Invasive mechanical ventilation (MV) is a fundamental tool for the management of patients with acute respiratory failure (ARF). The existing evidence shows that a number of interventions designed to prevent ventilator-associated lung injury and optimize weaning have a strong impact upon the duration of mechanical ventilation and mortality.1–5 Several epidemiological studies have investigated the different tendencies in the use of mechanical ventilation throughout the world.6–8 In addition, these studies have afforded useful information leading to updates in routine clinical practice referred to patients subjected to MV.

In Latin America, the epidemiological data on the implementation patterns referred to MV are limited. To our knowledge, no studies to date have provided epidemiological information on a general population of patients subjected to invasive mechanical ventilation in Argentina.9,10 To address this issue, we have described the mechanical ventilation practices in this country and explored the factors associated with mortality in the Intensive Care Unit (ICU) in patients of this kind.

A prospective, multicenter observational study was carried out between 1 September 2019 and 31 December 2019 in different ICUs in Argentina. The original study protocol was approved by the Ethics Committee of the Sociedad Argentina de Terapia Intensiva (ref. no. 1 2019) and was registered with clinicaltrials.gov (no. NCT04107467). Each participating center in turn received the corresponding approval from its own local Ethics Committee. The study was carried out following the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines for observational cohort studies.11

We included patients over 18 years of age admitted to any of the participating ICUs and who required invasive MV during at least 12 h from the time of admission to the participating institution; patients having started mechanical ventilation outside the ICU in a different Department (emergency or operating room); and patients admitted from a different institution and transferred to the ICU of one of the centers participating in the study. We excluded patients admitted to pediatric ICUs, anesthesia recovery or resuscitation wards, or coronary units. The patients were only enrolled in the study on the occasion of their first admission to the ICU. Patients with more than 10% of missing data referred to key study parameters used to generate the predictive model were excluded from the analysis.

This multicenter study provides new epidemiological data on mechanical ventilation in Argentina. The key findings can be summarized as follows: mortality in the ICU and in hospital among patients subjected to MV in this country is higher than reported at international level.6,8,10 Both the days of MV and the need for reintubation were high. With regard to the potential predictors, patient age, the diagnosis of sepsis, the use of NIV before MV, and extubation failure were correlated to increased mortality.

In coincidence with the observations of international studies, the most frequent reasons for starting MV were acute respiratory failure and coma.6,7,19

In their epidemiological study, Esteban et al.8 found VC-CMV to be the most widely used ventilation mode at the start of MV. Our own results are consistent with this; however, the percentage use of VC-CMV doubled that reported in the aforementioned study (EpVAr 2019: 74.9%; Esteban 2013: 38%).

Based on the results of our study, it can be concluded that the mortality rate among patients on MV for over 12 h is high in Argentina. Although the severity of the patients at the time of admission could explain this high percentage, other multicenter studies have revealed a mortality rate up to 15% lower in patient cohorts with similar characteristics in terms of variables such as age, SAPS II score upon admission and the reasons for MV.6,8,10 In a population of similar characteristics, Peñuelas et al. attributed mortality in the ICU to changes in clinical practice associated with the adoption of a protective ventilation strategy allowing the end-expiratory airway pressure to be maintained below the level considered to be harmful for the lungs.20 However, ventilator adjustment during the first day does not appear to be associated with mortality. In the same way as in other international reports, we found variables such as Vt and PEEP to be within ranges that could be regarded as acceptable in terms of lung protection.6,7,10 We also found that the values recorded for monitoring parameters such as driving pressure and plateau pressure were low.21–23 A high plateau pressure recording rate was observed (90%); this percentage was far higher than that recorded in the LUNG SAFE study (42%), and could reflect the interest in lung protection.23 This could be explained in terms of increased adherence to the use of protective mechanical ventilation protocols.24

We found four variables to be independently associated with increased mortality: age (over 70 years), the diagnosis of sepsis, the use of NIV before MV, and extubation failure. In coincidence with the study published by Estenssoro et al. in 2018,25 the need for mechanical ventilation in patients diagnosed with sepsis was an independent predictor of mortality. Unfortunately, we were unable to obtain information on the implementation of NIV as the first ventilatory support strategy, since the patients with successful initial NIV were not documented. Furthermore, in relation to extubation failure, the patients did not seem to have been prematurely extubated, since our median days on MV were similar to that reported by Esteban in 2013.8 However, the percentage of patients that were extubated in the first 24 h of the first SBT was lower than the figure reported in the WIND study (38.5% versus 57%).14 The T-tube SBT was used in a large proportion of patients (72.5%), despite the latest publications in relation to the process of weaning from MV, which point to this strategy as being more demanding than others for patients subjected to MV.26–29 It is known that sedation can have a negative impact upon ventilation-free days and on the days of ICU stay.30,31 According to the variables recorded by our investigators, many of the patients were subjected to deep sedation at some time during MV, even though almost half of the subjects had no organ failure. This may indicate an excessive use of sedation in our population – a fact that could have influenced the outcomes. Furthermore, another aspect that could indicate excessive use of sedatives is the fact that in our study the VC-CMV mode was the most frequently used ventilation mode, even on the tenth day of MV - while other studies report the use of spontaneous modes in preference to other modes before the end of the first week of MV.7–9

Although a confirmatory study specifically designed for the purpose would be needed, we believe that the high percentage of mortality recorded in our study could reflect the obstacles found throughout Argentina regarding the routine implementation of protocols referred to general patient care, such as those described in the PADIS guides.32 However, these results in relation to mortality could also be explained by differences in human and technological resources that were not considered in our study, or may reflect the findings of the LUNG SAFE and ICON studies, in which medium to low-income countries showed greater mortality due to sepsis and ARDS than high-income countries.33,34 Complex organizational and economic factors in medium to low-income countries account for the poorer outcomes in patients admitted to the ICU. Profound inequities, defined as systematic, unfair and avoidable differences in health determinants such as socioeconomic level, demographics and geography, can generate differences in access to healthcare services among different population groups – a situation which in turn has an impact upon health-related outcomes.35 Furthermore, in medium to low-income countries, the health systems tend to be divided into public and private sectors, and this sustains the differences based on socioeconomic level, affecting healthcare – particularly in the critical care setting.36 Of note in our study is the fact that we did not observe differences in mortality between these two healthcare sectors. This is possibly because we also did not observe structural differences in terms of human resources – with a similar distribution in the proportions physician/patient, nurse/patient and ventilatory assist specialist/patient in the two the mentioned sectors (unpublished data).

Our study has limitations. Firstly, this is an analysis of prospectively compiled clinical data corresponding to a large range of ICUs in our country, with different conditions regarding the patients and clinical practices – some of which could have influenced the results obtained. However, our analysis took this into account in the model to minimize possible bias related to the variables. Nevertheless, although use was made of multivariate models, the presence of unmeasured confounding factors cannot be discarded. Secondly, missing data are a problem in studies of this kind, since they may lead to misinterpretation of the results. To reduce this impact, we eliminated those records with over 10% missing data referred to the relevant study variables. Lastly, the data compiled in Argentina might not be representative of other medium to low income countries or other regions.

The present study describes the characteristics and clinical evolution of patients subjected to mechanical ventilation in Argentina. The main finding was a mortality rate higher than that reported in the international literature. Patient age, the diagnosis of sepsis, the use of noninvasive mechanical ventilation as an initial support measure, and extubation failure, were identified as independent mortality risk factors – some of them being potentially modifiable. The information obtained in this study is essential for our specialty and constitutes the basis for the development of care protocols seeking to optimize the management of patients on MV and thus improve the care outcomes in our country.

GAP has received funding for teaching programs from Medtronic Argentina and for teaching programs from Vapotherm Inc., USA.

EG is currently employed by Medtronic Argentina.

The mentioned sponsors had no role in the designing of the study, data collection and analysis, or in preparation of the manuscript.

GAP conceived and designed the study; collected, interpreted and analyzed the data; conducted the literature search; and drafted the manuscript. EG, MA and JHD designed the study; collected the data; and conducted a critical review of the manuscript. EN designed the study, collected and analyzed the data; and conducted a critical review of the manuscript.

All the authors approved the final version of the manuscript and assumed responsibility for all aspects of the study in order to guarantee that the issues referred to the accuracy and integrity of any part of the study were reviewed and adequately resolved.