Current issues of ACP Journal Club are published in Annals of Internal Medicine


Selective decontamination of the digestive tract in ICU patients

ACP J Club. 1992 July-Aug;117:13. doi:10.7326/ACPJC-1992-117-1-013

Source Citation

Gastinne H, Wolff M, Delatour F, Faurisson F, Chevret S. A controlled trial in intensive care units of selective decontamination of the digestive tract with nonabsorbable antibiotics. N Engl J Med. 1992 Feb 27;326:594-9.



To study the efficacy of selective decontamination of the digestive tract (SDD) in patients receiving mechanical ventilation (MV).


Randomized, double-blind, placebo-controlled trial.


15 intensive care units (ICUs) in referral hospitals.


Patients > 15 years of age requiring MV were eligible if they were intubated ≤ 48 hours before randomization. Patients were excluded if MV was discontinued < 24 hours after surgery, MV was started after sedative or alcohol overdose, or they had neutropenia, chronic degenerative central nervous system disease, spinal cord injury, enteropathy, were pregnant, or scored < 4 on the Glasgow Coma score, > 24 on the simplified acute physiologic score, were to receive only palliative treatment, were in another study, or refused.


4 times per day patients received either placebo (nonabsorbable calcium salt) or active intervention including both topical oropharyngeal application of 3 g of a gel containing 2% colistin, tobramycin, and amphotericin B and nasogastric tube administration of the same drugs (colistin and amphotericin, 100 g, and tobramycin, 80 mg).

Main outcome measures

Overall survival in the ICU, 60-day survival, duration of MV, and incidence of pneumonia.

Main results

75 of 220 (34%) patients treated by SDD and 67 of 225 (30%) patients in the placebo group died in the ICU. 60-day and total mortality were 37% and 40% in the SDD group and 34% and 36% in the placebo group. The differences between groups for mortality (both absolute and adjusted), duration of MV, and length of ICU stay were small and not statistically significant. Although the incidence of gram-negative pneumonia was lower in the SDD group, overall incidence of pneumonia did not differ (26 in SDD vs. 33 in placebo group; P = 0.42). The total mean cost for antibiotics was 2.2 times higher in the SDD group ($1287 vs. $577; P < 0.001).


Selective decontamination of the digestive tract as applied in this study does not have beneficial effects on mortality, duration of mechanical ventilation or intensive care unit stay, or incidence of pneumonia and was associated with a substantial cost increase.

Sources of funding: Lilly France and the Commission de la Recherche Clinique de l'Assistance Publique Hôpitaux de Paris.

Address for article reprint: Dr. H. Gastinne, Service de Réanimation, Hôpital Universitaire Dupuytren, 2 avenue Alexis-Carrel, 87042 Limoges CEDEX, France.


Nosocomial pneumonia is the second most common cause of hospital-acquired infection and may account for up to 60% of the infection-related mortality (1). The findings that oropharyngeal and tracheal colonization by gram-negative bacilli was associated with increased risk for pneumonia and that gut flora was determined to be the source stimulated studies of SDD in patients under intensive care. Decreased rates of gram-negative colonization and infection have led to the use of this approach in clinical practice to try to prevent morbidity and mortality. Unfortunately, many of these studies were not blinded or controlled, involved small numbers of trauma and postoperative patients, did not distinguish between early treatment of infection as opposed to prophylaxis, and did not examine effect on overall mortality. A recent meta-analysis of several trials concluded that, although the number of lower respiratory tract infections was decreased by SDD, its effect on mortality was not conclusive (2).

Gastinne and colleagues have done the largest well-designed, multicenter study to date of SDD in patients under intensive care. Although the control group had slightly fewer patients with multiorgan failure, the authors have conclusively shown that the routine use of SDD offers no survival advantage for patients in a predominantly medical ICU. The discrepancy with some earlier studies raises the question of comparability of surgical and medical patients in ICUs. Nonetheless, until improved outcome can be shown in a specific group of patients, the widespread adoption of this treatment cannot be supported.

Randolph J. Lipchik, MD
Medical College of Wisconsin Milwaukee, Wisconsin