Possible errors in diagnosis of bacterial sinusitis in tracheal intubated patients

Bacterial sinusitis is recognized as a complication of prolonged nasotracheal intubation. Verification of the diagnosis, however, requires culture findings from a "true" sample from the maxillary antrum. A "true" antral sample can only be obtained after passing either the nasal or oral mucosae with their rich aerobic and anaerobic bacterial flora. No properly controlled study has been made into the problem of contaminated samples in bacterial sinusitis complicating prolonged nasotracheal intubation. The design of the study was a case series in an intensive care unit. The patients had either had endotracheal tube or tracheal cannula for seven days or were septic, having an endothracheal tube or a tracheal cannula and therefore due to an investigation of infection foci. Antral cultures obtained via the canine fossa were compared with oral cultures to exclude the contribution of oral bacterial contamination and thereby to improve the reliability of diagnosis based on culture results. The contamination rate was lowered significantly from 67% to 5% when an area of free bone was prepared in the canine fossa thus avoiding instrument contact with the oral mucosa. A high contamination risk can be dealt with by adjusting the technique of sampling thereby increasing the culture reliability.