4.2. Study limitations
There are several potential limitations to this study. First, a non-randomized retrospective trial it is subject to selection bias. Second, during the study period the CSICU was converted from an open care model to a semi-closed model, and which may have impacted outcomes as well. Third, in the initial period between tracheostomies were performed by open approach by ENT surgeons, and follow tracheostomies were performed at the bedside using the percutaneous dilatation technique. Fourth, we don’t investigate secondary outcomes as use of prolonged sedation, rate of VAP, length of hospital stay, and number of ventilator-free days. It was beyond scope of the study. Five, authors have no data on the number of patients who were discharged home/nursing home/rehabilitation facility or remained in hospital. Also, we do not have information on the number of decanulated patients. And sixth: we agree that it is difficult to assess the causality of timing of tracheostomy and long-term survival. It is possible that those who received a later tracheostomy were not initially indicated for a tracheostomy at all.