Portal de Periódicos da CAPES

The BRONCH‐AHF study: effects on short‐term outcome of nebulized bronchodilators in emergency department patients diagnosed with acute heart failure

2017; Elsevier BV; Volume: 20; Issue: 4 Linguagem: Inglês

10.1002/ejhf.1028

1879-0844

Òscar Miró, Josep Tost, Víctor Gil, Francisco Javier Martín‐Sánchez, Pere Llorens, Pablo Herrero, Alexandre Mebazaa, Veli‐Pekka Harjola, José Ríos, Javier Marco‐Hernández, Sean P. Collins, W. Frank Peacock, Judd E. Hollander, María Teresa Lorca, Javier Jacob,

Asthma and respiratory diseases

In patients presenting to the emergency department (ED) with a chief complaint of dyspnoea of unknown origin, it is quite common for emergency physicians to provide intravenous diuretic plus nebulized bronchodilators while extensive rapid work-up is in progress. This serves to treat acute heart failure (AHF) and exacerbations of asthma or chronic obstructive pulmonary disease (COPD), which are the most common causes of dyspnoea seen in the ED. However, the safety and efficacy of acute administration of bronchodilators is poorly described in AHF patients without a specific (pulmonary) indication for bronchodilator therapy, as there are no studies specifically evaluating the effects of inhaled beta2-agonists (B2A) in patients with AHF.1 This lack of knowledge prompted the design of the present exploratory analysis of the EAHFE Registry, which includes consecutive patients diagnosed with AHF in 41 Spanish EDs during five different recruitment periods (2007/2009/2011/2014/2016).2 Our purpose was to ascertain the association of bronchodilator use (in addition to diuretics) with short-term outcomes in ED patients complaining of dyspnoea and ultimately diagnosed with AHF, but without a documented need for bronchodilators or diseases treated with bronchodilators such as COPD or asthma. The BRONCH-AHF (BRONCHodilators in Acute Heart Failure) study assessed patients recruited in the EAHFE Registry phase 4 (2014) and 5 (2016) who did not have a previous diagnosis of COPD or asthma, were not on chronic treatment with bronchodilators, and were not discharged (from ED or hospital) with bronchodilator treatment after the index episode. The patients were divided into four groups according to the treatment provided in the ED: 1) only intravenous diuretics (control group); 2) intravenous diuretics plus B2A nebulization (B2A group); 3) intravenous diuretics plus anticholinergic (Ach) nebulization (Ach group); and 4) intravenous diuretics plus B2A and Ach nebulization [dual nebulization (DN) group]. Forty-two independent variables that could potentially affect clinical outcomes were recorded to adjust outcomes (Supplementary material online, Table S1). The primary endpoint was 30-day all-cause mortality, and secondary endpoints were in-hospital all-cause mortality and length of stay (LOS) from ED admission to discharge. Comparison among groups was performed by one-way ANOVA (or the Kruskal–Wallis non-parametric test, if needed) and the chi-square test (with Yate's correction, if needed). For the primary endpoint, survival tables were obtained by the Kaplan–Meier method, and comparison was performed calculating hazard ratios (HR) with 95% confidence interval (CI) for 30-day death in the B2A, Ach and DN groups vs. the control group using the Cox regression model. For the secondary endpoints, we calculated odds ratio (OR) with 95% CI for in-hospital mortality, and absolute means difference (AMD) for LOS with 95% CI using logistic and linear regression models, respectively. Unadjusted HR, OR and AMD were adjusted for all independent variables with P < 0.10 in the univariate analysis. Missing value replacement was performed using a multiple imputation technique. Stratified analyses with interaction assessment for the primary endpoint were predefined for the following six variables: age (cut-off: 80 years), sex, history of ischaemic cardiomyopathy, chronic treatment with a beta-blocker, atrial fibrillation, tachycardia (cut-off: 100 b.p.m.), and left ventricular ejection fraction (cut-off: 35%). Statistical significance was accepted if 95% CI of HR or OR excluded the value 1, 95% CI of AMD excluded the value 0, or with P < 0.05. Of 7948 patients included in the EAHFE-4 (2014) and EAHFE-5 (2016) registries, 4207 fulfilled the inclusion criteria and 4146 had follow-up data and remained for the final analysis: 3403 (82.1%) controls, 129 (3.1%) B2A, 253 (6.1%) Ach, and 361 (8.7%) DN (Supplementary material online, Figure S1). Overall, the mean age was 80 ± 10 years, 59.3% were women, 72.7% presented multiple co-morbidities (at least 3/11 co-morbidities investigated), 59.2% had some functional limitation (Barthel Index <100) and 21.0% had a cardiorespiratory New York Heart Association class III–IV at baseline, and many did not undergo troponin (44.5%) and/or natriuretic peptide (51.4%) determination in the ED (Supplementary material online, Table S2). The primary endpoint was recorded in 173 (4.2%) patients, and with respect to the control group, the B2A, Ach and DN groups showed increased HRs for 30-day mortality: 1.855 (95% CI 0.944–3.644; P = 0.073), 1.230 (95% CI 0.681–2.222; P = 0.492), and 1.829 (95% CI 1.192–2.805; P = 0.006), respectively. However, all these HRs diminished after adjustment, and none achieved statistical significance (Figure 1). The stratified analysis of this adjusted model showed that interaction was only present for sex in patients receiving DN, with a significantly increased mortality for males (adjusted HR 2.147; 95% CI 1.143–4.031; P = 0.018) (Supplementary material online, Figure S2). Regarding secondary endpoints, 115 (2.8%) patients died in hospital and the median LOS was 5 days (interquartile range 1–9). We found a significant increase in in-hospital mortality for the B2A and DN groups, and longer hospitalizations for the DN group. However, after adjustment, only patients with DN maintained a significantly longer LOS (adjusted AMD +1.165 days; 95% CI 0.239–2.090; P = 0.014) (Table 1). We found that the use of nebulized bronchodilators, in addition to diuretics and other therapies for AHF, is present in 17.9% of patients, but this strategy is not associated with a significant positive or negative impact on short-term outcomes. However, some subtle findings also merit discussion. First, despite no statistical significance in the adjusted analysis, each study group had slightly different outcome profiles. While the direction of outcomes using B2A and DN (B2A plus Ach) was consistently towards an increase in adverse outcomes, the use of Ach alone was associated with a more neutral effect. Second, there was only one significant association observed between the use of DN and the increase in LOS; a finding supported by the suggestion that men could have increased 30-day mortality after DN treatment. Alternatively, as LOS is not a strong endpoint and may be influenced by many circumstances aside from treatment, this finding could be an artefact of multiple testing. Only the study of Singer et al.3 tested the association between bronchodilator use and outcomes in AHF patients without a history of COPD. They reported a neutral effect on in-hospital mortality (adjusted OR 1.02; 95% CI 0.67–1.56) but did not distinguish between bronchodilator classes. We found that Ach may be associated with more neutral effects than B2A or DN, which could indicate a better profile for use in patients with still undiagnosed dyspnoea in the ED. However, our data did not provide significant improvement in short-term outcomes, and thus, this still remains a hypothesis to be proved. On the other hand, DN is widely extended and a very recent Cochrane review of asthmatics demonstrated DN reduces hospitalization compared to patients treated with B2A alone.4 Conversely, this beneficial effect was not observed in our AHF patients, as the DN group had longer LOS. However, as commented above, prolonged hospitalization may be a poor surrogate for an adverse AHF outcome, as LOS varies greatly among hospitals, and could differ according to patient placement after ED diagnosis and management.5-7 Notwithstanding, it seems that when a DN is used, the outcome profile is closer to that observed for B2A than to that of Ach. Our study has some limitations. First, this is a secondary analysis of the EAHFE Registry limited to hypothesis generation. Second, exclusion of bronchodilator therapy need was based on assumptions from the retrospective review of a clinical datasheet and not on a thorough review of the complete medical reports. Third, emergency physicians were not asked about the reasons for bronchodilator use. This limits our capacity for determining whether bronchodilators where properly or unproperly administered, because while in some cases incomplete evaluation could have prompted incorrect use (many patients had a previous history of AHF but not COPD, and natriuretic peptide determination was lacking), in others, bronchodilators could have been indicated to treat symptoms of undiagnosed COPD or asthma in patients without a previous spirometry. Fourth, we did not collect doses, timing of administration, or the types of bronchodilators used in the ED, although in all the ED participating in the BRONCH-AHF study the first choice B2A and Ach are salbutamol and ipratropium bromide, respectively. Fifth, the use of bronchodilators by prehospital ambulances was not recorded. However, in Spain, only 15% of AHF patients come to the ED by advanced life support ambulances staffed with doctors allowed to provide these drugs.8 Finally, since this exploratory study had no sample size calculation, some groups were small in size (with few events) and there was an imbalance in groups size; consequently, we cannot exclude a type II error in some of our estimations, especially in the smaller subgroups (B2A group). In conclusion, the BRONCH-AHF study suggests that the addition of bronchodilators without a documented pulmonary need in patients ultimately diagnosed with AHF does not provide a benefit in survival compared to standard therapy (including diuretics), and the use of B2A or DN could have an increased risk for adverse outcomes. However, the use of bronchodilators by emergency physicians aimed at improving patients' signs and symptoms is still recommended, as these outcomes were not assessed in the present study. Further studies, especially a clinical trial designed to investigate the effect of this common therapeutic ED strategy, are warranted. Supported in part by grants from the Instituto de Salud Carlos III supported with funds from the Spanish Ministry of Health and FEDER (PI10/01918, PI11/01021, PI15/01019 and PI15/00773), Fundació La Marató de TV3 (2015/2510), and Catalonia Government for Consolidated Groups of Investigation (GRC 2009/1385 and 2014/0313). The ICA-SEMES Research Group has received unrestricted support from Orion Pharma and Novartis. The present study has been designed, performed, analysed and written exclusively by the authors independently from these pharmaceutical companies. Conflict of interest: none declared. Marta Fuentes, Cristina Gil (Hospital Universitario de Salamanca). María José Pérez-Durá, Eva Salvo (Hospital La Fe de Valencia). Rosa Escoda, Carolina Xipell, Carolina Sánchez, Josep M. Gaytan (Hospital Clínic de Barcelona). Antonio Noval (Hospital Insular de Las Palmas de Gran Canaria). José M. Torres (Hospital Reina Sofía de Córdoba). Maria Luisa López-Grima, Amparo Valero (Hospital Dr. Peset de Valencia). Alfons Aguirre, Maria Àngels Pedragosa (Hospital del Mar de Barcelona). Maria Isabel Alonso, Paco Ruiz (Hospital de Valme de Sevilla). José Miguel Franco (Hospital Miguel Servet de Zaragoza). Ana Belén Mecina (Hospital de Alcorcón). Susana Sánchez (Hospital Rio Ortega de Valladolid). Pascual Piñera (Hospital Reina Sofia de Murcia). Raquel Torres Garate (Hospital Severo Ochoa). Aitor Alquézar, Miguel Alberto Rizzi, Sergio Herrera (Hospital San Pau de Barcelona). Fernando Richard (Hospital de Burgos). Francisco Javier Lucas (Hospital General de Albacete). Irene Cabello, Álex Roset (Hospital Universitari de Bellvitge, Barcelona). José Manuel Garrido (Hospital Virgen de la Macarena, Sevilla). Héctor Alonso (Hospital Marqués de Valdecilla de Santander). Esther Rodríguez Adrada, Guillermo Llopis García (Hospital Clínico San Carlos, Madrid). Fernando Richard, José María Álvarez Pérez, María Pilar López Diez (Hospital Universitario de Burgos). Javier Lucas (Hospital General de Albacete). Joaquín Vázquez Álvarez, Marta Sánchez González, Belén Prieto, María García García (Hospital Universitario Central de Asturias). Víctor Marquina, Inmaculada Jiménez, Patricia Javaloyes, Néstor Hernández, Benjamin Brouzet, Ana López (Hospital General de Alicante). Juan Antonio Andueza (Hospital General Universitario Gregorio Marañón de Madrid). Rodolfo Romero (Hospital Getafe de Madrid). Roberto Calvache (Hospital de Henares de Madrid). María Teresa Lorca, Luis Calderón (Hospital del Tajo de Madrid). Beatriz Amores Arriaga, Beatriz Sierra (Hospital Clínico Lozano Blesa de Zaragoza). Enrique Martín Mojarro (Hospital Sant Pau i Santa Tecla de Tarragona). Lisette Travería Bécquer (Hospital Universitario de Canarias de Tenerife). Lluís Llauger García, Gerard Corominas La Salle (Hospital Universitari de Vic de Barcelona). Carmen Agüera Urbano (Hospital Costa del Sol de Marbella, Málaga). Ester Soy Ferrer (Hospital Josep Trueta de Girona). Table S1. Dictionary of the variables included in the present study. Table S2. Characteristics of patients included in the BRONCH-AHF study and comparison among the different groups of treatment. Figure S1. Patients' flow chart. Figure S2. Stratified analysis and evaluation of interaction for the primary endpoint for patients that were treated in the emergency department with bronchodilators compared to control group. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.