Introduction: Surgical therapy of carcinoma of the esophagus or cardia by transthoracic esophageal resection is associated with a high morbidity in which nosocomial infections have a great importance. This study investigates the influence of prophylactic selective bowel decontamination on the course and results of transthoracic resection of the esophagus.
Methods: Seventy patients with carcinoma of the esophagus and cardia were included in this prospective and partially randomized study at the University of Heidelberg. Twenty-five patients received prophylactically selective bowel decontamination with tobramycin, polymyxin B and amphotericin B. The treatment course was documented uniformly. In addition, microbiological screening was performed by swab examinations of nose, throat and anus, by urine and blood cultures, and the documentation of results of additional microbiological diagnostic studies.
Results: Bacteriological screening confirmed a reduction in infectious agents and a change of their spectrum in the respiratory and digestive tract without an increase in multiresistant bacteria. Patients who received selective bowel decontamination had a lower infection rate, a shorter artificial respiration period and a shorter intensive care stay without statistically significant differences. The mortality rate was 4% vs 9% in the control group (95% confidence interval -0.172-0.116).
Conclusion: This study confirms the feasibility and microbiological effectiveness of selective bowel decontamination in the context of surgical therapy which is associated with a high nosocomial infection rate. The result of the clinical treatment seems slightly more favorable in the treatment group. Decisive are complications caused by surgery which fundamentally determine the clinical course and frequently cause infectious complications. The prophylactic use of selective bowel decontamination may be useful in patients with an increased risk of prolonged ventilation support or colon interposition but it is not to be generally recommended.