Both the delayed withdrawal and the premature interruption of mechanical ventilation have been associated with increased mortality in intensive care units (ICUs).1,2 The decision of whether a patient can tolerate extubation is important as there is failed extubation in 10%–20% of patients, and this is associated with poorer clinical results.3,4 During this process, it is not uncommon to find errors in the prediction of extubation when using regular evaluation methods to measure the ability of the patient to breathe without help, such as blood gas analysis, vital capacity (VC) determination, and minute ventilation (VE). However, airway resistance (Raw) and elasticity (static) of the respiratory system are simple, non-invasive methods that are widely used to predict the success of extubation in patients with respiratory disease.5–7

Patients with cerebral infarction, especially those who present with cerebral edema and an extensive affectation of the encephalic trunk, often present a deteriorated level of consciousness. These patients generally have a poor prognosis due to the affectation of vital centers or respiratory complications.8–10 In ICUs, the question frequently comes up of how to extubate stroke patients with satisfactory parameters when there are concerns about the patient's level of consciousness and capacity to maintain spontaneous breathing. It is well known that there is a high incidence of pneumonia in patients with cranial trauma in whom prolonged intubation becomes necessary.11

The rapid shallow breathing index (RSBI), which is reflected in the relationship of respiratory frequency and tidal volume (f/Vt), is a simple datum to obtain for predicting the result of the extubation more precisely than isolated parameters.12 Chatila et al.13 observed that the f/Vt ratio was more exact than other clinical tools that are frequently used to predict the result of mechanical ventilation extubation. The predictive exactness and the simplicity of the f/Vt parameter justify their use in the management of patients with mechanical ventilation. The maximal transdiaphragmatic pressure ratio (Pdi/Pdimax) could be another important parameter for predicting the success of extubation.14–16 However, little is known about the integrated evaluations of respiratory mechanics, maximal transdiaphragmatic pressure and the PaO2/FiO2 ratio.

Although there are few studies that establish indices for predicting the success of extubation in stroke patients and there are no available guidelines, according to our experience in this disorder, the evolution seems to be good without an artificial airway. Our hypothesis was that Pdi and Pdi/Pdimax are important indirect indicators for predicting the satisfactory extubation of the mechanical respirator in stroke patients. The objective of this study is to identify components of respiratory mechanics that are associated with the success of extubation in stroke patients in whom mechanical ventilation is used.

We carried out an observational, prospective cohort study in 20 consecutive patients with stroke who were hospitalized with mechanical ventilation for more than 10 days in our ICU. The protocol included the periods between July and October 2008. The study was approved by the internal review committee at our university, and the family members of all the patients gave their written informed consent.

The continuous variables are presented as means, standard deviation, ratios, and percentages. The simple comparisons of the groups were done with the t-test for unpaired data or with the χ2 test. The factors associated with failed extubation were studied with a logistic regression model and the strength of the association is expressed with odds ratios and 95% confidence intervals. The variables evaluated were the following: Ti/Ttot, Pdi/Pdimax ratio, PaO2/FiO2 ratio, PImax, RSBI and respiratory mechanics. The level of statistical significance for each test was established at P<.05. The sample size was calculated depending on the variability of the maximal transdiaphragmatic pressure with the use of the E/S equation, corresponding with the magnitude of the expected effect divided by standard deviation of the result assessment variable. We used an α error of 5% and a β error of 20%, based on the results of previous studies.12 18 patients were necessary in order to meet the statistical power demands established for this research. The statistical analysis was done with a commercial statistical computer program (InStat3®).

The 20 patients included in this study had a mean age of 56.4±4.7 (48–67); mean BMI was 20.1±3.4(14.3–27.1)kg/m2, PImax was 43.8±9.1(30–64)cmH2O, transdiaphragmatic pressure (Pdi) was 15.1±5.5 (9–26), maximal transdiaphragmatic pressure (Pdimax) was 45±10(30–64)cmH2O, the transdiaphragmatic pressure/maximal transdiaphragmatic pressure ratio (Pdi/Pdimax) was 0.34±0.1 (0.17–0.65), PaO2/FiO2 was 199.8±35 (122–281) and RSBI – Tobin was 107±12 (104–111). The effective inspiratory time (Ti/Ttot) was 0.38±0.1, the mechanical ventilation time was 302±35, lung elasticity was 43.4±8.8 and Raw was 7.9±2.4 (Table 1).

In the group in which the extubation process was satisfactory, the mean number of days of mechanical ventilation was 12.5±2.2 versus 13.1±2 in the group with failed extubation. In 12 patients (60%), tracheostomy was not used (extubation success). The characteristics of the 2 groups of patients are presented in Table 2.

We considered the extubation to be successful if the patient remained extubated 48h later. The first 4h following the extubation were the worst for these patients, given that half of the group that failed in the process needed invasive mechanical ventilation support within this time period. The other patients were re-intubated over the course of the first 24h (P=.0001).

The muscle fatigue index was determined with the Ti/Ttot and Pdi/Pdimax ratios. It was considered within the area of respiratory muscle fatigue if any of these values were higher than 0.4. The mean Ti/Ttot and Pdi/Pdimax for the extubation failure group were 0.4±0.08 (95% CI, 0.36–0.44) and 0.5±0.7 (95% CI, 0.43–0.56), respectively. In the extubation failure group, there were only 2 patients with a Pdi/Pdimax ratio lower than 0.4, and 2 presented a Ti/Ttot ratio equal to 0.4. Thus, 6 patients were in the fatigue zone and 2 in the risk zone for developing respiratory muscle fatigue. The Raw was significantly higher in this group (P=.0004). Static lung elasticity did not present significant differences between the two groups (P=.34) (Table 2).

None of the patients in the satisfactory extubation group were within the respiratory muscle fatigue zone: 6 were in the zone at risk of developing fatigue and 3 were in the lack of fatigue zone. The Ti/Ttot ratio was 0.37±0.05 (95% CI, 0.34–0.41) and the Pdi/Pdimax ratio was 0.25±0.05 (95% CI, 0.21–0.28). The groups with failed extubation presented values of the Ti/Ttot ratio and the Pdi/Pdimax ratio higher than those of the satisfactory extubation group (P<.0001 and P=.0008, respectively) (Table 2).

We observed a significant correlation between the Pdi/Pdimax ratio and the RSBI (r=0.55; P=.009) and the PaO2/FiO2 (r=−0.59; P=.005) (Figs. 1 and 2, respectively).

Fig. 1.

Correlation between the transdiaphragmatic pressure/maximal transdiaphragmatic pressure ratio (Pdi/Pdimax) and the rapid shallow breathing index (RSBI – Tobin).

(0.07MB).

Fig. 2.

Correlation between the transdiaphragmatic pressure/maximal transdiaphragmatic pressure ratio (Pdi/Pdimax) and the arterial oxygen pressure/inspiratory fraction of oxygen (PaO2/FiO2).

(0.06MB).

All the significant variables are included in a logistic regression model with the aim to determine the factors that predisposed patients for reintubation. The patients in whom there was an RSBI<80, a PaO2/FiO2 ratio<200, a Pdi/Pdimax ratio<40cmH2O, a Pdi>15 and a Raw<8 presented mechanical ventilation extubation failure (Table 3).

Our study has obtained two main results:

• •

  The values of the muscle fatigue index (Ti/Ttot/Pdi/Pdimax and the Pdi/Pdimax ratio) higher than 40% play a decisive role as predictive factors of a failed extubation process in stroke patients.

• •

  The first 4h after extubation is decisive in terms of the success or failure of the extubation process.

Some data of respiratory mechanics and oxygenation are frequently evaluated in the extubation process of patients in whom prolonged mechanical ventilation has been used.2,13 Factors such as age, sex (mainly women), heart dysfunction, and low cardiac output could also require mechanical ventilation in excess.14,15 Despite there being few studies20,21 about the evaluation of this extubation process with the use of data on respiratory mechanics, none of them have been centered on the evaluation of respiratory muscle strength as a predictive factor for the success of extubation in these patients in order to withdraw mechanical ventilation. However, in our study we have observed the importance of measuring respiratory mechanics as well as the strength of the respiratory muscles (Table 3). It could be supposed that our patients did not have enough muscle strength because their BMIs were at the lower limit of normal (Table 1).7 Nevertheless, this situation is predictable in ICU patients who have been kept on mechanical ventilation for long periods of time.16

Several studies have recommended the use of static lung elasticity as a reliable variable for the prediction of successful extubation in patients with lung disease (e.g. COPD, pneumonia, ARDS).2,16 Furthermore, these data are easy to obtain with non-invasive methods. Nonetheless, as other studies have indicated, we observed that static lung elasticity presents lower-than-normal values,22 and this datum was not sensitive enough to differentiate the prognosis in the extubation process (P=.77). One probable explanation of this observation would be related with the time spent by patients in the ICU and the effect of the positive pressure during mechanical ventilation.

We have also observed that Raw was greater in the failed extubation group than in the satisfactory extubation group (P=.003). The high Raw in this group of patients could be related with edema, the presence of liquids or secretions in the airway walls that reduce functional residual capacity, and bronchial spasms.23 Oddo et al.24 demonstrated that high Raw is one of the most important factors that predict the need for mechanical ventilation, as well as the difficulty of satisfactory extubation, and that these alterations increase the effort required for respiration and could lead to respiratory muscle fatigue and a high risk of mortality. Failed extubation is frequently the result of a lack of protection of the airway and, therefore, it depends on the combination of cough, secretions and mental state.

We have demonstrated that the increased muscle fatigue index Ti/Ttot/Pdi/Pdimax (3.1 [2.13–5.5]) and the Pdi/Pdimax ratio were the main driving factors for extubation failure (OR, 4.7 [2.5–5.8]). This would occur mainly in the presence of a high Raw. Not even with the use of non-invasive ventilation was muscle fatigue avoided in the 8 patients in whom mechanical ventilation was re-instated.

PImax continues to be widely used for the prediction of extubation failure and the early detection of muscle fatigue, especially when this value is less than 20cmH2O. However, Meade et al.25 revealed that PImax is extremely inexact and should not be used for predicting the results of respirator disconnection or extubation. Nonetheless, our results indicate that the muscle fatigue index and the Pdi/Pdimax ratio, as an isolated measurement of diaphragm effectiveness, have a predictive value higher than that of PImax, especially in patients with values higher than 20cmH2O.

It has been reported that respiratory skeletal muscle fatigue occurs when the muscles generate more than 15% of their maximal effort during a sustained contraction. It is well-known that fatigue is produced only when the pressure–time level generated surpasses that threshold level. In addition, the fatigue threshold is higher during the intermittent contractions and depends on the relative duration of the contraction and relaxation, especially if the inspiratory muscles of the individuals examined are subjected to an external inspiratory load. For this reason, we measured the respiratory muscle fatigue index, which is defined as Ti/Ttot/Pdi/Pdimax, where Ti is the inspiratory time, Ttot is the total respiration time, Pdi is the mean transdiaphragmatic pressure generated by breathing and Pdimax is the maximal transdiaphragmatic pressure. The value of Ti/Ttot/Pdi/Pdimax reflects the function of generating contractions by the intercostal muscles and diaphragm. When breathing is produced with an inspiratory load, the threshold of the muscle fatigue index is from 0.35 to 0.40. Under this threshold, breathing can be maintained for more than 1h without evidence of fatigue. Above this threshold, there is a failure in the task after a time limit that is inversely related with muscle fatigue index.17,18

Our data demonstrated that in 60% of our patients from the failed extubation group, mechanical ventilation was reinstated in the first 4h after extubation. Yang et al.12 demonstrated that the first 4h following extubation are crucial for determining either the success or failure of said process. This is related with the difficult adaptation of the patient to spontaneous breathing, the variability of the levels of consciousness and the high resistance of the airways. Several authors12,17,21 have indicated that the work of respiration is the sum of the resistive and elastic work, and that the higher values during a long period of time leads to fatigue of the respiratory musculature. The authors also suggest that the muscle fatigue index (Ti/Ttot/Pdi/Pdimax) test the energy output of the respiratory muscles during mechanical work, constituting a predictive muscle fatigue variable.25 As we have shown, the high values of Pdi and the Pdi/Pdimax ratio were the reason why some patients presented failed extubation during the first 4h before spontaneous breathing (Table 2).

Our study has certain limitations. We did not determine the intracranial pressure, although we constantly verified the maximal respiratory pressures of our patients in order to avoid all undesirable vascular hypertension.26 Another limitation is that we did not have a control group available, based on the supposition that our patients were their own controls.

The muscle fatigue index is an important parameter for predicting the success of the extubation process in patients with stroke who have used prolonged mechanical ventilation. The success of extubation in stroke patients is related with a careful evaluation of respiratory musculature indices, such as the Pdi and Pdi/Pdimax.

The protocol was presented to the ethics committee of our university and approved by said committee, and all patients gave their written informed consent.

No fund was received, for this study, from any tobacco company or from other companies.

This material has not previously been published, either completely or partially, nor is it being considered for publication elsewhere. Any unexpected adverse effects or change in protocol have been declared. All the authors have made important contributions to the study and have read and approved of the manuscript. All the authors agree with the declarations indicated and cede the editorial rights to this article.

The authors declare having no conflict of interest.