We read with great interest the recent article by Xu et al. that discusses how the Swallowing, Tongue protrusion, Airway protection reflected by spontaneous and suctioning cough, and Glasgow coma scale Evaluation (STAGE) score of neurologic patients may predict the success of extubation. However, we have four points of concern.

First, it should be determined whether the STAGE scoring is superior to other previously studied scoring systems, such as the Visual pursuit, Swallowing, Age, and Glasgow for Extubation (VISAGE) score and Extubation trategies in Neuro-Intensive care unit patients and associations with Outcome (ENIO) study. The definition of outcomes should ideally be uniform to compare these systems. However, in the VISAGE scoring system, extubation failure is defined as reintubation within 48 h of extubation, and in the ENIO study, this period was extended to 5 days. Extubation criteria and definition of reintubation vary among intensive care units. Moreover, supportive devices postextubation may also differ. For research purposes, a standardized definition should be established.

Second, we question the inclusion of patients in this study. The authors included patients who were extubated. As mentioned, patients who underwent a tracheostomy due to a high possibility of difficulty in weaning from a ventilator were excluded from the study. We believe that this exclusion of patients was unjustified and clinically irrelevant. The results showed that 19% of patients were categorized as extubation failure and associated with a higher mortality. We agree that prolonged mechanical ventilation or the use of a tracheostomy is strongly associated with mortality. However, an extubation attempt should be counted as a separate event since the patients are not in the same condition before intubation. High-risk patients with low STAGE scores are advised to undergo a tracheostomy to improve patient safety. Thus, in utilizing the STAGE score as a decision-making tool in extubation, the differentiation of extubation success from the high possibility of tracheostomy is clinically relevant.

Third, items of the STAGE score were subjectively evaluated. Subjective dichotomic value of swallowing and cough intensity depends on the patient’s sedation level and the timing of scoring. Moreover, some items in the scoring system are correlated, such as patients showing poor suctioning cough having poor spontaneous cough or swallowing; thus, similar scores may be given. Moreover, the relationship between the STAGE scores given by the two respiratory therapists and the nonsystematic assessment by physicians or neurosurgeons remains unclear. A quantitative measurable value, such as the force expiratory flow, should be examined. Thus, STAGE scores, if correlated adequately to other clinical measures, may be more convincing.

Finally, traumatic brain injury phenotyping may aid in assessing the weaning process. In a previous study, traumatic brain injury phenotype was referred to as any traits or characteristics that distinguish a specific state, or any of its subtypes, and offers opportunity to improve patient outcomes. Traumatic brain injury phenotyping would be useful in determining the timing of extubation of patients with other cerebral disease.

Adequate knowledge of the STAGE scoring system can allow us to integrate this into machine learning, enabling us to easily predict extubation success.