). version: I.50) at stage IV of the New York Heart Associa- tion (NYHA) on 31st December, 2013. To be included in the analysis, patients must have been attended at least once during the study period at any of the 279 primary health care centers. Patients were included when they were registered as hav- ing NYHA IV in the patient’s EMR in the period between 1st January, 2010, and 31st December, 2013. All patients had a follow-up for at least one year or until they died during this period. Variables included in the analysis were: socio demo- graphic (gender, age), cardiovascular risk factors and comorbidities (hypertension, diabetes, coronary heart dis- ease, stroke, and atrial fibrillation), other comorbidities (chronic renal disease, chronic pulmonary disease, and can- cer), laboratory tests (sodium, potassium, creatinine, and Results hemoglobin), considering anemia an hemoglobin <13 g/dL in men and 12 g/dL in women, and Hyponatremia if Na < 140 mmol/L, daily living activities assessment, and Descriptive analysis treatment related to HF (beta-blockers (BB), angiotensin- converting enzyme inhibitors (ACEi), angiotensin receptor Baseline characteristics of patients are shown in Table 1. blockers (ARB), mineral corticoid antagonists (MRA), and Among the 1148 patients, women represented 61.7%, mean loop diuretics). In order to perform the analysis ACEi and age was 82 (SD 9) years, and 82.8% were more than 75 ARB were grouped together. years old. The most prevalent cardiovascular comorbidi- Blood pressure, body mass index, and heart rate were ties at baseline were hypertension (79.2%), atrial fibrillation categorized as hypotension (systolic blood pressure (SBP) (46.7%), and diabetes (42.2%). One third of the patients had <90 mmHg), low body mass index (<20 kg/m2 ), and high heart a history of coronary heart disease and 14.3% had had a rate (>100 b/min), respectively. previous stroke. Among the non-cardiovascular comorbidi- The main outcome was all-cause mortality (at one, three, ties, the most frequent was chronic kidney disease (34.1%). six, and twelve months) after NHYA IV registration in EMR. The most commonly prescribed medication were ACEi/ARB (66.6%) followed by BB (45%), and MRA (26.5%). Women were more commonly older, hypertensive, and Statistical analysis had atrial fibrillation whilst men suffered more frequently from coronary heart disease, chronic obstructive pulmonary Continuous variables were expressed as mean and stan- disease, and cancer. Mineral corticoid receptors antagonist dard deviation while categorical variables were described as were more frequently prescribed in men and no differences total number and percentages. Univariate analysis between were found regarding the rest of the medication. patients’ characteristics (socio-demographic, comorbidi- Fig. 1 depicts the evolution in mortality during the first ties, clinical and analytic information, daily living activities, year after reaching NYHA IV. It can be observed that the and medical treatment related to HF) and mortality outcome slope of the curve was higher in the first six months. were performed using student’s t test and chi-square test or Fisher’s exact test as appropriate. The unadjusted odds ratio Univariate analysis was also computed. Multivariate logistic regression analyses were performed to analyze the association of mortality (at During the first year mortality constantly increased (4.2% at one, three, six, and twelve months after reaching the NYHA 1 month, 11.8% at 3 months, 22.6% at 6 months, and 34.6% IV stage of HF) and the patients’ characteristics. A multi- at 12 months, respectively). ple imputation with the predictive mean matching method Cardiovascular comorbidity did not significantly affect was carried out to deal with the missing values of the varia- mortality at any time during the first year after reaching bles considered in the multivariate model. To generate the NYHA IV. Only age and hyponatremia were related to higher imputations all the patients’ information was taken into mortality at very early stages (one month). From the sixth account. Twenty imputations were generated and the values month onwards, male gender, age, high dependency on daily combined using Rubin’s rules.11 The significance level was activities, cancer, hyponatremia, chronic renal disease, low fixed at 0.05. Analysis was performed using R software for SBP, and decreased body mass index were associated with a Windows version 3.3.2 (R project for statistical computing; higher risk of mortality. Reduced levels of hemoglobin and 480 J.M. Verdú-Rotellar et al. Table 1 Baseline characteristics of heart failure patients. Total N = 1148 Women N = 708 Men N = 440 p Value Age (%) <0.001 <50 3 (0.26) 2 (0.28) 1 (0.23) 50 to 74 195 (17.0) 90 (12.7) 105 (23.9) ≥75 950 (82.8) 616 (87.0) 334 (75.9) Age (years) 81.6 (8.87) 83.1 (8.42) 79.2 (9.04) <0.001 Smoking habit (%) Non-smoker 838 (74.2) 648 (93.4) 190 (43.7) <0.001 Active smoker 41 (3.63) 8 (1.15) 33 (7.59) Former smoker 250 (22.1) 38 (5.48) 212 (48.7) Barthel <20 (1 to 100) (%) 90 (12.3) 67 (14.5) 23 (8.46) 0.022 Comorbidity Hypertension (%) 909 (79.2) 594 (83.9) 315 (71.6) <0.001 Diabetes (%) 485 (42.2) 286 (40.4) 199 (45.2) 0.121 Coronary heart disease (%) 375 (32.7) 184 (26.0) 191 (43.4) <0.001 Stroke (%) 164 (14.3) 91 (12.9) 73 (16.6) 0.094 Atrial fibrillation (%) 536 (46.7) 356 (50.3) 180 (40.9) 0.002 Chronic renal disease (%) 391 (34.1) 253 (35.7) 138 (31.4) 0.146 Chronic pulmonary disease (%) 378 (32.9) 145 (20.5) 233 (53.0) <0.001 Cancer (%) 216 (18.8) 114 (16.1) 102 (23.2) 0.004 Clinical variables Systolic blood pressure ≤90 mmHg (%) 37 (3.46) 18 (2.71) 19 (4.67) 0.127 Body mass index <20 kg/m2 (%) 36 (4.21) 19 (3.75) 17 (4.87) 0.532 Heart rate > 100 min---1 (%) 55 (5.41) 33 (5.25) 22 (5.67) 0.887 Charlson ≥ 6 (%) 155 (13.5) 77 (10.9) 78 (17.7) 0.001 Glomerular filtration <30 mL/min (%) 97 (11.9) 63 (12.4) 34 (11.1) 0.675 Anemia (<13 g/dL in men and <12 g/dL in women) 421 (53.8) 245 (49.9) 176 (60.3) 0.006 Hyponatremia (<140 mmol/L) 234 (30.7) 143 (29.7) 91 (32.4) 0.493 Laboratory tests Sodium (mg/dL) 141 (3.62) 141 (3.61) 141 (3.63) 0.337 Potassium (mg/dL) 4.53 (0.55) 4.50 (0.53) 4.57 (0.58) 0.113 Hemoglobin (g/dL) 12.2 (1.80) 12.0 (1.68) 12.5 (1.94) <0.001 Creatinine (mg/dL) 1.24 (0.60) 1.15 (0.51) 1.39 (0.71) <0.001 Medication ACEi/ARB *(%) 758 (66.0) 478 (67.5) 280 (63.6) 0.199 Beta blockers (%) 517 (45.0) 325 (45.9) 192 (43.6) 0.490 MRAs† (%) 304 (26.5) 162 (22.9) 142 (32.3) 0.001 Doses of loop diuretics (mg) 55.7 (44.3) 54.5 (42.5) 57.8 (47.1) 0.243 Values are given as mean (standard deviation) or frequency (percentage). * ACEi/ARB indicates angiotensin converting enzyme inhibitor/Angiotensin II receptor blockers. † MRAs, Mineralocorticoid Receptor Antagonists. raised loop diuretic doses were related to mortality at six Three months after reaching NYHA IV the main factors and twelve months (Table 2). related to higher mortality were male gender, increased age, and reduced body mass index. Six months after reaching NYHA IV the effect of the former variables remained signif- Multivariate analysis icant and other variables presented statistical significance in the multivariate model, such as reduced glomerular fil- Some variables had missing values, especially the Barthel tration, low SBP, cancer, and the loop diuretic doses used to index (36%), hemoglobin levels (32%), and glomerular filtra- manage HF. tion rate (29%). We present the final adjusted multivariate models after performing multiple missing imputation. In the first month the prescription of ACE inhibitors or ARB Discussion resulted in a protective effect (OR 0.46 95% CI 0.26---0.84). This effect remained along the first year but was not statis- Our study, based on real world data from clinical records, tically significant from the third month onwards (Fig. 2). adds information to previous evidence about HF prognosis. Short-term mortality in end-stage heart failure patients 481 Figure 1 Evolution in the cumulate incidence of mortality during the first year after reaching New York Heart Association IV (N = 1148). It describes the evolution and impact short-term, and at one Comparative analysis of mortality determinants year, of the main prognostic factors after reaching NYHA IV, and is specifically focused on stable, advanced HF patients Regarding gender, it has already been argued that women managed in primary care. At three months, only age, male present higher comorbidity (with the exception of coro- gender, and reduced body mass index resulted in higher nary heart disease such as HF) and symptom severity than mortality. At six and 12 months these factors were still men, but better survival. It is plausible to think that related to higher mortality with the addition of decreased this could have influenced prognosis, nevertheless, it does glomerular filtration, low SBP, cancer, and higher doses of not completely explain differences in mortality between furosemide. genders.20 A number of scores have been developed to predict the Patients with SBP levels lower than 90 mmHg had a more prognosis of HF patients, in most of them, however, survival than two-fold risk of death in comparison with the oth- was analyzed at long-term (one year or more), partici- ers, as previously observed by Vidán et al. They justified pants were recruited in hospitals, the proportion of NYHA this result by the fact that patients with higher SBP on IV was lower than 10%, and many findings were based on hospital admission were younger, healthier, and received variables generally complicated to obtain in non-hospital more aggressive treatment. In addition, the pathophysiology settings.12---15 With respect to mortality, the PRAISE study might differ in patients with HF and higher blood pressure.21 was conducted in patients with NYHA IIIb and IV and showed Similar results were also reported in advanced chronic sys- an annual rate of 11%. The CHARM study included a sample tolic HF patients although the SBP cut-off was 120 mm Hg. of 173 patients with NYHA IV and observed a figure of 29%, The authors hypothesized that their findings could be related which was closer to our findings.16,17 Ambardekaer et al. also to a greater incidence of ischemic heart disease in patients reported a high rate of mortality after a 16-month follow-up with lower blood pressure.22 In our study such an effect in a cohort of ambulatory advanced HF patients who were on mortality was found to be independent of age, treat- not eligible for transplantation/left ventricular assisted ment, and comorbidity. The latest version of the European device.18 Currently, morbidity and mortality for individuals Society of Cardiology Clinical Practice Guidelines includes with advanced HF continues to be elevated and comorbid- hypotension as a factor of poor ischemic cardiomyopathy ity common, since most patients are elderly, as previously prognosis and proposes these patients be candidates for the described.19 implantation of assistive devices.23 482 Table 2 Risk factors related to mortality at 1, 3, 6, and 12 months. N Mortality 1 month Mortality 3 months Mortality 6 months Mortality 1 year N (%) p Value N (%) p Value N (%) p Value N (%) p Value Sex 0.738 0.043 0.055 0.002 Women 708 28 (3.95) 72 (10.2) 146 (20.6) 220 (31.1) Men 440 20 (4.55) 63 (14.3) 113 (25.7) 177 (40.2) Age (1 year) 1148 84.3 (8.95) 0.039 84.8 (8.19) <0.001 84.3 (8.25) <0.001 83.7 (8.35) <0.001 Age (two cathegories) 45---74 years 198 6 (3.0) 0.048 13 (6.5) 0.018 26 (13.1) 0.006 47 (23.7) 0.001 ≥75 years 950 42 (4.4) 122 (12.8) 233 (24.5) 350 (36.8) Smoking habit 0.837 0.156 0.469 0.935 Non-smoker 838 35 (4.2) 101 (12.1) 194 (23.2) 292 (34.8) Active smoker 41 1 (2.4) 1 (2.44) 7 (17.1) 14 (34.1) Former smoker 250 12 (4.8) 31 (12.4) 51 (20.4) 84 (33.6) Comorbidity (%) Barthel index 0.070 0.194 0.012 0.032 ≥20 644 28 (4.3) 85 (13.2) 159 (24.7) 242 (37.6) <20 90 8 (8.9) 17 (18.9) 34 (37.8) 45 (50.0) Hypertension 0.584 0.589 0.426 0.176 No 239 12 (5.0) 31 (13.0) 59 (24.7) 92 (38.5) Yes 909 36 (3.9) 104 (11.4) 200 (22.0) 305 (33.6) Diabetes 0.259 0.032 0.323 0.247 No 663 32 (4.8) 90 (13.6) 157 (23.7) 239 (36.0) Yes 485 16 (3.3) 45 (9.28) 102 (21.0) 158 (32.6) Coronary heart disease 0.955 0.755 0.640 0.876 No 773 33 (4.3) 93 (12.0) 178 (23.0) 269 (34.8) Yes 375 15 (4.0) 42 (11.2) 81 (21.6) 128 (34.1) Stroke 0.125 0.751 0.920 1.000 No 984 37 (3.7) 114 (11.6) 221 (22.5) 340 (34.6) Yes 164 11 (6.7) 21 (12.8) 38 (23.2) 57 (34.8) J.M. Verdú-Rotellar et al. Atrial fibrillation 0.748 0.787 0.517 0.607 No 612 24 (3.9) 70 (11.4) 133 (21.7) 207 (33.8) Yes 536 24 (4.5) 65 (12.1) 126 (23.5) 190 (35.4) Chronic renal disease 0.720 0.042 0.010 0.008 No 757 30 (3.9) 78 (10.3) 153 (20.2) 241 (31.8) Yes 391 18 (4.6) 57 (14.6) 106 (27.1) 156 (39.9) Short-term mortality in end-stage heart failure patients Table 2 (Continued) N Mortality 1 month Mortality 3 months Mortality 6 months Mortality 1 year N (%) p Value N (%) p Value N (%) p Value N (%) p Value COPD* 1.000 0.992 1.000 0.918 No 770 32 (4.1) 90 (11.7) 174 (22.6) 265 (34.4) Yes 378 16 (4.2) 45 (11.9) 85 (22.5) 132 (34.9) Cancer 0.860 0.096 0.002 <0.001 No 932 38 (4.1) 102 (10.9) 193 (20.7) 296 (31.8) Yes 216 10 (4.6) 33 (15.3) 66 (30.6) 101 (46.8) Anemia 0.951 0.150 0.006 <0.001 No 362 15 (4.1) 36 (9.9) 68 (18.8) 105 (29.0) Yes 421 16 (3.8) 57 (13.5) 115 (27.3) 176 (41.8) Hyponatremia 0.006 0.094 0.045 0.009 No 528 14 (2.6) 57 (10.0) 110 (20.8) 170 (32.2) Yes 234 17 (7.2) 34 (14.5) 65 (27.8) 99 (42.3) Total Mortality 1 month Mortality 3 months Mortality 6 months Mortality 1 year N N (%) p Value N (%) p Value N (%) p Value N (%) p Value Clinical variables (%) Systolic blood pressure 0.070 0.072 0.001 0.006 >90 mmHg 1033 42 (4.07) 120 (11.6) 227 (22.0) 347 (33.6) ≤90 mmHg 37 4 (10.8) 8 (21.6) 17 (45.9) 21 (56.8) Body mass index 0.137 <0.001 <0.001 <0.001 ≥20 kg/m2 819 28 (3.42) 74 (9.04) 149 (18.2) 244 (29.8) <20 kg/m2 36 3 (8.33) 11 (30.6) 19 (52.8) 23 (63.9) Heart rate 0.728 0.207 0.294 0.809 ≤100 min---1 961 41 (4.27) 111 (11.6) 212 (22.1) 325 (33.8) >100 min---1 55 3 (5.45) 10 (18.2) 16 (29.1) 20 (36.4) Charlson 0.384 0.942 0.078 0.012 <6 993 39 (3.93) 116 (11.7) 215 (21.7) 329 (33.1) ≥6 155 9 (5.81) 19 (12.3) 44 (28.4) 68 (43.9) 483 484 Table 2 (Continued) Total Mortality 1 month Mortality 3 months Mortality 6 months Mortality 1 year N N (%) p Value N (%) p Value N (%) p Value N (%) p Value Glomerular filtration 1.000 0.032 0.001 <0.001 ≥30 mL/min 719 29 (4.03) 76 (10.6) 150 (20.9) 233 (32.4) <30 mL/min 97 3 (3.09) 18 (18.6) 36 (37.1) 52 (53.6) Medication (%) ACEi/ARB† 0.004 0.040 0.009 <0.001 No 390 26 (6.67) 57 (14.6) 106 (27.2) 162 (41.5) Yes 758 22 (2.90) 78 (10.3) 153 (20.2) 235 (31.0) Beta blockers 0.741 0.672 0.085 0.016 No 631 28 (4.44) 77 (12.2) 155 (24.6) 238 (37.7) Yes 517 20 (3.87) 58 (11.2) 104 (20.1) 159 (30.8) MRAs‡ 0.550 1.000 0.759 0.450 No 844 33 (3.91) 99 (11.7) 188 (22.3) 286 (33.9) Yes 304 15 (4.93) 36 (11.8) 71 (23.4) 111 (36.5) Doses of loop diuretics (Mean, SD§ ) 1148 57.9 (49.4) 0.763 60.0 (48.9) 0.287 61.7 (48.8) 0.025 60.8 (47.3) 0.008 Laboratory tests (Mean, SD) Sodium (mg/dL) 1148 140 (4.12) 0.094 140 (4.13) 0.512 141 (4.08) 0.361 140 (3.99) 0.143 Potassium (mg/dL) 1148 4.53 (0.59) 0.982 4.48 (0.63) 0.466 4.54 (0.62) 0.799 4.56 (0.59) 0.304 Creatinine (mg/dL) 1148 1.25 (0.58) 0.932 1.39 (0.69) 0.026 1.36 (0.73) 0.007 1.34 (0.65) 0.001 Hemoglobin (g/dL) 1148 12.1 (2.02) 0.869 11.8 (1.91) 0.083 11.8 (1.90) 0.003 11.9 (1.89) 0.001 * Chronic Obstructive Pulmonary Disease. † ACEi/ARB indicates angiotensin converting enzyme inhibitor/Angiotensin II receptor blockers. ‡ MRAs, Mineralocorticoid Receptor Antagonists. § J.M. Verdú-Rotellar et al. Standard deviation. Short-term mortality in end-stage heart failure patients 485 Figure 2 Evolution of the impact of factors related to mortality in patients with heart failure at stage IV of the New York Heart Association. In relation to the patients’ weight, in our study, the risk Although it may appear obvious that HF patients requiring of dying for patients with a body mass index <20 kg/m2 was higher loop diuretic doses might be more unwell than those more than three times higher than in the rest. Considerable receiving lower ones, the powerful and independent asso- research has been devoted to the role body mass index plays ciation with mortality warrants further consideration. Loop in HF patients.24 It has been hypothesized that individuals diuretic activation of the renin-angiotensin-aldosterone and with chronic HF and obesity have significantly lower sym- sympathetic nervous systems has been shown to play a major pathetic activation. This may partially explain the obesity role in the pathophysiology of HF and may be associated with paradox described in chronic HF patients.25 The higher mor- HF progression.30 tality related to low body mass index has been associated The protective effect on early mortality of ACEi/ARB dis- with higher Tumoral Necrosis Factor-alfa, adiponectin, tro- appeared along the follow up. We cannot ascertain why this ponin T and pulmonary arterial systolic pressure.26 Whilst effect is not maintained, but probably may be due to the we could not measure either cachexia or sarcopenia in fact of the advanced stage of the disease, where could be our cohort we did make an approximation of the obesity hard to lengthen the survival with any intervention. effect by measuring body mass index. It has been shown An association between anemia and increased all- to accurately correlate with the thickness of the subcuta- cause and short-term cardiovascular mortality has been neous fat layer and is accepted as a measure of obesity reported.31 In our study, the trend regarding anemia was in HF.27 toward a higher risk although it was not statistically signifi- Most HF prognostic scores consider glomerular filtration cant at multivariate analysis. as an independent factor of death. We found an almost two-fold odds ratio of mortality in patients presenting a glomerular filtration <30 mL/min. When attempts to alle- Strengths and limitations viate congestion are limited by impaired renal function the cardiorenal syndrome appears.28 It consists of a neuro Our main limitation is the retrospective design and the fact hormonal adaptation leading to renal-reduced perfusion that data were gathered from electronic medical records. and right ventricular dysfunction. Its prevalence is around Nevertheless, SIDIAP has already proven to be a valid source 60% in patients with acute decompensation and up to 51% for cardiovascular disease research.10 Some variables pre- mortality has been reported when glomerular filtration is sented missing values which were imputed in order to avoid <53 mL/min.29 substantial loss of precision and power in the statistical 486 J.M. Verdú-Rotellar et al. analysis. A strength is the large cohort of patients with sta- failure: a scientific statement from the American Heart Asso- ble, advanced HF managed in the community. ciation. Circulation. 2012;125:1928---52. It would be necessary further studies to properly explain 4. Blinderman CD, Homel P, Billings JA, Portenoy RK, Tennst- some of the findings, such as the role of ACEi/ARB on the edt SL. Symptom distress and quality of life in patients with survival of patients at advances stages of HF. advanced congestive heart failure. J Pain Symptom Manag. 2008;35:594---603. Nevertheless the initial objective of our study was not to 5. Pastor DK, Moore G. Uncertainties of the heart: palliative care test the efficacy of treatments in enlarging the survival but and adult heart failure. Home Healthc Nurse. 2013;31:29---36. to identify factors related with a poorer prognosis in order 6. Allen LA, Matlock DD, Shetterly SM, Xu S, Levy WC, Portalupi to be able to discuss with the patients and their families the LB, et al. Use of risk models to predict death in the next year best way to face the end of life. among individual ambulatory patients with heart failure. JAMA Cardiol. 2017;2:435---41. 7. Levy WC, Mozaffarian D, Linker DT, Sutradhar SC, Anker SD, Conclusions Cropp AB, et al. The Seattle Heart Failure Model: prediction of survival in heart failure. Circulation. 2006;113:1424---33. Among stable, ambulatory HF patients, male gender, age, 8. Chyu J, Fonarow GC, Tseng CH, Horwich TB. Four-variable risk and low body mass index are involved in higher mortality model in men and women with heart failure. Circ Heart Fail. 2014;7:88---95. at both short and long-term after reaching NYHA IV stages. 9. Goda A, Williams P, Mancini D, Lund LH. Selecting patients for In addition, low SBP, reduced glomerular filtration, malig- heart transplantation: comparison of the Heart Failure Survival nancy, and higher necessity in the doses of loop diuretics Score (HFSS) and the Seattle Heart Failure Model (SHFM). J contribute to increasing the risk of mortality at medium and Heart Lung Transplant. 2011;30:1236. long-term. Such variables are easily measurable in primary 10. Ramos R, Balló E, Marrugat J, Elosua R, Sala J, Grau M, care and may help health professionals identify HF patients et al. Validity for use in research on vascular diseases of the with worse outcomes and decide with patients and their SIDIAP (Information System for the Development of Research families the best way to face the most advanced stages of in Primary Care): the EMMA study. Rev Esp Cardiol (Engl Ed). the disease. 2012;65:29---37. 11. Rubin RB. Multiple imputations for nonresponse in surveys. New York: John Wiley and Sons; 1987. 12. Lanfear DE, Levy WC, Stehlik J, Estep JD, Rogers JG, Shah KB, et al. Accuracy of seattle heart failure model and heart mate Key Points II risk score in non-inotrope-dependent advanced heart failure patients: insights from the ROADMAP study (risk assessment and comparative effectiveness of left ventricular assist device and What is known about the subject/What does this medical management in ambulatory heart failure patients). Circ study contribute Heart Fail. 2017;10(5.), pii: e003745. 13. Vazquez R, Bayes-Genis A, Cygankiewicz I, Pascual-Figal D, 1. Clinical course of HF is progressive and, in advanced Grigorian-Shamagian L, Pavon R, et al. The MUSIC Risk score: a stages leads to a number of hospitalizations, trans- simple method for predicting mortality in ambulatory patients with chronic heart failure. Eur Heart J. 2009;30:1088---96. plantation or death. 14. Harjola VP, Follath F, Nieminen MS, Brutsaert D, Dickstein K, 2. Palliative measures can help to better manage Drexler H, et al. Characteristics, outcomes, and predictors of patients at final stages. mortality at 3 months and 1 year in patients hospitalized for 3. Since the uncertainty of the evolution at final stages acute heart failure. Eur J Heart Fail. 2010;12:239---48. it is difficult to ascertain which patients could ben- 15. Lagu T, Pekow PS, Shieh MS, Stefan M, Pack QR, Kashef MA, et al. efit from homecare and palliative measures. Validation and comparison of seven mortality prediction models 4. To know the factors related to a worse short term for hospitalized patients with acute decompensated heart fail- prognostic may help to take decisions about their ure. Circ Heart Fail. 2016;9(8.), pii: e002912. care. 16. Packer M, O’Connor CM, Ghali JK, Pressler ML, Carson PE, Belkin RN, et al. Effect of amlodipine on morbidity and mor- tality in severe chronic heart failure. N Engl J Med. 1996;335: 1107---14. 17. Pocock SJ, Wang D, Pfeffer MA, Yusuf S, McMurray JJ, Swedberg KB, et al. Predictors of mortality and morbidity in patients with References chronic heart failure. Eur Heart J. 2006;27:65---75. 18. Ambardekaer AV, Forde-McLean RC, Kittleson MM, Stewart GC, 1. Bader F, Atallah B, Brennan LF, Rimawi RH, Khalil ME. Heart Palardy M, Thibodeau JT, et al. High early event rates in patients failure in the elderly: ten peculiar management considerations. with questionable eligibility for advanced heart failure ther- Heart Fail Rev. 2017;22:219---28. apies: results from the medical arm of mechanically assisted 2. Sayago-Silva I, García-López F, Segovia-Cubero J. Epidemiology circulatory support (Medamacs) registry. J Heart Lung Trans- of heart failure in Spain over the last 20 years. Rev Esp Cardiol. plant. 2016;35:722---30. 2013;66:649---56. 19. Ahluwalia SC, Gross CP, Chaudhry SI, Leo-Summers L, Van Ness 3. Allen LA, Stevenson LW, Grady KL, Goldstein NE, Matlock DD, PH, Fried TR. Change in comorbidity prevalence with advanc- Arnold RM, et al., American Heart Association; Council on Qual- ing age among persons with heart failure. J Gen Intern Med. ity of Care and Outcomes Research; Council on Cardiovascular 2011;26:1145---51. Nursing; Council on Clinical Cardiology; Council on Cardiovas- 20. Frazier CG, Alexander KP, Newby LK, Anderson S, Iverson E, cular Radiology and Intervention; Council on Cardiovascular Packer M, et al. Associations of gender and etiology with out- Surgery and Anesthesia. Decision making in advanced heart comes in heart failure with systolic dysfunction: a pooled Short-term mortality in end-stage heart failure patients 487 analysis of 5 randomized control trials. J Am Coll Cardiol. failure: The role of sympathetic activation. Int J Cardiol. 2007;49:1450---8. 2015;181:120---6. 21. Vidán MT, Bueno H, Wang Y, Schreiner G, Ross JS, Chen J, et al. 26. Takiguchi M, Yoshihisa A, Miura S, Shimizu T, Nakamura Y, The relationship between systolic blood pressure on admission Yamauchi H, et al. Impact of body mass index on mortality in and mortality in older patients with heart failure. Eur J Heart heart failure patients. Eur J Clin Invest. 2014;44:1197---205. Fail. 2010;12:148---55. 27. Keys A, Fidanza F, Karvonen MJ, Kimura N, Taylor HL. Indices of 22. Banach M, Bhatia V, Feller MA, Mujib M, Desai RV, Ahmed MI, relative weight and obesity. Int J Epidemiol. 2014;43:655---65. et al. Relation of baseline systolic blood pressure and long-term 28. Bock JS, Gottlieb SS. Cardiorenal syndrome: new perspectives. outcomes in ambulatory patients with chronic mild to moderate Circulation. 2010;121:2592---600. heart failure. Am J Cardiol. 2011;107:1208---14. 29. Smith GL, Lichtman JH, Bracken MB, Shlipak MG, Phillips CO, 23. Ponikowski P, Voors AA, Anker SD, Bueno H, Cleland JG, Coats DiCapua P, et al. Renal impairment and outcomes in heart fail- AJ, et al. 2016 ESC Guidelines for the diagnosis and treat- ure: systematic review and meta-analysis. J Am Coll Cardiol. ment of acute and chronic heart failure: The Task Force 2006;47:1987---96. for the diagnosis and treatment of acute and chronic heart 30. Benedict CR, Weiner DH, Johnstone DE, Bourassa MG, Ghali failure of the European Society of Cardiology (ESC) Devel- JK, Nicklas J, et al. Comparative neurohormonal responses in oped with the special contribution of the Heart Failure patients with preserved and impaired left ventricular ejection Association (HFA) of the ESC. Eur J Heart Fail. 2016;18: fraction: results of the Studies of Left Ventricular Dysfunction 891---975. (SOLVD) Registry The SOLVD Investigators. J Am Coll Cardiol. 24. Zamora E, Díez-López C, Lupón J, de Antonio M, Domingo M, 1993;22 Suppl A:146A---53A. Santesmases J, et al. Weight loss in obese patients with heart 31. Ponikowski P, van Veldhuisen DJ, Comin-Colet J, Ertl G, Komajda failure. J Am Heart Assoc. 2016;5:e002468. M, Mareev V, et al. Beneficial effects of long-term intra- 25. Farré N, Aranyó J, Enjuanes C, Verdú-Rotellar JM, Ruiz S, venous iron therapy with ferric carboxymaltose in patients with Gonzalez-Robledo G, et al. Differences in neurohormonal activ- symptomatic heart failure and iron deficiency. Eur Heart J. ity partially explain the obesity paradox in patients with heart 2015;36:657---68.