Integrated Prediction of Extubation Outcome by the Spontaneous Breathing Trial in Newborn Infants

Prolonged endotracheal mechanical ventilation has significant adverse effects in newborn infants including subglottic stenosis, bacterial colonisation and bronchopulmonary dysplasia. Neonatal health professionals aim to limit the period of mechanical ventilation to the minimum possible duration. The decision to extubate is commonly a clinical one, based on current ventilator settings, blood gases analysis and the overall clinical condition of the infant. On the other hand, newborn infants commonly fail extubation and require reintubation which is a traumatic invasive procedure which could further prolong the total duration of invasive mechanical ventilation. Thus, accurate prediction of readiness for extubation outcome will probably incur a notable reduction in respiratory morbidity.

A number of studies have evaluated methods to predict extubation outcome. Of them, the spontaneous breathing trial (SBT) has been shown to hold the highest predictive value: it involves placing an infant from invasive mechanical pressure limited time cycled ventilation ventilation to endotracheal continuous positive end-expiratory pressure (ET-CPAP) for a period of 5 minutes during which saturation and heart rate are closely monitored. The infant would pass the test if there is no hypoxia or bradycardia during the test. A successful SBT has a positive predictive value of 93%, sensitivity of 97% and specificity of 73%. The combination of a successful SBT with increased variability in respiratory parameters during the SBT has been reported to further increase the sensitivity of the trial; alas the specificity remains below 75%.

It should be noted that at King's College Hospital the SBT forms part of routine clinical care as per clinical guideline and it is at the discretion of the attending physician to conduct the trial prior to extubation.

The work of breathing is undertaken by the respiratory muscles. Impaired respiratory muscle function is associated with a higher risk of respiratory muscle fatigue and respiratory failure and need for mechanical ventilation. Respiratory muscle function has been identified as an independent predictor of extubation outcome in mechanically ventilated infants. One simple way to quantify respiratory muscle function is the rate of relaxation of the inspiratory muscles which can be depicted by the rate of the decline of the airway pressure signal following a spontaneous breath. In concept, inefficient muscle function is manifested by a more gradual drop to baseline of the airway pressure over time. The principle on which the method is based is that when skeletal muscles contract against increased external loads, their relaxation slows as a result of slower calcium uptake by skeletal muscle fibres. In terms of physical properties, fatigued muscle fibres take longer to relax after contraction and this reduction in the relaxation rate can be quantified by an increase in the time constant of relaxation and a decrease in the maximal relaxation rate.

The rate of relaxation of the inspiratory muscles has been shown to accurately describe respiratory muscle function in non-neonatal populations. The information on the inspiratory muscle rate of relaxation is readily available in the ventilator graphics which, as part of routine clinical care, use a pressure port and a flow sensor to record airway flow, delivered volume and pressure over time. To the knowledge of the investigator, the potential value of this information in predicting extubation outcome in newborn infants has not been explored to date.

Typically, respiratory muscle dysfunction would occur in premature infants owing to immaturity, decreased muscle mass and prematurity-associated co-morbidities. Furthermore infants with Congenital Diaphragmatic Hernia (CDH) and Anterior Wall Defects (AWD) such as gastroschisis and exomphalos, have also been reported to have impaired respiratory muscle function which could probably preclude timely successful extubation. King's College Hospital provides care for a large number of infants with CDH and AWD while also has a long tradition in the assessment of respiratory muscle function in infants making it the ideal site to recruit infants with CDH and AWD.

The hypothesis of the investigator is that respiratory muscle function assessment using the rate of relaxation of the inspiratory muscles during SBT can accurately predict extubation outcome. This prediction might stand independently for the rate of relaxation or in conjunction with SBT-derived respiratory function parameters such as the outcome of the SBT and the variability of respiratory parameters during the SBT. An integrated algorithm incorporating respiratory muscle function data could thus be produced. This would be validated in a large cohort and could increase the prediction accuracy of extubation outcome and thus reduce re-intubation associated respiratory morbidity.

This will be a prospective, observational cohort study. Aims of the study are to measure the rate of relaxation, airway flow, airway pressure, tidal volume and transcutaneous oxygen saturation in all planned extubations over the coming 12 months in the neonatal unit at King's College Hospital. Data to be recorded include basic epidemiologic parameters such as gestation at birth, birth weight, corrected gestational age at measurement, day of life, weight at measurement, blood gas parameters within 6 hours of the SBT and associated co-morbidities that might impact on the success of the SBT such as open ductus arteriosus and current evidence of infection. Infants with known genetic or chromosomal abnormalities, infants with congenital anomalies other than CDH or AWD and infants with focal acute lung pathology such as atelectasis or pneumothorax will be excluded from the study.

Based on the unit's activity over the past two chronological years, the investigators anticipate that with a successful consent rate of 80% it will be possible to measure 100 infants in total: 40 premature infants, 40 term infants and 20 infants with CDH and AWD. As this is a novel index, direct power calculation cannot be undertaken before the commencement of the study. An interim power calculation based on pilot data will be conducted half way through the project at six months, to determine the required sample size. The investigators believe however, that given that previous studies with similar outcome measures incorporated significantly smaller numbers of participants, the achieved sample size will be sufficient for the objectives of the study to be achieved.

Inform written consent will be requested from the parents or legal guardians of the infants and the attending Neonatal Consultant will be requested to assent to the study.

When the clinical team decides an infant is ready for extubation, the investigator will switch the ventilator to ET-CPAP at the same pressure as the positive end expiratory pressure during mechanical ventilation. Airflow, integrated tidal volume and airway pressure will be recorded with the NM3 respiratory function monitor (Philips Respironics, Connecticut, USA). The monitor will be connected to a Laptop (Dell Latitude, Dell, Bracknell, UK) with customised Spectra software (3.0.1.4) (Grove Medical, London, UK). The NM3 respiratory profile monitor has a combined pressure, flow and carbon dioxide sensor and this will be placed in line between the inspiratory limb of the ventilator circuit and the endotracheal tube. The NM3 monitor is automatically calibrated for flow and pressure. Flow will be measured using a fixed orifice pneumotachograph. One of the tubes from the pneumotachograph will be connected to a second pressure transducer to measure airway pressure. Oxygen saturation and heart rate monitoring will also be performed.

A failed SBT will be recorded if the infant has either a bradycardia for more than 15 seconds and/or a fall in SpO2 below 85% despite a 15% increase in FIO2. At this point the study will be stopped and ventilation will be restarted.

Infants will be extubated to either nasal CPAP or high flow nasal cannula oxygen at the clinician's discretion.

The primary study outcome will be reintubation within 72 hours of extubation. The indications for reintubation will be: (a) more than six episodes of apnoea requiring stimulation in six hours, or more than one significant episode of apnoea requiring bag and mask ventilation; (b) respiratory acidosis: PaCO2 >8.5 kPa and pH,7.25; (c)FIO2>0.60 to maintain SpO2 in the range of 90-95%.

If the infant has either bradycardia for more than 15 seconds and/or a fall in SpO2 below 85% despite a 15% increase in FIO2, the study will be stopped and ventilation will be restarted. Strict infection control policies will be followed according to the unit's guidelines.

Data will be stored until the child is 25 years old. Continuous outcomes will be compared between premature and term infants and between term infants and infants with CDH/AWD. Student's t test will be used when the data are normally distributed and Mann-Whitney U test when skewed. Categorical data will be assessed using the Fisher two tailed exact test. P<0.05 will be considered significant. The ability of respiratory variables to accurately discriminate between successful and failed extubation will be assessed using receiver operating characteristic curves. A value of the area under the curve of 0.5 will indicate no discriminatory value, and a value >0.8 will suggest a high predictive value. Standard formulae will be used to calculate sensitivity, specificity, and positive and negative predictive values and likelihood ratios. Statistical analyses will be performed using SPSS software (SPSS Inc, Chicago, Illinois, USA).

The investigators anticipate that by incorporating information on respiratory muscle function during the SBT, it will be possible to create an algorithm that will accurately predict extubation outcome in mechanically ventilated infants. The investigators also want to demonstrate that the rate of respiratory muscle relaxation might be different in premature infants, in full-term infants and in infants with surgically corrected anomalies such as CDH and AWD. These results will be incorporated in local clinical practice and form a guideline on "when to safely extubate" infants on mechanical ventilation. This will reduce respiratory morbidity associated both with prolonged ventilation as well as with failed extubation and re-intubation.