Reflex cough PEF as a predictor of successful extubation in neurological patients*

Abstract

Objective:

To evaluate the use of reflex cough PEF as a predictor of successful extubation in neurological patients who were candidates for weaning from mechanical ventilation.

Methods:

This was a cross-sectional study of 135 patients receiving mechanical ventilation for more than 24 h in the ICU of Cristo Redentor Hospital, in the city of Porto Alegre, Brazil. Reflex cough PEF, the rapid shallow breathing index, MIP, and MEP were measured, as were ventilatory, hemodynamic, and clinical parameters.

Results:

The mean age of the patients was 47.8 ± 17 years. The extubation failure rate was 33.3%. A reflex cough PEF of < 80 L/min showed a relative risk of 3.6 (95% CI: 2.0-6.7), and the final Glasgow Coma Scale score showed a relative risk of 0.64 (95% CI: 0.51-0.83). For every 1-point increase in a Glasgow Coma Scale score of 8, there was a 36% reduction in the risk of extubation failure.

Conclusions:

Reflex cough PEF and the Glasgow Coma Scale score are independent predictors of extubation failure in neurological patients admitted to the ICU.

Introduction

Mechanical ventilation (MV) is an important tool in the treatment of respiratory failure in patients with acute disease or acute exacerbation of chronic disease; however, patients should be weaned from MV as soon as the condition leading to their being placed on MV is resolved. ¹Screening tests can be used in order to make decisions regarding weaning and extubation, which have been the subject of many studies in recent decades. ¹

Weaning failure is often attributed to impaired gas exchange, respiratory muscle fatigue, and an imbalance between respiratory load and ventilatory demand; however, in patients with neurological injury, impaired airway protection can be one of the reasons for weaning failure.^(2,3)

The parameter that is most commonly used in order to evaluate the ability of patients to breathe spontaneously and weaning potential is the ratio between RR and tidal volume, which is known as the rapid shallow breathing index (RSBI). ⁴Although the RSBI is widely used, it has not proved to be a reliable predictor in patients with neurological injury.^(5-10) In such patients, extubation failure appears to be more closely related to their inability to protect the airway than to their ability to breathe spontaneously. ¹¹

In 20% of patients admitted to the ICU, acute neurological disorders (neuromuscular disease, in 10%, and coma or central nervous system dysfunction, in 10%) are the main reasons for initiating invasive MV. ¹²Therefore, there is a need for parameters that are more accurate in order to make weaning decisions in such patients.

Cough effectiveness or strength before extubation, as measured by voluntary cough PEF with an open glottis, has been studied as a predictor of the ability to protect the airway after extubation in critically ill patients, being an important measure of airway protection ability. ^(13-18) However, there are physiological differences in motor activation of expiratory and accessory muscles between voluntary and reflex cough; the latter is associated with widespread and simultaneous activation of expiratory and accessory muscles, generating two or more PEFs of lower amplitude when compared with the former. ¹⁹

The objective of the present study was to evaluate the use of reflex cough PEF as a predictor of successful extubation in neurological patients who were candidates for weaning from MV.

Methods

This was a cross-sectional study conducted in the ICU of Cristo Redentor Hospital, in the city of Porto Alegre, Brazil, between January of 2011 and June of 2013. The research project was approved by the Research Ethics Committee of the Conceição Hospital Group (Protocol no. 10-150).

The inclusion criteria were as follows: being on MV for more than 24 h; having been placed on MV because of a neurological condition; and being a candidate for weaning from MV.

The criteria for extubation were as follows: adequate oxygenation, with an FiO₂ of < 0.4; hemodynamic stability (HR of < 130 bpm); mean arterial pressure > 60 mmHg with minimal or no use of vasopressors; axillary temperature of < 37.5°C; hemoglobin level > 8 g/dL; Glasgow Coma Scale score ≥ 8; and acid-base and electrolyte balance.

A digital manometer (MVD-500 v.1.1; Globalmed, Porto Alegre, Brazil) was used in order to measure MIP and MEP, the best of three consecutive maximal inspiratory and expiratory maneuvers performed with the one-way valve closed for 30 s being recorded. The RSBI was measured immediately before a spontaneous breathing trial (SBT), by means of a spirometer connected to the endotracheal tube. The Glasgow Coma Scale score and reflex cough PEF were determined by the Cristo Redentor Hospital ICU physical therapy team immediately before the SBT, with the head of the bed elevated 45°.

A portable peak flow meter (Mini-Wright AFS; Clement Clarke International Limited, Harlow, England) was used in order to measure reflex cough PEF. The peak flow meter was connected to the endotracheal tube via a T-piece with a one-way valve to allow free inhalation and exhalation through it. To initiate the cough reflex, mechanical stimulation was provided by introducing an 8F suction catheter through the nose; when the stimulation delivered above the glottis was insufficient to elicit cough, the catheter was inserted into the endotracheal tube through the front opening of the T-piece in order to stimulate tracheal receptors.

The SBT was performed with a T-tube and supplemental oxygen (maximum FiO₂ of 40%) for 30 min. Extubation failure was defined as the need for reintubation within less than 48 h after extubation. All clinical and demographic data were collected from the medical records of the patients. The need for reintubation was determined by the attending physician, and the information regarding the reasons for reintubation was collected from the medical records of the patients.

The Statistical Package for the Social Sciences, version 18.0 (SPSS Inc., Chicago, IL, USA) was used for statistical analysis. The Kolmogorov-Smirnov test was used in order to determine whether the data were normally distributed. Continuous variables were expressed as means and standard deviations or medians and interquartile ranges, whereas categorical variables were expressed as percentages. The Student's t-test or the Mann-Whitney test (for continuous variables) and Pearson's chi-square test or Fisher's exact test (for categorical variables) were used. The relative risk was calculated in order to determine the associations among predictor variables. Variables with a value of p < 0.2 were included in the multivariate analysis in order to compare extubation success and failure rates. To evaluate the predictive ability of reflex cough PEF, Glasgow Coma Scale scores, and the RSBI, sensitivity, specificity, and accuracy were measured with a ROC curve. The level of statistical significance was set at p < 0.05.

Results

A total of 135 neurological patients extubated between January of 2011 and June of 2013 were included in the present study. The mean age of the patients was 47.8 ± 17 years. Male patients predominated (71.1%), and the extubation failure rate was 33.3%.

The most common clinical diagnosis was that of traumatic brain injury, in 62 patients (47%), followed by those of subarachnoid hemorrhage, in 35.6%, intracranial hemorrhage, in 11.4%, and brain tumor (in the postoperative period), in 6.1%. Other clinical characteristics of the sample are presented in Table 1. Of the patients who required reintubation, 7% did so because of a decreased level of consciousness, 31% did so because of accumulation of bronchial secretions, and 62% did so because they were unable to protect their airway.

Table 1. Clinical and epidemiological characteristics of the 135 patients studied, together with their ventilatory parameters.^a .

Variable	Total	Extubation success	Extubation failure	p
	(n = 135)	(n = 90)	(n = 45)
Age, years	47.80 ± 17.01	48.17 ± 17.50	49.82 ± 16.93	0.875b
Male gender, n (%)	96 (71.10)	66 (73.30)	30 (66.70)	0.561c
APACHE II score	18.87 ± 5.41	18.20 ± 5.70	20.40 ± 4.40	0.024b
GCS at admission	7.77 ± 2.14	7.94 ± 2.12	7.20 ± 2.16	0.79b
GCS score at extubation	9.69 ± 1.12	10.07 ± 0.93	8.90 ± 0.51	< 0.001b
PEEP, cmH2O	5.27 ± 0.46	5.25 ± 0.45	5.31 ± 0.47	0.516
FiO2, %	34 ± 0.49	34 ± 0.41	34 ± 0.63	0.921b
VT, mL	522 ± 134	535 ± 135.40	493 ± 121	0.180b
Duration of MV, daysd	8.62 ± 5.70	7.21 ± 4.85	11.46 ± 6.26	< 0.001d
MIP, cmH2Od	65.22 ± 23.81	70.43 ± 22.30	54.80 ± 23.53	< 0.001d
MEP, cmH2Od	69.10 ± 43.87	75.65 ± 48.80	55.73 ± 27.59	0.003d
RSBI, breaths/min/Ld	46.34 ± 17.79	43.86 ± 16.76	51.30 ± 18.92	0.028d
Reflex cough PEF	102.09 ± 41.13	115.34 ± 38.95	75.76 ± 31.23	< 0.001b

APACHE II: Acute Physiology and Chronic Health Evaluation II; GCS: Glasgow Coma Scale; PEEP: positive end-expiratory pressure; V_T: tidal volume; MV: mechanical ventilation; and RSBI: rapid shallow breathing index. ^aValues expressed as mean ± SD. ^bStudent's t-test. ^cPearson's chi-square test. ^dMann-Whitney test.

With regard to the ventilatory, hemodynamic, and gas exchange parameters related to the SBT, there were no statistically significant differences between the patients in whom extubation was successful and those in whom extubation failed ( Table 2). However, there were statistically significant differences between the two groups of patients regarding the length of ICU stay, the length of hospital stay, and mortality ( Table 3).

Table 2. Ventilatory, hemodynamic, and gas exchange parameters measured within 30 min after initiation of a spontaneous breathing trial.^a .

Variable	Extubation success	Extubation failure	pb
	(n = 90)	(n = 45)
SBP, mmHg	138 ± 22	140 ± 16	0.411
DBP, mmHg	80 ± 11	81 ± 8.90	0.677
HR, bpm	89 ± 12	88 ± 12	0.838
RR, breaths/min	23 ± 8.90	25 ± 3.80	0.359
Arterial blood gases, pH	7.40 ± 0.31	7.40 ± 0.03	0.436
PaCO2, mmHg	40 ± 6.20 (5.37 ± 0.83)	39 ± 5.47 (5.16 ± 0.73)	0.177
PaO2, mmHg	117 ± 35 (15.60 ± 4.67)	125 ± 31 (16.67 ± 4.13)	0.243
SaO2, %	98 ± 1.51	98 ± 1.75	0.230
PaO2/FiO2	346 ± 116	356 ± 112	0.507

SBP: systolic blood pressure; and DBP: diastolic blood pressure. ^aValues expressed as mean ± SD. ^bStudent's t-test.

Table 3. Outcomes in the patients in whom extubation was successful and in those in whom extubation failed.^a .

Variable	Extubation success	Extubation failure	p
	(n = 90)	(n = 45)
Length of ICU stayb	12 (7-17)	17 (14-23)	< 0.001c
Length of hospital stayb	25 (17-30)	30 (21-52)	0.009c
Outcome of hospitalization	0.017d
Discharge	77 (85.6)	29 (64.4)
Death	4 (4.4)	9 (20)e
Transfer	9 (10)	7 (15)
Outcome of ICU care	< 0.001d
Discharge	84 (93.3)	38 (84.4)
Death	1 (1.1)	6 (13.6)e
Transfer	5 (5.6)	1 (2.2)
VAP	32 (35.5)	24 (53.3)	0.027

VAP: ventilator-associated pneumonia. ^aValues expressed as n (%), except where otherwise indicated. ^bValues expressed as median (interquartile range). ^cMann-Whitney U test. ^dPearson's chi-square test. ^eStatistically significant association with adjusted standardized residuals (level of significance of 5%).

A reflex cough PEF of < 80 L/min and the Glasgow Coma Scale score proved to be independent predictors of extubation failure. A reflex cough PEF of < 80 L/min showed a relative risk of 3.6 (95% CI: 2.0-6.7; p < 0.001), and the final Glasgow Coma Scale score showed a relative risk of 0.64 (95% CI: 0.51-0.83; p < 0.001). For every 1-point increase in a Glasgow Coma Scale score of 8, there was a 36% reduction in the risk of extubation failure. Similarly, the ROC curve showed that reflex cough PEF and the Glasgow Coma Scale score are good predictors of successful extubation ( Figure 1). It is of note that the accuracy of the RSBI in predicting successful extubation was low in our sample of patients.

Figure 1. Accuracy of reflex cough PEF and Glasgow Coma Scale scores in predicting successful extubation in neurological patients admitted to the ICU.

Discussion

The present study showed that reflex cough PEF can be a predictor of successful extubation in neurological patients who are candidates for weaning from MV.

It is known that 20% of all individuals admitted to the ICU require ventilatory support because of acute neurological complications. ¹⁰Although there have been reports of successful extubation in such patients, it is important to assess their mental state and their ability to protect the airway. ²⁰Studies have shown that patients with traumatic brain injury and subarachnoid hemorrhage have a 20-45% chance to develop ventilator-associated pneumonia, MV being directly related to prolonged ICU stay and mortality.^(14,15) In our study, ventilator-associated pneumonia was more common in the patients whose extubation failed than in those whose extubation was successful.

In addition to increasing hospitalization costs significantly, extubation failure is associated with prolonged MV, an increased incidence of ventilator-associated pneumonia, and an increased risk of mortality.^{(16-18,21-24)} In the present study, pneumonia was significantly more common, mortality rates were significantly higher, and ICU and hospital stays were significantly longer in the patients in whom extubation failed, findings that are consistent with the literature.

The influence of age on weaning outcomes can vary depending on the population studied. ²⁴In the present study, there was no significant association between patient age and MV weaning outcomes. This finding is consistent with those of other studies of patients with neurological injury, suggesting that age is not an important predictor of MV weaning outcomes in such patients.^(9-11)

Our study population consisted predominantly of males (71.1%). Epidemiologically, this is due to the fact that the ICU of Cristo Redentor Hospital is a trauma ICU, where the prevalence of male patients is associated with lifestyle or greater exposure to risk factors for traumatic brain injury. ²⁵

In the present study, there were no significant differences between the patients in whom extubation was successful and those in whom extubation failed in terms of the ventilatory, hemodynamic, and gas exchange parameters during the SBT. This finding is consistent with those of studies of neurological patients, in whom the major obstacles to weaning and extubation are associated with their inability to protect the airway rather than with their inability to breathe spontaneously, given that there are no differences between the groups.^(9,10)

In the present study, there were significant differences between the patients in whom extubation was successful and those in whom extubation failed regarding the RSBI, MIP, and MEP evaluated separately. However, in the multivariate analysis, there were no significant differences between the two groups of patients regarding those parameters. Although they are among the most widely assessed parameters for weaning and extubation, they are not good predictors of extubation failure in neurological patients.^(6-10) The accuracy of the RSBI in predicting successful extubation was found to be low in the present study. This finding is consistent with those of previous studies, showing that the RSBI is not a good predictor in such patients.^(9,10)

Consciousness level assessment by the Glasgow Coma Scale is extremely important in neurological patients. A study by Namen et al. showed that the Glasgow Coma Scale score was associated with successful extubation regardless of the protocol used; a score ≥ 8 was associated with successful extubation in 75% of cases, whereas a score of < 8 was associated with successful extubation in 33% of cases.^(26,27)

In the present study, all patients had a Glasgow Coma Scale score > 8, the mean score being 10.07 ± 0.93 points in the group of patients in whom extubation was successful and 8.9 ± 0.51 points in the group of patients in whom extubation failed. Our logistic regression analysis showed that a Glasgow Coma Scale score of 8 is an independent predictor of extubation failure, every 1-point increase in that score reducing the risk of extubation failure by 36%. The present study showed that the Glasgow Coma Scale score has good accuracy in predicting successful weaning in neurological patients. These findings are similar to those of Namen et al., who demonstrated that the chances of success increased by 30% for every 1-point increase in the Glasgow Coma Scale score. ²⁶Vidotto et al. studied 92 neurosurgical patients who had Glasgow Coma Scale scores ≥ 8 and who were extubated after an SBT and noted that 16% required reintubation. ⁴Mokhelesi et al. found similar rates, with an extubation failure rate of 50% for scores ≤ 10 and of 9% for scores > 10. ²⁸

Airway protection can be evaluated by measuring peak cough flow (PCF) with a peak flow meter or a pneumotachograph. Analysis of PCF as measured with a peak flow meter in 95 patients before extubation showed that PCF was significantly lower in those in whom extubation failed than in those in whom extubation was successful (64.2 ± 6.8 L/min vs. 81.9 ± 2.7 L/min); the risk of extubation failure was 5.1 times higher in those in whom PCF was lower than 60 L/min. ¹⁵However, analysis of PCF as measured with a pneumotachograph showed that PCF was similar between the patients in whom extubation was successful and those in whom extubation failed, mean PCF values being 79.7 L/min and 58.1 L/min, respectively. ⁷The differences between the two aforementioned studies regarding PCF values can be explained by the different measurement instruments used. In a study of 150 patients admitted to a medical ICU, ²⁹reflex cough PEF was measured with a portable respiratory mechanics monitor. Mean reflex cough PEF was 74 L/min in the patients whose extubation was successful and 42 L/min in those whose extubation failed. ²⁹The authors of that study concluded that PCF is a potential predictor of successful extubation in patients who pass an SBT. ²⁹

The present study has some limitations. Mean reflex cough PEF values (115.34 ± 38.95 L/min in the patients in whom extubation was successful and 75.76 ± 31.23 L/min in those in whom extubation failed) were higher than those found in the literature. This difference can be partly explained by the use of a portable peak flow meter and by the fact that the study population consisted solely of patients with neurological injury, unlike the clinical populations evaluated in other studies. It is of note that reference values for reflex cough PEF have yet to be established, and that studies aimed at doing so can be useful in determining the best timing for weaning and extubation. ³⁰

Although several studies have suggested that it is essential to evaluate cough and, consequently, airway protection ability before extubation,^(7,15,23) there is still no consensus on the methodology to be used or on cut-off points to predict extubation outcomes.

We can conclude that reflex cough PEF as measured with a peak flow meter and the Glasgow Coma Scale score are independent predictors of extubation failure in neurological patients admitted to the ICU.