Objective: There are varieties of clinical practices for intensive care respiratory support. However, there has been no published report characterizing its current practice in Thailand. The present study was undertaken to characterize the practice of respiratory support for intensive care patients in Thai tertiary hospitals.
Material and Method: A cross-sectional survey and retrospective historical cohort of intensive care units (ICUs) was performed on May 30, 2011 from ten tertiary hospitals in Thailand. The participating ICUs were asked to complete the following data of all patients who were mechanically ventilated in the ICUs: demographic data, characteristics of respiratory support, ICU type, causes of respiratory failure, and weaning technique.
Results: A total of 258 patients from ten tertiary hospitals were included and analyzed. The medical ICU patients remained in the ICU significantly longer than patients in other ICUs. Patients in surgical ICUs were significantly younger than patients in other ICUs. The prevalence of mechanically ventilated patients in this survey was 64.7% with a significantly higher proportion in the medical ICUs. The median of ventilator days was also significantly higher in the medical ICUs. An invasive ventilator was more commonly used in all ICUs rather than non-invasive ventilators. The three common causes of respiratory support were severe sepsis or septic shock, respiratory failure and post-operation, respectively. Volume-controlled continuous mandatory ventilation (VC-CMV) ventilation was more commonly used as the initial mode of ventilation in both surgical and medical ICUs. The maximum plateau pressure was significantly higher in the medical ICU patients but there were no differences in maximum tidal volume and PEEP level. One-third of the patients were in the weaning process, mostly in the medical ICUs. Pressure support was the predominant weaning mode in the medical ICUs, while synchronized intermittent mandatory ventilation (SIMV) was more predominant in the surgical ICUs. Protocol-based weaning was used in approximately two-thirds of patients who were in the weaning process. With repeated estimation equation logistic model and left censors’ cohort to 28 days, the medical ICUs had significantly lower ventilator free overtime individual patients when compared with surgical ICUs, while there was no difference within mixed ICUs.
Conclusion: The VC-CMV was more commonly used as the initial mode of ventilation in both surgical and medical ICUs. Pressure support was the predominant weaning mode in the medical ICUs, while SIMV was more predominant in the surgical ICUs. Individual patients in medical ICU had a greater number of ventilator days and less probability of being ventilatorfree.

Keywords: Ventilatory support, Thai ICUs, Weaning of ventilator, Mode of ventilation