Selective decontamination of the digestive tract: to stimulate or stifle?
- 68 Downloads
Since its introduction as a prophylactic strategy for intensive care unit (ICU) patients 20 years ago, selective decontamination of the digestive tract (SDD) has driven scientific opinions towards extremes. Although intense scrutiny often stimulates discovery, at times polarization can stifle progress. In Intensive Care Medicine the contribution by van Saene and coauthors  reviews the evidence that in their view supports the widespread use of SDD in ICU patients. Does the review stimulate or stifle? "C'est le ton qui fait la musique" and, in our opinion, the tone of the review is overly aggressive in chiding colleagues, editors, and conference organizers for "being antagonistic towards SDD in a fundamentalist way." According to the authors the polarized views of SDD have split the ICU world into "traditionalists" who reject SDD and "heretics" who propagate it. Because we consider that the arguments proposed by van Saene and colleagues in favor of SDD represent a selective use of the available scientific evidence, we present our analysis of the issues, focusing on the effects of SDD on patient mortality rates and on antibiotic resistance rates. We hope the readers will agree that we should not be considered as "traditionalists with a nihilistic attitude towards the concept of SDD" since each of us has conducted controlled trials of SDD [2, 3, 4, 5].
Effects of SDD on ICU-acquired infections
SDD has been shown by many studies to be a successful preventive measure for ICU-acquired infections, especially respiratory tract infections. Importantly, however, the studies that do not support a benefit were carried out in ICUs with high levels of antibiotic resistance [5, 6, 7, 8, 9, 10]. Although the evident impact of SDD on reducing incidence rates of ventilator-associated pneumonia (VAP) appears to be inversely related to the methodological quality of the SDD study, the benefits remain significant in "high-quality" studies .
An important question, however, is whether the total package of SDD (topical prophylaxis in oropharynx and intestines and systemic preventive therapy) is necessary for this benefit. The relevant importance of the different aspects of the SDD regimen for the prevention of VAP has not been elucidated. It should be recalled that the incorporation in the ICU regimen of systemic antibiotics in the so-called "SDD" was an addition to the original concept of SDD, as initially applied to neutropenic patients [12, 13, 14].
Four studies have evaluated the effects of oropharyngeal decontamination, in one study in combination with systemic prophylaxis , and all found significant reductions in VAP [2, 15, 16, 17]. The magnitude of these preventive effects are similar to those reported in successful SDD trials. In one of these studies a 63% relative risk reduction in the incidence of VAP was achieved without disturbing gastric and enteral colonization . The results of this and other studies [18, 19, 20] seriously question the importance of the gastropulmonary route of infection in the pathogenesis of VAP. However, a head-to-head comparison of the complete SDD package vs. oropharyngeal decontamination in a randomized fashion has never been performed. And a nagging concern in all of these trials has been the potential for bias in the interpretation of respiratory tract secretion cultures, especially tracheal aspirates, in light of potential spillover of the prophylactic oropharyngeal antibiotics , which may hamper microbiological diagnosis of VAP.
Effects of SDD on patient survival
For 20 years researchers have been unable to demonstrate a significant reduction in patients' mortality rates in individual studies. Survival benefit was demonstrated only in meta-analyses [11, 22, 23]. D'Amico et al.  reported a mortality odds ratio of 0.8 (95% confidence interval 0.69–0.93) in a meta-analysis of 16 trials that evaluated SDD regimens that included systemic prophylaxis. The odds ratio was 0.98 (0.73–1.33) for 7 studies that compared topical and systemic prophylaxis with systemic prophylaxis and 1.01 (0.84–1.22) for 11 studies that evaluated topical prophylaxis alone. These results imply that systemic antibiotics are the most important part of the SDD regimen, presumably reducing mortality by preemptive therapy of early-onset infections in patients at risk (e.g., surgical and trauma patients, or patients with impaired airway reflexes). Unfortunately, this issue is unresolved because there are too few studies with adequate power to compare systemic vs. systemic plus topical antibiotics.
Nathens and Marshall  compared the effects of SDD on survival in surgical and medical populations. In surgical patients they found a mortality odds ratio of 0.6 (0.41–0.88) when analyzing the effects of topical and systemic prophylaxis vs. controls. There was no significant reduction when only topical prophylaxis was used. In medical patients mortality reductions were not apparent for the combination of systemic and topical or for topical prophylaxis alone. Van Nieuwenhoven et al.  analyzed the relationship between methodological study quality and outcome effects of SDD. In contrast to the inverse relationship between study quality and pneumonia prevention, no such effect was demonstrated for ICU mortality. Overall, SDD was associated with a relative risk reduction for mortality of 0.12 (0.03–0.21), and this reduction was 0.14 (0.03–0.25) for the studies with the highest methodological study quality. The results of these meta-analyses suggest that SDD can improve patient outcome with an estimated relative reduction in mortality of approx. 0.1–0.2 for a mixed population. Whether the results of meta-analyses should be used as a proof of principle or more as a hypothesis generator indicating the potential benefit is a matter of debate. Potential methodological pitfalls of meta-analyses in general, and of SDD trials in particular, have been reviewed .
More recently the results of two trials support the contention that SDD improves patient survival. After stratification on Acute Physiology and Chronic Health Evaluation II (APACHE II) scores, Krueger and coworkers  randomized 265 patients in a double-blind design to a regimen containing intravenous ciprofloxacin for 4 days and topical colistin and gentamicin applied to nostrils, mouth, and stomach; control patients received intravenous and topical placebo. The overall relative risk for ICU mortality was 0.76 (0.53–1.09), but in the subgroup of moderately ill patients (APACHE II scores of 20–29) the relative risk was 0.51 (0.3–0.88). A recently completed Dutch SDD trial presented at a scientific meeting  reports an ICU mortality odds ratio of 0.6 (0.4–0.8). Of note, instead of randomizing individual patients this group randomized new admissions to a ward where either SDD or standard care was used. The ICU mortality reduction, which persisted after ICU discharge, is unexpectedly large, being twice that expected from the results of the most optimistic meta-analyses to date.
Has the question of the effect of SDD on ICU-mortality been answered conclusively? Specifically, are the results of the latter two studies generalizable to other centers? Very importantly, baseline levels of antibiotic resistance were extremely low in both studies. Methicillin-resistant Staphylococcus aureus (MRSA) was rarely  or not at all  isolated from surveillance cultures. Moreover, at least some will question how such a large reduction in mortality could be found, while other studies have failed to do so for 20 years. Nevertheless, the results of these studies are provocative, and the accumulating evidence that SDD improves patient outcome warrants confirmatory studies. It is also clear that patient outcome, and not the incidence of nosocomial infections, should be the endpoint of such trials. We also believe that further studies on the relative importance of the individual components of SDD are needed.
SDD and antibiotic resistance
Antibiotic resistance is an emerging global problem, driven in large part by antibiotic use. Rational and restrictive use of antibiotics, limitation of cross-infections, and prevention of device-associated infections are the tools that we have for controlling resistance. There are multiple examples—at the level of single wards, hospital settings, and even countries—that these tools can be effective. Despite this knowledge and experience, presented in many guidelines, editorials, and reviews, the problem of antibiotic resistance continues to grow.
Van Saene and coauthors report that SDD is the answer to this problem. ICUs are generally considered the "epicenters" of antibiotic resistance. Problems with resistant micro-organisms become most evident here because the intensity of antibiotic use and vulnerability of patients to infections are highest. Mechanisms contributing to an increase in the number of patients colonized (or infected) with antibiotic resistant pathogens are: (a) selection (or induction) of resistant flora due to antibiotic use, (b) introduction of patients already colonized, and (c) transmission of micro-organisms or resistance genes due to lapses in infection control. These pathways are mutually interactive, and the relative importance of each of the routes is usually unknown.
How can SDD affect these processes, thereby decreasing the development and spread of antibiotic resistance? The chance of cross-transmission will decrease when the number of patients who are colonized by problematic pathogens (i.e., "colonization pressure") is reduced. Presumably SDD would reduce colonization pressure for the micro-organisms susceptible to the antibiotics used. Conversely, it is unlikely that SDD would control micro-organisms that are resistant to the regimen used [e.g., MRSA, vancomycin-resistant enterococci (VRE), Acinetobacter species, multiresistant Pseudomonas aeruginosa]. In fact, studies suggest that in settings where resistance is endemic SDD may actually increase resistance [6, 27, 28, 29]. The potential beneficial effect of SDD in appropriate settings must be considered in light of other infection control measures, such as hand disinfection and glove use. For example, when on average 20–40% of the patients in an ICU are colonized with resistant bacteria, and adherence to infection control measures exceeds the threshold necessary to prevent transmission, SDD will not contribute to control of resistance. However, when average compliance falls below that threshold, transmission could be prevented by a reduction in colonization pressure due to the use of SDD. In this way SDD could be seen as an adjunct to an infection control program, and there is experience with the temporary use of SDD in outbreak situations . Hand hygiene is always mentioned as an essential complement to SDD, but levels of adherence to this or other infection control measures have never been quantified in SDD trials.
Van Saene et al. state that overgrowth of intestinal flora and suppressing the anaerobic flora "guarantee development of mutant antibiotic resistant strains." However, the data cited to confirm this hypothesis describe only outbreaks or selection of preexisting resistant flora, such as VRE . On the other hand, the hypothesis has not been refuted, and this mechanism may apply for resistance to cephalosporins due to inducible production of extended-spectrum β-lactamases. Since resistance is so widespread, and SDD is not, the argument, however, that resistance occurs frequently in settings not using SDD does not prove that SDD is protective!
After 20 years of clinical experience with SDD the initial fear that SDD would increase induction or mutation rates towards resistance seems less important, at least in areas with an initial low prevalence of resistance. Selection of preexisting resistant pathogens, however, remains a serious threat. The most important pathogens in this regard are MRSA , multiresistant Acinetobacter species , and VRE. Because of their lack of activity on these micro-organisms conventional SDD regimens do select for enterococci , and unexpected clinical infections have been reported . To our knowledge, SDD has never been studied in a setting where VRE are endemic. It can be expected that SDD will increase selective pressure for VRE, and, even more worrisome, for vancomycin-resistant Staphylococcus aureus. The first clinical isolates of these bacteria, harboring the vanA transposon previously found in VRE, have recently been reported in the United States [33, 34]. Many questions remain on the issue of antibiotic resistance and SDD.
Interestingly, only few countries (i.e., Scandinavian countries, The Netherlands, and Switzerland) still have reasonably low rates of antibiotic resistance (such as MRSA, extended-spectrum β-lactamase producing Enterobacteriaceae, and multiresistant P. aeruginosa) among nosocomial pathogens. SDD was not necessary to control resistance in these countries where stringent preventive measures including both control of antibiotic use and infection control have been adhered to since the early antibiotic era. Because most other countries have been less successful in controlling antibiotic resistance, their ICUs are less attractive settings for SDD. Thus there are countries that do not need SDD to control antibiotic resistance, and where SDD could be used to prevent nosocomial infections and may be to improve patient outcome. In the majority of the countries antibiotic resistance has increased to such a level that the widespread use of SDD may only worsen the situation.
In conclusion, SDD can prevent some episodes of VAP, and there is early tantalizing evidence that SDD may improve patient outcome in moderately ill ICU patients. These provocative results should be confirmed in various ICU settings, and the relative importance of the individual components of SDD should be evaluated. Moreover, more studies should determine the relative benefits of SDD in specific patient populations (i.e., medical, surgical or trauma) and in subgroups of patients with different levels of illness. Notwithstanding, SDD has provided us with multiple interesting and testable hypotheses that should continue to be investigated. In ICUs where antimicrobial resistance is endemic SDD can create a selective growth advantage and should be avoided.
- 1.Saene HFK van, Petros AJ, Ramsay G, Baxby D (2003) All great truths are iconoclastic: selective decontamination of the digestive tract moves from heresy to level 1 truth. Intensive Care Med (http://dx.doi.org/10.1007/s00134-003-1722-2)Google Scholar
- 2.Bergmans DC, Bonten MJ, Gaillard CA, Paling JC, van der Geest S, Van Tiel FH, Beysens AJ, de Leeuw PW, Stobberingh EE (2001) Prevention of ventilator-associated pneumonia by oral decontamination: a prospective, randomized, double-blind, placebo-controlled study. Am J Respir Crit Care Med 164:382–388CrossRefGoogle Scholar
- 20.Bonten MJM, Gaillard CA, van der Geest S, van Tiel F, Beysens AJ, Smeets HGW, Stobberingh EE (1995) The role of intragastric acidity and stress ulcer prophylaxis on colonization and infection in mechanically ventilated patients. A stratified, randomized, double blind study of sucralfate versus antacids. Am J Respir Crit Care Med 152:1825–1834CrossRefGoogle Scholar
- 25.Krueger WA, Lenhart FP, Neeser G, Ruckdeschel G, Schreckhase H, Eissner HJ, Forst H, Eckart J, Peter K, Unertl KE (2002) Influence of combined intravenous and topical antibiotic prophylaxis on the incidence of infections, organ dysfunctions, and mortality in critically ill surgical patients: a prospective, stratified, randomized, double-blind, placebo-controlled clinical trial. Am J Respir Crit Care Med 166:1029–1037CrossRefGoogle Scholar
- 26.de Jonge E, Schultz M, Spanjaard L, et al (2002) Effects of selective decontamination of the digestive tract on mortality and antibiotic resistance. Intensive Care Med 28: S12Google Scholar
- 33.Miller D, Urdaneta V, Weltman A (2002) Vancomcyin-resistant Staphylococcus aureus–Pennsylvania. MMWR Morb Mortal Wkly Rep 51:902Google Scholar
- 34.Sievert DM, Boulton ML, Stoltman G (2002) Staphylococcus aureus resistant to vancomcyin-United States. MMWR Morb Mortal Wkly Rep 51:565–566Google Scholar