INTRODUCAO: A assincronia paciente-ventilador e comum durante a ventilacao mecanica invasiva e tem sido associada com piores desfechos clinicos. Os objetivos deste estudo sao estimar a associacao entre a mecanica respiratoria e a incidencia de assincronia; avaliar o impacto do indice de assincronia e de "clusters" de duplo disparo sobre desfechos clinicos; e avaliar a acuracia do exame clinico a beira leito na deteccao de assincronia. METODOS: Este foi um estudo de coorte prospectivo em pacientes adultos sob ventilacao mecanica por mais de 24 horas. Foram coletadas caracteristicas basais dos pacientes e medida a mecanica respiratoria dentro das primeiras 72 horas pos intubacao. Nosso desfecho primario foi o indice de assincronia (IA), definido como a porcentagem de eventos de assincronia em relacao ao numero total de ciclos respiratorios e esforcos perdidos. As curvas de ventilacao mecanica foram capturas e analisadas durante todo o periodo de ventilacao por um software especifico (BetterCareR, Espanha), capaz de identificar automaticamente varios tipos de assincronia e calcular o IA. Calculamos o IA durante toda a ventilacao mecanica e durante a ventilacao assistida ( > 50% dos ciclos disparados pelo paciente). Consideramos alta incidencia de assincronia quando IA >= 10%. "Cluster" de duplo disparo foi definido como periodo de 3 minutos com pelo menos seis eventos de duplo disparo. Os pacientes tambem foram avaliados diariamente para deteccao de assincronia a beira leito e acompanhados ate a alta hospitalar. As analises estatisticas foram realizadas no programa R e incluiram modelos de regressao linear e logistica, para identificacao de preditores de assincronia e de mortalidade, e curvas de Kaplan- Meier e modelos de risco proporcional de Cox, para analises de sobrevivencia. RESULTADOS: Foram triados 125 pacientes em ventilacao mecanica, dos quais 103 foram incluidos no estudo. A mediana do IA foi de 5,1% (IQ: 2,6-7,8) durante toda a ventilacao mecanica e 5,4% (IQ: 2,9-9,1) durante a ventilacao assistida. Esforco perdido e duplo disparo foram os tipos de assincronia mais comuns e 22% dos pacientes tiveram IA >= 10%. "Simplified Acute Physiology Score 3" (SAPS3) e pressao positiva no final da expiracao intrinseca (PEEPi) foram associados com IA >= 10%, com OR=1,03 (IC95%: 1-1,06) e OR= 1.72 (IC95% 1.1-2.68), respectivamente. A presenca de IA >= 10% foi associada com falha de extubacao (33% nos pacientes com IA >= 10% e 6% naqueles com IA < 10%), mas nao com mortalidade. Noventa (87%) pacientes apresentaram "clusters" de duplo disparo e aqueles com mais de 12 horas cumulativas de "clusters" tiveram pior desfecho clinico, incluindo menor numero de dias livres de ventilacao mecanica (p < 0,01) e menor sobrevivencia (HR=2,09; IC95%:1,04-4,19; ajustado para tempo de ventilacao e SAPS3). Observamos ainda que os sinais clinicos de desconforto respiratorio avaliados isoladamente apresentaram acuracia moderada, entre 50% e 80%, para detectar assincronia. CONCLUSOES: As variaveis basais associadas com alta indecencia de assincronia sao SAPS3 e PEEPi; alta incidencia de assincronia esta associada com falha de extubacao; "clusters" de duplo disparo sao comuns e estao associados com piores desfechos clinicos

INTRODUCTION: Patient-ventilator asynchrony is common during invasive mechanical ventilation and have been associated with worse clinical outcomes. Our objectives are to estimate the association between respiratory mechanics and the incidence of asynchrony; to assess the impact of asynchrony index and clusters of double triggering on clinical outcomes; and to address the accuracy of clinical exam at the bedside to detect asynchrony. METHODS: This was a prospective cohort study in adult patients under invasive mechanical ventilation for more than 24 hours. Were collected patient's baseline characteristics and were measured respiratory mechanics within 72 hours after intubation. Our primary outcome was the Asynchrony Index (AI), defined as the percentage of asynchronous events divided by the number of respiratory cycles and ineffective efforts. Ventilator waveforms were captured and analyzed throughout the entire period of mechanical ventilation using a dedicated software (BetterCareR, Spain), capable of automatically identify a number of asynchronies and to compute the AI. We calculated the AI during the enterer period of mechanical ventilation and during assisted ventilation only ( > 50% of breaths triggered by patient). We defined high incidence of asynchrony as AI >= 10%. Cluster of double triggering was defined as a 3-minute period with at least sex events of double triggering. Patients were daily examined to detect asynchrony at the bedside and followed until hospital discharge. Statistical analyses were performed using R and included linear and logistic regression analyses, to address predictors of asynchrony and mortality, and Kaplan-Meier curves and Cox regression, to survival analysis. RESULTS: Were screened 125 patients and 103 were included in this study. The median AI was 5.1% (IQR: 2.6-7.8) during the entire period of mechanical ventilation and 5.4% (IQR: 2.9-91) during assisted ventilation. Ineffective effort and double triggering were the most common types of asynchrony and 22% of patients had AI >= 10%. Simplified Acute Physiology Score 3 (SAPS3) and Intrinsic Positive End Expiratory Pressure (PEEPi) were associated with AI >= 10%, OR=1.03 (95%CI: 1- 1.06) e OR= 1.72 (95%CI: 1.1-2.68), respectively. The IA >= 10% was associated with extubation failure (33% among patients with AI >= 10% and 6% among those with AI < 10%), but not with mortality. Ninety (87%) patients hat clusters of double triggering and those with >= 12 cumulative hours of clusters had worse clinical outcomes, including less ventilator free-days (p < 0.01) e shorter survival (HR=2.09; 95%CI:1.04-4.19; adjusted by duration of mechanical ventilation and SAPS3). We also found that clinical signs of respiratory failure had a moderate accuracy, between 50% and 80%, to detect asynchrony. CONCLUSIONS: Patient's baseline variables associated with high incidence of asynchrony are SAPS3 and PEEPi; high incidence of asynchrony is associated with extubation failure; and clusters of double triggering are common and are associated with worse clinical outcomes