Blood gases, weaning, and extubation.

In this data-driven, evidence-based era of medicine, there remains a substantial portion of medical practice that lacks the ‘definitive’ clinical trial or other supporting science. This aspect of medicine is often referred to as the “art” of medicine and addresses that portion of disease management that is also referred to as clinical judgment or experience. There is probably an element of “art” in every medical decision. The timing of extubation or discontinuation of mechanical ventilation is no exception to this process. In recent years there has been more science used to base management. Guidelines developed by the major health care professional groups involved in the management of these patients provide a framework for weaning from mechanical ventilation and extubation.1 Recommendations include the control and elimination of factors that may contribute to ventilator dependence, early assessment after clinical stability for discontinuation of mechanical ventilation, as well as assessment for the potential discontinuation of mechanical ventilation during spontaneous breathing (as opposed to during supported breathing). The physiologic variables and indexes that have been studied to date have been only moderately successful in predicting the likelihood of successful extubation. With patients on mechanical ventilation the variables/indexes with statistically significant likelihood ratios (for predicting extubation success) are minute ventilation, negative inspiratory pressure, maximum inspiratory pressure, occlusion pressure at 0.1 second, and the CROP index (which integrates compliance, respiratory rate, oxygenation, and maximum inspiratory pressure). During spontaneous breathing trials, the variables/indexes with statistically significant likelihood ratios are respiratory rate, tidal volume, and rapid shallow breathing index. The (commonly used) respiratory rate and rapid shallow breathing index were found to have positive likelihood ratios of only 1.50 (95% confidence interval 1.23–1.83) and 1.58 (95% confidence interval 1.30–1.90) in pooled analyses.2 Once a patient has met the threshold of these weaning criteria, a spontaneous breathing trial (SBT) is conducted. There can be great differences in the conduct of SBTs (unsupported, supported, level of support, duration), but irrespective of the methods, patients who successfully tolerate the SBT are considered ready for extubation. The SBT serves several purposes, providing a real-time assessment of a patient’s respiratory mechanics, ability to oxygenate and ventilate, and comfort while breathing through a mechanical resistance that probably exceeds airway conditions when extubated. This strategy has been instrumental in reducing the duration of mechanical ventilation—by 2 days in the Ely et al study3—and has been the basis of other protocol-driven investigations,3–6 with improvement in the efficiency and efficacy of patient management. Reintubation rates have been 5–15%, in contrast with failure rates over 30% among those not undergoing SBTs.7 However, there are other causes of extubation failure that can not be identified with an SBT. Investigators have begun to focus on the possibility of post-extubation upper airway obstruction and the patient’s ability to cough and clear secretions as other factors that contribute to post-extubation respiratory failure.8–9 What other variables may contribute toward assessing these patients? What about arterial blood gas (ABG) values?