Of note all patients in this study who developed HCAP had a tracheostomy, whereas in the other studies the patients were intubated via the oral route.14, 15 and 16 Reflux of gastric contents into the oropharynx and subsequent aspiration into the lung may be a less important route by which pneumonia develops on patients with a tracheostomy and exogenous infection via the tracheostomy may be more important than endogenous infection from the oropharynx.19 Of note, patients in this setting had a surgical tracheostomy rather than the percutaneous (PERC) tracheostomies

more commonly used in ICUs in developed countries. The 30° angle may be insufficient to prevent the reflux of gastric contents into the oropharynx and subsequent aspiration into the lung. In the study of Drakulovic15 the patients were semi-recumbent at 45° whereas in the study of van Nieuwenhoven16 it proved impossible to maintain the planned angle BAY 73-4506 45°, an average treatment position of 28° was the result on day 1 and was down to 23° by day

7.15 and 16 In the current study we aimed for a 30° angle and this was checked twice daily. It was noted that patients tended to slip down the bed and that it was difficult to maintain the 30° elevation. A limitation of this study is that we did not formally document the adherence to the intended degree of elevation. It has also been suggested that maintaining a supine position in Omipalisib clinical trial the control group as in the study of Drakulovic15 led to a higher than normal rate of HCAP than is the case if a smaller 10° angle is maintained as in the study of van Nieuwenhoven.16 The rate of HCAP in this study was 38–39/1000 ventilated days. This rate is high compared with developed country settings but within the range reported in mechanically ventilated patients in developing countries.20, 21 and 22 It was lower than we had expected based on previous ward experience. In the period

Venetoclax in vitro leading up to the study several changes were made in the ward infrastructure and nursing care to improve infection control. This may have contributed to the lower pneumonia frequency during the course of the study. The study size as a result, lacked adequate power to show the 50% reduction in pneumonia frequency that was the target. However, at the time of this analysis, there was no suggestion of a lower pneumonia frequency in the semi-recumbent patients. The development of pneumonia was independently associated with an older age and a longer duration of mechanical ventilation consistent with other studies of pneumonia in patients receiving mechanical ventilation.14 We used a blind non-directed bronchial lavage method with quantitative cultures to determine the organism causing pneumonia.12 This method was appropriate for the local situation and gave a range of organisms consistent with studies of VAP from other similar locations.