Correspondence

Prevention of Gastrointestinal Bleeding during Mechanical Ventilation

N Engl J Med 1998; 339:266-268July 23, 1998

Article

To the Editor:

Cook et al. (March 19 issue)1 found that among critically ill patients requiring mechanical ventilation, those receiving ranitidine had a significantly lower rate of clinically important gastrointestinal bleeding than those treated with sucralfate. There were no significant differences in the rates of ventilator-associated pneumonia, the duration of the stay in the intensive care unit (ICU), or mortality. Despite the authors' best efforts, their results remain difficult to interpret. It is not clear whether the two groups were actually matched for severity of illness, since the authors provide only the mean scores for the Acute Physiology and Chronic Health Evaluation (APACHE II) and not the APACHE II estimates of the risk of death in the hospital. Furthermore, given the high mean (±SD) scores (24.6±7.3 in the sucralfate group and 24.7±7.1 in the ranitidine group) and the relatively low mortality rates in the ICU (22.8 percent and 23.5 percent, respectively), either there was a high death rate after discharge from the ICU (which might substantially alter the results of the study) or the APACHE II system as a method of calibration did not work in this group of patients.2 This failure of calibration (in this case, the predicted mortality rate was high and the actual rate was low, presumably across the entire spectrum of risk) may well have been associated with a failure to discriminate between hospital survivors and patients who died.

Enteral nutrition by itself may well prevent both gastrointestinal bleeding3 and ventilator-associated pneumonia.4 The rates of early, successful enteral nutrition are not given and may have differed between the two groups.

Given the risks and costs of the various prophylactic regimens and the absence of any meaningful difference in clinical outcome with different but very low rates of bleeding, we remain to be convinced that any specific prophylaxis against stress ulcers is necessary in this group of patients.

Imogen Mitchell, F.R.A.C.P.
Royal Prince Alfred Hospital, Camperdown, Sydney, NSW 2050, Australia

David Bihari, F.R.A.C.P.
St. George Hospital, Kogarah, Sydney, NSW 2217, Australia

4 References

1. 1

   Cook D, Guyatt G, Marshall J, et al. A comparison of sucralfate and ranitidine for the prevention of upper gastrointestinal bleeding in patients requiring mechanical ventilation. N Engl J Med 1998;338:791-797
   Full Text | Web of Science | Medline

2. 2

   Hosmer DW, Hosmer T, Le Cessie S, Lemeshow S. A comparison of goodness-of-fit tests for the logistic regression model. Stat Med 1997;16:965-980
   CrossRef | Web of Science | Medline

3. 3

   Pingleton SK, Hadzima SK. Enteral alimentation and gastrointestinal bleeding in mechanically ventilated patients. Crit Care Med 1983;11:13-16
   CrossRef | Web of Science | Medline

4. 4

   Heyland DK, Cook DJ, Guyatt GH. Enteral nutrition in the critically ill patient: a critical review of the evidence. Intensive Care Med 1993;19:435-442
   CrossRef | Web of Science | Medline

To the Editor:

The pH of most commercially available enteral-nutrition formulations is 6 to 7. A continuous infusion of one of these formulations at the rate of about 100 ml per hour keeps the pH high enough to render the administration of ranitidine unnecessary, unless ranitidine has some effect other than merely increasing the intraluminal gastric pH. Moreover, the theory behind the potential beneficial effect of sucralfate relies on the stomach's being able to generate a pH low enough to eradicate microbial growth (normally, <3.5 to 4). Only if patients are being fed intermittently is it possible for intraluminal gastric pH to fall to levels that are bactericidal. Cook and colleagues did not measure intraluminal pH on admission or at any stage during the patients' stay in the ICU, mainly because this might have allowed the therapies to be identified.

Were all the patients capable of maintaining a low pH, a high-energy, blood-flow–dependent process that is not always possible in the case of about 50 percent of critically ill patients? It is possible that the results of Cook et al. differ from those of previous meta-analyses1,2 because they did not control for these variables and they included a large number of patients at low risk for gastrointestinal bleeding (i.e., those who tolerate enteral nutrition3).

Duncan L.A. Wyncoll, M.B., B.S.
Royal Brompton Hospital, London SW3 6NP, United Kingdom

3 References

1. 1

   Cook DJ, Reeve BK, Guyatt GH, et al. Stress ulcer prophylaxis in critically ill patients: resolving discordant meta-analyses. JAMA 1996;275:308-314
   CrossRef | Web of Science | Medline

2. 2

   Tryba M. Sucralfate versus antacids or H2-antagonists for stress ulcer prophylaxis: a meta-analysis on efficacy and pneumonia rate. Crit Care Med 1991;19:942-949
   CrossRef | Web of Science | Medline

3. 3

   Cook DJ, Fuller HD, Guyatt GH, et al. Risk factors for gastrointestinal bleeding in critically ill patients. N Engl J Med 1994;330:377-381
   Full Text | Web of Science | Medline

To the Editor:

Did Cook et al. stop the tube feedings one hour before the administration of sucralfate or its placebo and for at least one hour afterward, as suggested by the manufacturer? Sucralfate acts locally rather than systemically by reacting with hydrochloric acid in the stomach to form a pastelike substance that adheres electrostatically to proteins on the surface of damaged gastric mucosa.1,2 These dosing precautions are needed for sucralfate to provide optimal gastric protection.

How does one give oral medications to endotracheally intubated patients without the risk of bronchial aspiration? Furthermore, did the patients have only nasogastric tubes in place, or did some have post-pyloric tubes? Sucralfate given by a post-pyloric tube would not provide gastric protection.

Paul Marik, M.D.
St. Vincent Hospital, Worcester, MA 01604

Joseph Varon, M.D.
Baylor College of Medicine, Houston, TX 77030

2 References

1. 1

   Rees WD. Mechanisms of gastroduodenal protection by sucralfate. Am J Med 1991;91:Suppl 2A:58S-63S
   CrossRef | Web of Science | Medline

2. 2

   McCarthy DM. Sucralfate. N Engl J Med 1991;325:1017-1025
   Full Text | Web of Science | Medline

To the Editor:

In designing their study, Cook et al. assumed that the efficacy of prophylaxis against stress-ulcer–induced hemorrhage is established. This is questionable, particularly in the case of clinically important bleeding. A recent meta-analysis showed significant heterogeneity in the comparison between histamine H₂-receptor antagonists and either placebo or no therapy.1 Another trial found no difference in the rates of bleeding or in the number of units of blood transfused between treated patients and untreated patients.2 In an earlier study by Cook et al.,3 it appears that among patients with respiratory failure, coagulopathy, or both, bleeding was more frequent in the group that received prophylaxis than in the group that did not receive prophylaxis (22 of 369 patients, or 6.0 percent, vs. 9 of 447 patients, or 2.0 percent). We wonder why, in their current study, they did not include a control group that received no treatment. We also would like to know whether they can provide data on the number of units of blood transfused.

Cook et al. mention on the basis of their previous study3 that the incidence of gastrointestinal bleeding is 3.7 percent among untreated patients who are receiving mechanical ventilation. This incidence is similar among patients who received prophylaxis.

Georges Offenstadt, M.D.
Bertrand Guidet, M.D.
Eric Maury, M.D.
Hôpital Saint-Antoine, 75571 Paris CEDEX 12, France

3 References

1. 1

   Cook DJ, Reeve BK, Guyatt GH, et al. Stress ulcer prophylaxis in critically ill patients: resolving discordant meta-analyses. JAMA 1996;275:308-314
   CrossRef | Web of Science | Medline

2. 2

   Ben-Menachem T, Fogel R, Patel RV, et al. Prophylaxis for stress-related gastric hemorrhage in the medical intensive care unit: a randomized, controlled, single-blind study. Ann Intern Med 1994;121:568-575
   Web of Science | Medline

3. 3

   Cook DJ, Fuller HD, Guyatt GH, et al. Risk factors for gastrointestinal bleeding in critically ill patients. N Engl J Med 1994;330:377-381
   Full Text | Web of Science | Medline

Author/Editor Response

The authors reply:

To the Editor: Drs. Mitchell and Bihari wonder whether the two groups were significantly different at base line despite the randomization of the 1200 patients. Patients were similar with respect to the physiologic and clinical characteristics that we measured in addition to the 20 we reported. The results were the same whether or not the analyses were adjusted for APACHE II scores. Although the observational evidence and the experimental evidence of the magnitude and direction of the effect of enteral feeding on the rates of bleeding and pneumonia yield discordant results, interpreting our trial in the light of nutritional support is crucial. The type, timing, and tolerance of feeding were similar in the two groups. We believe that a 50 percent reduction in the relative risk of a serious bleeding event in patients with a base-line risk of 4 percent constitutes a clinically important difference, in keeping with our estimate that 50 patients would need to be treated with ranitidine instead of sucralfate to prevent one episode of bleeding. Decisions in this area should ideally consider individual base-line risks and estimates of benefit, harm, and cost with the use of different drugs and prophylactic approaches (universal or targeted to high-risk groups).

Dr. Wyncoll raises key mechanistic questions about gastric pH and gastric blood flow that our blinded efficacy trial was not designed to address. The lower (by 50 percent) bleeding rate associated with ranitidine was consistent in the total cohort of patients as well as those receiving full, partial, or no enteral nutrition.

Drs. Marik and Varon ask about the administration of sucralfate. Patients received sucralfate through a nasogastric tube; if enteral feeding was distal, such as by means of jejunostomy, sucralfate was still delivered into the stomach by nasogastric tube. At the discretion of each ICU team, feeding was withheld for up to one hour after the administration of sucralfate. We agree that prevention of aspiration is an important, though perhaps elusive, goal in patients receiving ventilation. Randomized trials evaluating the effect of body position on the likelihood of aspiration suggest that a semirecumbent position is associated with a lower rate of scintigraphic evidence of aspiration than a supine position.1-3

Dr. Offenstadt and colleagues refer to our meta-analysis comparing histamine H₂-receptor antagonists with no prophylaxis.4 We found that the outcome was heterogeneous with respect to the rates of overt bleeding (macroscopic evidence of blood, regardless of the amount) but not with respect to the outcome of clinically important bleeding (associated with hemodynamic instability and the need for transfusion). We caution against making inferences about the effect of prophylaxis from our natural-history study5 and suggest turning to randomized trials designed for this purpose. Patients with clinically important bleeding in the sucralfate group received a total of 142 units of blood (mean, 6.2; median, 5), whereas those receiving ranitidine received 76 units (mean, 7.6; median, 4).

Deborah Cook, M.D.
Gordon Guyatt, M.D.
McMaster University, Hamilton, ON L8N 4A6, Canada

David Leasa, M.D.
University of Western Ontario, London, ON N6A 5A5, Canada

for the Canadian Critical Care Trials Group

5 References

1. 1

   Torres A, Serra-Batlles J, Ros E, et al. Pulmonary aspiration of gastric contents in patients receiving mechanical ventilation: the effect of body position. Ann Intern Med 1992;116:540-543
   Web of Science | Medline

2. 2

   Ibanez J, Penafiel A, Raurich JM, Marse P, Jorda R, Mata F. Gastroesophageal reflux in intubated patients receiving enteral nutrition: effect of supine and semirecumbent positions. JPEN J Parenter Enteral Nutr 1992;16:419-422
   CrossRef | Web of Science | Medline

3. 3

   Orozco-Levi M, Torres A, Ferrer M, et al. Semirecumbent position protects from pulmonary aspiration but not completely from gastroesophageal reflux in mechanically ventilated patients. Am J Respir Crit Care Med 1995;152:1387-1390
   Web of Science | Medline

4. 4

   Cook DJ, Reeve BK, Guyatt GH, et al. Stress ulcer prophylaxis in critically ill patients: resolving discordant meta-analyses. JAMA 1996;275:308-314
   CrossRef | Web of Science | Medline

5. 5

   Cook DJ, Fuller HD, Guyatt GH, et al. Risk factors for gastrointestinal bleeding in critically ill patients. N Engl J Med 1994;330:377-381
   Full Text | Web of Science | Medline