Over the last few decades numerous clinical trials have been conducted with patients admitted to Intensive Care Units (ICU) to prevent or reduce ventilator-associated lung injury—use of non-invasive ventilation1 (NIV), lung-protective ventilation,2,3 adjustment of positive end-expiratory pressure (PEEP),4 decubitus prone position,5 use of neuromuscular blockers6. Also, to reduce the duration of mechanical ventilation—proper use of sedatives7,8 and early identification of when to start weaning from mechanical ventilation.9 Some of these interventions, which initially for the management of patients with acute respiratory distress syndrome (ARDS) now seem to be applicable to all the patients on mechanical ventilation.10 The possible clinical impact of the results from these studies in the specific context of a country has not been described extensively.

The objectives of our analysis were these: (1) to describe the evolution of the clinical practice of mechanical ventilation in Mexico, (2) to estimate whether the changes seen have followed the current scientific evidence, and (3) to estimate whether mortality has changed over time.

This is a retrospective subanalysis study of patients admitted to Mexican intensive care units (ICU) who participated in 4 prospective, observational, non-interventional, international studies on mechanical ventilation conducted in 1998,11 2004,12 2010,13 and 2016. All of them were prospective and observational studies of patients admitted for one month who received invasive mechanical ventilation for over than 12h or NIV for over an hour. The follow-up of the patients included was during the time of mechanical ventilation up to a maximum of 28 days after inclusion.

The following variables were recorded in all patients: age, sex, severity on admittance estimated by SAPS II, daily arterial blood gas test, treatment-associated variables (ventilatory parameters, sedation, neuromuscular blockers), and complications (ARDS, sepsis, ventilator-associated pneumonia, cardiovascular failure, kidney failure, liver failure, hematologic failure), date, and condition upon ICU and hospital discharge.

The protocol and the definitions used have previously been published.13 Only the researcher from each ICU knew the characteristics of the study in order to not influence the daily routine clinical practices (list of researchers in Annex). The study was approved by the ethics committee of each of the participating institutions and the need for informed consent was applied according to local regulations. The study was conducted in full compliance with the Strengthening the Reporting of Observational Studies in Epidemiology standards for cohort observational studies.14

The objective of this analysis is to evaluate whether the changes observed from the first study conducted back in 1998 until the last study was conducted in 2016 meet the following hypotheses resulting from the clinical trials published over the last few decades: (a) increased use of NIV, both for the management of patients with chronic lung disease and for the management of acute hypoxemic respiratory failure; (b) increased use of a protective ventilation strategy defined as low or adjusted tidal volumes to maintain a low plateau pressure and use of high PEEP in patients with ARDS; (c) evaluate whether protective ventilation is also used in patients without ARDS criteria; (d) evaluate changes in the weaning from mechanical ventilation such as the frequency of use of support pressure vs. spontaneous breathing trial using a T tube, both during the first weaning attempt and in the weaning of patients with difficult-prolonged weaning, and (e) describe the practice of tracheotomy at the ICU setting over time. Finally, we evaluate whether some of these changes have had an impact on mortality rate.

The practice of mechanical ventilation in the Mexican ICUs that participated in 4 international studies on mechanical ventilation conducted between 1998 and 201611–13 has changed over time, according to the scientific evidence available. The biggest change operated here has been a reduced programmed tidal volume. Although there is a tendency toward a lower mortality rate, no significant differences were seen between the first to the last study.

A significant finding that could in part account for the absence of important changes in the mortality observed is the differences in the characteristics of the patients included in the 4 studies, above all in the reason why mechanical ventilation was started in the first place. It has gone from a prevalence of patients with postoperative respiratory failure and trauma patients seen in 1998 to the fact that, in 2016, 1 of every 3 patients included suffered from neurological disease. This finding, the population included in our study and the progression of the most common diseases, is consistent with other series published including patients from a specific country15–21 or from several different geographical regions.11,22,23

We should mention here the low incidence of patients with exacerbated chronic obstructive pulmonary disease (COPD). However, in this case the situation of the Mexican ICUs is similar to what has been reported worldwide. In the remaining countries that participated in the international studies on mechanical ventilation11–14 this type of disease has gone down from 10% to 7% over a period of 18 years. This finding seems to be explained by the low incidence of COPD in Mexico as confirmed by the PLATINO study.24 In this study conducted with a general population of ≥40 years, the percentage of patients with COPD-like clinical criteria in Mexico City was 7.8% (almost half of the percentage of patients found in 4 other South American capitals included in the study). The factors cited by the authors to explain these findings are the altitude in Mexico City (the majority of the ICU that participates in our study are located in this city) and ethnical aspects. However, these are hypotheses that require further analysis.

Research on mechanical ventilation over the last 2 decades has been changing the routine clinical practice regularly.12 Nonetheless, there is still great heterogeneity on this regard.25

One of the changes would be greater use of NIV as the first attempt of ventilatory support. Although to this day there are few studies available that have analyzed this evolution, a study conducted in France reported an increase from 16% in 1997 to 24% in 2011.1 In our study, that started from a low use of NIV, a reduction was seen. A possible explanation may be the case-mix of our population: a low percent of patients with COPD and heart failure in whom NIV has its main indication. Also, a high percentage of patients with neurological disease in whom NIV has not proven beneficial. On the other hand, most of the hospitals that participated in the study are public hospitals where the pressure sustained by the healthcare system is very high and the demand for ICU beds is much higher. This leads to a constant lack of ICU beds. This difficulty to accept patients in the ICU has triggered the development of NIV in the ER area and in some hospital wards like pneumology and internal medicine units.

Probably, the greatest change induced by clinical research is the so-called lung-protective ventilation strategy that ventilates with low tidal volumes (<6mL/kg of the ideal weight) and high PEEP with the objective of maintaining the plateau pressure <30cmH2O. This strategy, initially indicated for patients with ARDS2,3 has been used in the general population of patients on mechanical ventilation.10 In general, in the epidemiologic studies on mechanical ventilation published so far, we can see a progressive reduction of the programmed tidal volume. In the studies conducted during the first years of this century15,16,18,22 tidal volume was between 8 and 10mL/kg. Studies published more recently,19,26 have reported a progressive reduction to tidal volumes <8mL/kg, but >6mL/kg while keeping plateau pressure within safe limits. In our study, the evolution was similar in patients with ARDS criteria and without ARDS: the programmed tidal volume gradually dropped to a median of 6mL/kg in 2016.

The lowest compliance with the recommendations is showed by the programmed PEEP level, that remains at levels <10cmH2O in all the studies analyzed. In our series, we saw a significant change in patients with ARDS, in whom a median of 10cmH2O was reached in the last cohort included. These changes reduced the driving pressure to levels that have been associated with higher survival rates in patients with ARDS.27

Another aspect of mechanical ventilation that has changed significantly over the last 20 years is weaning.28 Changes can be summarized in the early identification of those patients who can be weaned from mechanical ventilation, running a test to check the tolerance to spontaneous ventilation and, in patients with difficult or prolonged weaning, withdrawals with a daily spontaneous breathing test29 or gradual reduction of support pressure.30 Although all these changes are well supported by scientific evidence, there are very many changes in the routine clinical practice.

A survey31 conducted among 1144 intensivists from 6 different geographical areas (Canada, India, the United, Europe, Australia/New Zealand and the United States) reported a daily identification rate of patients who meet the criteria for starting weaning of around 70% and 96%, support pressure as the preferred method for the first spontaneous breathing trial (between 57% and 72% as opposed to 9–59% that prefer the T-tube), and the weaning of patients with difficult/prolonged weaning (31% to 72% of the responders). In our study, we saw significant clinical changes, some of which match the results reported by the survey. Therefore, in time, it is possible to see a preference for the use of support pressure both for the first spontaneous breathing trial and for difficult weaning. An important piece of information here is the high use of SIMV (with or without support pressure) for weaning (26% of the patients in 2016 were weaned using this method) despite the fact that clinical trials show that this method is the least recommended of all.29,30 In the above-mentioned survey31 it was reported that SIMV is rarely used as a weaning method, although there are regions such as India and Australia/New Zealand where the percentage is closer to the figures confirmed by our study.

The mortality rate recorded in our series was lower than the one reported in similar epidemiologic studies,18,32 probably due to differences in the populations included. The lack of a significant reduction of mortality over the study period may be explained by the relatively low mortality rate seen over the 1998 period (15%). That is why the changes seen in the baseline characteristics of the population included during the study periods could counteract the possible benefit of the changes seen in the management of mechanical ventilation such as the protective ventilation strategy for example. As a matter of fact, since the 2004 period there has been a non-significant reduction of mortality that became more pronounced until 2016 that could be associated with the progressive increase of the protective ventilation strategy.

Our study has several limitations, which should be considered when interpreting the data available. For example, the different sample size included in each cohort can bring about imprecision in some comparisons and estimations. Also, the participating ICUs were not exactly the same in all the studies. This means that there could be a factor associated with variability in the clinical practice that was not possible to assess.

In conclusion, the clinical management of mechanical ventilation in Mexican ICUs has been modified over a period of 18 years. The most significant change is the adoption of a ventilation strategy based on low tidal volumes to maintain a safe plateau pressure. These changes have not been associated with significant changes in mortality, which should be associated with mechanical ventilation but, also, with factors that this study did not analyze.

MCM, JE, AVG, UC, MP, IP, JRS, and EM were the coordinators of the international study on mechanical ventilation in Mexico during the different periods and participated in the recruitment stage of the participants. They also collaborated with the writing of the manuscript. AM, OP, and FFV collaborated with the statistical design and analysis of the study and participated in the writing of the manuscript as well. MCM, FFV, OP, and AE collaborated as intellectual consultats of the study and participated in its design, data mining, database analysis, and in the writing of the final version of the manuscript.

None.

None declared.