Learning From Experience

Improving Early Tracheal Extubation Success After Congenital Cardiac Surgery

Peter D. Winch, MD, MBA; Anna M. Staudt, MD; Roby Sebastian, MD; Marco Corridore, MD; Dmitry Tumin, PhD; Janet Simsic, MD; Mark Galantowicz, MD; Aymen Naguib, MD; Joseph D. Tobias, MD


Pediatr Crit Care Med. 2016;17(7):630-637. 

In This Article


Our data again demonstrate that early tracheal extubation is achievable in small infants undergoing cardiovascular surgery. Successful early tracheal extubation avoids the potential complications and associated morbidity of prolonged tracheal intubation in the CTICU and thus results in a shorter length of hospitalization. Over the last decade, there have been tremendous changes in the overall intraoperative management of pediatric cardiac surgical patients. Not only have we seen improved surgical techniques, but also better CPB management such as reduced circuit sizes for pediatric patients and low-flow bypass techniques to avoid circulatory arrest. With regards to the anesthetic management, the evolution of inhaled volatile anesthetic agents along with shorter-acting analgesics and sedatives has also changed management strategies. This has also led to a smoother and more stable emergence and recovery in these patients. All of these factors have considerably contributed to the practice of successful early tracheal extubation.

The early tracheal extubation strategy returns the patient to a physiology of negative pressure ventilation sooner, a physiology particularly important in those patients with Blalock-Taussig shunts and Glenn or Fontan anatomy. It also minimizes the need for pharmacologic sedation and its attendant impact on cardiovascular physiology. While there are clear advantages of the early tracheal extubation, both physiologic and in terms of parental satisfaction, it is of paramount importance to be able to identify those patients who are and are not candidates for early tracheal extubation and continuously improve their management. Early tracheal extubation failure that either directly causes morbidity or prolongs hospital stay does far more harm than benefit. In a large study looking at extubation failure in a multi-institutional cohort of CTICU patients, the duration of ventilation prior to extubation was an important independent risk factor predictive of extubation failure, with the rate of failure increasing two-fold after only 24 hours of ventilation.[7] In this analysis and in our study, extubation failure was correlated with increased in-hospital mortality and prolonged LOS, confirming these as risks to consider when planning the extubation strategy. Thus, it is imperative that we monitor our results with this strategy and continue to attempt to identify challenges that arise from this approach.

To this point, it is important to consider how our practice has progressed since our last report in 2009. In 2009, we detailed predictors of successful early tracheal extubation following congenital cardiac surgery in neonates and infants. We concluded, "in order to obtain a probability of extubation of at least 50%, 0–3 month-old patients need a weight at or above 3.6 kg and a CPB time of less than 80 minutes".[5] However, in our recent experience, the success rate for patients with both of these risk factors was 89%, with similarly high success rates observed among non-CPB patients weighing less than 3.6 kg. These updated data demonstrate considerable improvement in early extubation success rates over the predicted probability of 50% computed from a past cohort.[5] Since 2009, we have developed a small cohesive group of cardiac anesthesia practitioners, which has enabled us to pursue a practice guideline that uses a standardized anesthetic. Such practice guidelines have been demonstrated to improve outcomes throughout all areas of medical practice.

Preoperative pulmonary hypertension is perceived to be one of the major contraindications to early tracheal extubation.[8–11] Many institutions often keep these patients sedated and chemically paralyzed due to the fear of a pulmonary hypertensive crisis which remains a major cause of morbidity and mortality.[12] On the other hand, positive intrapulmonary pressure and alveolar over-distention caused by prolonged mechanical ventilation might increase pulmonary vascular resistance.[13] Suctioning through the endotracheal tube has been shown to be a potent trigger for pulmonary hypertensive crisis in patients with preoperative pulmonary hypertension (PAH).[13,14] PAH in itself does not seem to contraindicate early tracheal extubation, which remains a feasible way to decrease postoperative ICU stay and hospital costs.[15] Early tracheal extubation also maintains stable hemodynamic parameters in the immediate postoperative period. Changes that may occur with alterations in arterial blood values are often transient without a major impact on the overall outcome.[16,17]

As demonstrated in Table 3, there was no meaningful difference between the tracheal extubation failure rate for the early (group 2) (11%) or late (group 4) patients (15%). Arterial blood gas analyses from the OR through the CTICU stay until the time of reintubation were examined for potential predictors of deterioration in cardiac output or respiratory drive, such as lactic acidosis, hypercapnia, or anemia. Unfortunately, there was nothing in the laboratory data that predicted the need for reintubation in these 10 patients. Of the 22 patients who required reintubation within 24 hours of arrival in the CTICU, 45% failed because of excessive agitation. This agitation resulted in escalation or change to the standard nurse controlled analgesia (NCA) used for nearly all postoperative patients (< 1 yr old): fentanyl 0.5 μg/kg basal rate with 0.5 μg/kg bolus every 10 minutes as needed. In these 10 patients, there was an increase in pain medication administration, additional sedative agents, or a variation from the normal NCA order that appeared to contribute to respiratory depression and failure, requiring reintubation. Delirium may have also been a contributing factor, as the escalation of the sedative and analgesic agents did not result in control of the symptoms, but rather a further escalation.

Our success in early tracheal extubation has resulted in the emergence of a new challenge related to postoperative pain management and sedation for this unique population of patients, particularly during the first 24 hours. Clearly, not all agitation in the CTICU is caused by pain. Hunger, the unfamiliar environment of the CTICU with its attendant noises and lighting, the inability or permissibility of parents holding their infant, and the uncertainty and anxiety transferred by parents to their children are all possible contributors. Additionally, delirium in the ICU setting in the adult population is well documented and is felt, in part, to arise from a disruption of the circadian rhythm by routine ICU operating procedures. Infants may be particularly susceptible to this phenomenon.[18–20]

We loosely use the pediatric anesthesia emergence delirium scale as detailed by Sikich et al[21] at bedside in the CTICU when attempting to decide which patients will benefit from the escalation of pain medications. This scale awards points (from 1 to 4) based on the applicability of a series of statements; if the child makes eye contact with the caregiver, if the child's actions are purposeful, if the child is aware of their surroundings, and if the child is restless or inconsolable.

As a result of our analysis, we have embarked on an effort in our CTICU to provide non-pharmacologic interventions to decrease patient agitation. Appropriately selected patients may be placed in adult size beds to allow for parental presence (holding and comforting). In our experience, it has been possible to reduce agitation and avoid further escalation of opioid administration through this approach. There is a wide range of evidence to support the use of nonpharmacologic methods to provide complementary pain relief in children which may include cognitive-behavioral methods.[22,23] It is also important to assess the needs of patients on an individual basis. This may be based on various factors such as age, type of surgery, or preoperative status, with the intervention planned accordingly.

As our youngest patients cannot tell us why they are agitated, we, as creative, educated providers must seek better, perhaps more imaginative and holistic ways to treat the agitated infant. The "Kangaroo Care" model has been embraced in the neonatal ICU setting as a benefit in the care of the smallest patients. Perhaps a similar benefit may be realized for the population of patients included in our study. Realizing that not all patients are the same, nor are all parents appropriate candidates for this approach, careful selection will be needed in order to best utilize this method. Whether it is reducing the noise and light level in the wards at night or placing infants in larger beds to allow parental access, we must remain open to such changes. Pain and agitation often coexist and hence our management plan should be multimodal with the safe combination of medications and nonpharmacologic means such as cognitive-behavioral and emotional support and strategies to create a comfortable environment.