Incidence and predictors of aspiration pneumonia in endoscopic sedation

Study aims: We investigated the incidence of aspiration pneumonie during outpatient endoscopic sedation.

Methods: 24.441 patients received endoscopic sedation with either Propofol mono or Propofol in combination with Midazolam during EGD or colonoscopy. We prospectively assessed signs for aspiration pneumonia in a questionnaire that 15.690 patients completed 24 hours after the endoscopic procedure. Using multivariate analysis we screened for predictors of aspiration pneumonia during endoscopic sedation.

Results: 829 of the 15.690 patients reported newly developed coughing the day after the examination (5.3%). In 130 out of the 829 patients we observed either concomitant coughing and shortness of breath (n = 107), coughing in combination with fever (n = 17) or fever with coexisting shortness of breath (n = 6). Therefore, 0.8% of the examined patients showed indicative symptoms of aspiration pneumonia 24 hours after the endoscopic procedure.

Aspiration pneumonia was significantly more frequent in patients that had received EGD compared to patients with colonoscopy (p < 0.001). An interaction between sedative and treatment showed (p < 0.04) that an aspiration pneumonia occurred less often for patients during colonoscopies who received propofol mono (14.6%) compared to propofol and midazolam (26.2%), while the effect of the sedative reversed for patients during gastroscopy (Propofol: 34.6%, Propofol and Midazolam: 12.3%).

Using multivariate analysis, we were able to identify coughing and vomiting during the endoscopic procedure to be of major predictive value. If coughing or vomiting occurred during the endoscopic procedure the relative risk of developing aspiration pneumonia the day after the examination was significantly increased by 25.65 (95% CI: 12.26 – 53.63).

Notably, we observed accumulating aspiration pneumonia during the winter months suggesting an environmental impact to the incidence reported (p < 0.0001).

Discussion: In the present study we were able to show that aspiration pneumonia has a comparable high incidence in endoscopic sedation. Furthermore, we were able to identify major predictors of aspiration pneumonia which based on the pronounced relative risk of 26.65 should lead to preventive antibiotic treatment if present.