- Email: [email protected]
A randomized, double-blind trial for stress ulcer prophylaxis shows no evidence of increased pneumonia
A Randomized, Double-Blind Trial for Stress Ulcer Prophylaxis Shows No Evidence of Increased Pneumonia Ernst W. Hanisch, MD, Albrecht Encke, MD, FACS, Frank Naujoks, MD, Joachim Windolf, MD, Frankfurt/Main, Germany
H2-receptor antagonists are commonly used for stress ulcer prophylaxis on intensive care units. However, there is evidence that via the route of an elevated gastric pH, followed by bacterial overgrowth and subsequent tracheal aspiration, pneumonia could occur. In line with this assumption total gastrectomized patients should develop a very high incidence of pneumonia, which is actually not the case. We therefore formulated the hypothesis that stress ulcer prophylaxis with H2-receptor antagonists does not lead to an increased pneumonia rate. METHODS: A total of 158 patients with mechanical ventilation ≥48 hours of a surgical intensive care unit were randomized to the following groups: A, placebo (n 5 57); B, pirenzepine (3 3 10 mg intravenously, n 5 44); and C, ranitidine (3 3 50 mg intravenously, n 5 57). RESULTS: The pneumonia rate in ranitidine-, pirenzepine-, and placebo-treated patients is 10 of 57, 10 of 44, and 12 of 57, respectively. CONCLUSIONS: Pneumonia rate is not adversely affected by H2-receptor antagonists in stress ulcer prophylaxis. Am J Surg. 1998;176:453– 457. © 1998 by Excerpta Medica, Inc. BACKGROUND:
I
n the last 15 years a decline in the incidence of stress ulcer bleeding has been observed. One recent study reported even a virtual absence of stress-ulcer–related bleeding of patients in intensive care units (ICU) receiving prolonged mechanical ventilation without any prophylaxis.1 It is generally believed that this phenomenon is not exclusively related to the widespread use of H2-receptor antagonists in ICU, but has to be interpreted in association with multiple improvements in anesthesiological and surgical techniques. Surprisingly, the diminished rate of stress ulcer bleeding had no influence upon the overall mortality of patients in ICU.2 In this context the argument of an increasing effect of H2-receptor antagonists on the pneumonia rate, one of the leading causes of death in ICU
From the Department of Surgery (EWH, AE, FN), Division of General and Trauma Surgery (JW), Johann Wolfgang Goethe University, Frankfurt am Main, Germany. Requests for reprints should be addressed to E. W. Hanisch, MD, Department of Surgery, Johann Wolfgang Goethe University, Theodor-Stern-Kai 7, 60590 Frankfurt/Main, Germany. Manuscript submitted February 26, 1998 and accepted in revised form August 17, 1998.
© 1998 by Excerpta Medica, Inc. All rights reserved.
emerged.3,4 It was discussed that via the route of an elevated gastric pH, followed by bacterial overgrowth and subsequent tracheal aspiration, pneumonia would develop.5 This assumption, however, is subject to controversial discussions. In line with the above outlined hypothesis, total gastrectomized patients should develop a very high incidence of pneumonia, which is actually not the case. We therefore formulated the hypothesis that stress ulcer prophylaxis with H2-receptor antagonists does not lead to an increased pneumonia rate. Since only few double-blind, placebo-controlled trials do exist (for an overview see Table I) with quite differing pneumonia rates and results, patients were randomly assigned to a group with no stress ulcer prophylaxis. The trial considers also pirenzepine, which has only minor effects upon the gastric pH in comparison with H2receptor antagonists.
METHODS All patients referred to the intensive care unit of the surgical department of the Johann Wolfgang Goethe-University Frankfurt/Main were considered for the study. Exclusion criteria were patients with an active peptic ulcer disease and a concomittant ulcer medication; patients with upper gastrointestinal bleeding; patients ,18 years; transplanted patients (kidney, liver, heart); and patients with preexisting pneumonia and gastric resection. At the time of referral to the intensive care unit of the surgical department all patients were randomized to the following groups: A, placebo; B, pirenzepine (Gastrozepin, 3 3 10 mg intravenously, Thomae, Biberach, Germany); or C, ranitidine (Sostril, 3 3 50 mg intravenously, Cascan, Wiesbaden, Germany). The primary outcome measure was the pneumonia rate in patients under mechanical ventilation $48 hours. The definition of pneumonia was according to Daschner13: radiological signs of pneumonia and purulent tracheal secretion or positive microbiological findings in tracheal aspiration and temperature .38°C and leucocytosis .10,000 mm3. In all patients with radiological signs of pneumonia a bronchioalveolar lavage was routinely performed, but no quantitative cultures were done. Besides the primary study endpoint, pneumonia, clinically relevant stress bleeding was also documented. The definition of clinically relevant bleeding was bright red blood via gastric tube or melena combined with hemodynamic changes (systolic blood pressure ,100 mm Hg, tachycardia .100 beats per minute) and requirement of blood transfusion (fall in hemoglobin 0002-9610/98/$19.00 PII S0002-9610(98)00239-6
453
STUDY OF STRESS ULCER PROPHYLAXIS/HANISCH ET AL
TABLE I Placebo-controlled Studies: Effect on Pneumonia Rate n
Patients
Cheadle 1985 Reusser7 1989
200 40
Karlstadt8 1990 Apte9 1992 Metz10 1993 Martin11 1993 Ben-Menachem12 1994
87 34 167 131 300
Surgical Neurosurgical ventilation .48 hours Surgical/medical Tetanus, tracheotomized Severe head injury Surgical/medical Medical
6
Intervention
Pneumonia Rate
Cimetidine double blind Ranitidine/antacid
Placebo 1.5%, cimetidine 6.5% Placebo 38%, ranitidine 36%
Cimetidine double blind Ranitidine and/or antacid Ranitidine double blind Cimetidine double blind Cimetidine, sucralfate single blind
Placebo 6%, cimetidine 0% Placebo 50%, ranitidine 81% Placebo 19%, ranitidine 14% Placebo 9%, cimetidine 0% Placebo 6%, cimetidine 13%, sucralfate 12%
Figure. Trial profile.
.2 g/dL within 24 hours) and endoscopic identification of bleeding site and activity. Assuming a pneumonia rate of 25% in ranitidine-treated patients and a pneumonia rate of 15% in placebo-treated patients, the complete sample size was calculated to be at least 100 patients (a 5 0.05; b 5 0.20). A complete and balanced randomization schedule was generated by the Institute of Biomathematics of the University of Frankfurt. At the time of entering the ICU 454
patients were assigned to a consecutive study number, and the application of the blinded drug regimen was started. Blinding of drugs (placebo, ranitidine, pirenzepine) was achieved by dissolving them in glucose (100 mL; 5%; Braun, Melsungen, Germany) resulting in an equal infusion volume for each patient. Drugs were prepared in advance by a staff not involved in the treatment of patients according to the randomization schedule and identified by the above mentioned consecutive study number, ie, the
THE AMERICAN JOURNAL OF SURGERY® VOLUME 176 NOVEMBER 1998
STUDY OF STRESS ULCER PROPHYLAXIS/HANISCH ET AL
TABLE II General Patient Characteristics, Pneumonia, Stress Bleeding, and Mortality in Patients Mechanically Ventilated ≥48 Hours Ranitidine Patients (n) Age (years) Intensive care unit stay (days) APACHE II score Mechanical ventilation (days) Pneumonia (n)* Stress bleeding (n)† Mortality (n)
Pirenzepine
Placebo
57
44
55/22–88
53/18–86
58/22–88
57
9.7/2–95 19/2–30
9.9/2–39 21/12–34
12.6/2–58‡ 18/1–28
8.2/2–93 10 3 7
8/2–32 10 3 12
10.2/2–55§ 12 2 12
Data are presented as means and full range. * Chi square 5 3.55; DF 5 2; P 5 0.17. † Chi square 5 1.79; DF 5 2; P 5 0.41. ‡ Chi square 5 7.54; DF 5 2; P 5 0.02. § Chi square 5 8.54; DF 5 2; P 5 0.01.
code was located exclusively in this group during the trial. All blinded drugs (only identifiable by their consecutive study number) were given to the patients by the nurse staff of the ICU. Data documentation for each patient14 was performed prospectively and daily by different staff not involved in the treatment of patients and in the randomized assignment of drugs. Only one of the authors (FN) was a member of these two staff groups at different time periods. The complete trial ran between December 1, 1991, and December 30, 1992. Following several discussions of the study protocol at workshops of the “Project Group for Clinical Studies” of the German Surgical Society, the trial was approved by the Ethics Committee of the Johann Wolfgang Goethe University, Frankfurt/Main. Statistical Analysis Qualitative data were compared by the chi-square test; quantitative data were discriminated by the Kruskal-Wallis test.
RESULTS For the trial profile, see the Figure. Of 827 randomized patients, a total of 158 were mechanically ventilated $48 hours. The mean duration of the mechanical ventilation was 8.2, 8, and 10.2 days in patients receiving ranitidine, pirenzepine, and placebo, respectively. General characteristics, pneumonia, stress bleeding, and mortality of these patients are summarized in Table II. There is no statistical difference between the groups with respect to the pneumonia rate. Stress ulcer bleeding had no serious clinical sequelae in patients with ranitidine (blood transfusions, 2 units each) and pirenzepine (blood transfusion, 4 units each) except in 2 placebo patients. Both underwent cardiac surgery (aortic valve replacement; bypass surgery) with postoperative medication of coumarin, aspirin, and cortisone, respec-
tively. The first patient died owing to his ulcer bleeding, the second one received 12 units of blood. Bleeding had been controlled by endoscopic intervention. After the extension of exclusion criteria, no further bleeding was observed in placebo-treated patients. (The directly related death of a patient due to ulcer bleeding led the monitor to break the code. Since the patient belonged to the placebo group, exclusion criteria were extended by postoperative drugs like aspirin, cortisone, and coumarin.) Of 741 notrandomized patients, 22 developed pneumonia and 3 demonstrated clinical relevant bleeding. Table III summarizes the data of clinical relevant stress ulcer bleeding during the complete trial.
COMMENTS Prophylaxis has unequivocally decreased the formerly frightening stress ulcer bleeding of patients in ICU, although the general mortality of this population could not be reduced. This obvious discrepancy stimulated numerous investigations, and the knowledge of nosocomial infection as critical determinant in this context was established.15 According to this hypothesis, the pH-elevating effect of H2-receptor antagonists is prompting gastric bacterial growth, which in turn causes more pneumonias via an esophageal ascension and subsequent tracheal aspiration. Thus, giving sucralfate, which does not alter gastric pH, yields significantly less pneumonias.16 In this prospective randomized study, late pneumonias occurred to a significantly smaller extent in the sucralfate group than in the ranitidine group (5% versus 21%). In a recent study that compared the results of sucralfate versus maximal H2blocker infusion therapy plus antacids, the overall pneumonia rate was 27.5% in the H2-blocker antacid group and 29.8% in the sucralfate group, with a significant level of P 5 0.48. 17 This study implies a considerable reduction of charges/cost for the sucralfate approach. Contrary to the above-mentioned studies, in another randomized trial with trauma patients, no difference could be found.18 Our data, placebo-controlled, are corroborating these results (Table II). Furthermore, there are also placebo-controlled studies in long-term ventilated patients supporting the results of our study. Thus, Reusser et al7 randomized 40 neurosurgical patients requiring mechanical ventilation for .48 hours to either a control group receiving no stress ulcer prophylaxis or to a ranitidine/antacid group. There was no significant difference in the pneumonia rate of patients with and without stress ulcer prophylaxis. In line with these findings, Metz et al10 also could not find any difference in 167 placebo-controlled, randomized patients with severe head injury. On the contrary, Apte et al9 demonstrated for 34 randomized tracheotomized patients with tetanus, receiving intravenous ranitidine (50 mg, 6 hourly and/or antacids) or neither ranitidine nor antacids, significantly more pneumonias in the ranitidine group (81%) than in the control group (50%) were found. Moreover, patients treated with ranitidine developed pneumonia significantly earlier than patients in control groups. With respect to pirenzepine only a few trials are available, the largest one having been conducted by Tryba et al.19 In that study, 400 patients were randomized to a ranitidine (200 mg) or a pirenzepine (50 mg) group. Eight out of 28
THE AMERICAN JOURNAL OF SURGERY® VOLUME 176 NOVEMBER 1998
455
456
THE AMERICAN JOURNAL OF SURGERY® VOLUME 176 NOVEMBER 1998
Survivor
Survivor
Death at 31st day postoperative
Forrest Ia (duodenal ulcer) Death at 12th day postoperative Erosive gastritis Erosive gastritis
Survivor Death at 43rd day postoperative Mortality
Forrest IIb (angulus)
Forrest IIb (corpus) Survivor
Erosive gastritis
4 13 1 12 18 10 29 21
Pirenzepine Pirenzepine
13 24
Erosive gastritis
TABLE III
Ranitidine
REFERENCES 1. Zandstra DF, Stoutenbeek CP. The virtual absence of stressulceration related bleeding in ICU patients receiving prolonged mechanical ventilation without any prophylaxis. A prospective cohort study. Intens Care Med. 1994;20:335–340. 2. Schuster DP. Stress ulcer prophylaxis: in whom? With what? Crit Care Med. 1993;21:4 – 6. 3. Tryba M, Zevounou F, Wruck G. Streßblutungen und postoperative Pneumonien bei Intensivpatienten unter Ranitidin oder Pirenzepin. Dtsch Med Wschr. 1988;113:930 –936. 4. Driks MR, Craven DE, Celli BR, et al. Nosocomial pneumonia in intubated patients given sucralfate as compared with antacids or histamine type 2 blockers. NEJM. 1987;317:1376 –1382. 5. Cook DJ, Laine LA, Guyatt GH, Raffin ThA. Nosocomial pneumonia and the role of gastric pH. A meta-analysis. Chest. 1991;11:7–13. 6. Cheadle WG, Vitale GC, Mackie CR, Cuschieri A. Prophylactic postoperative nasogastric decompression. Ann Surg. 1985;202: 361–366. 7. Reusser P, Zimmerli W, Scheidegger D, et al. Role of gastric
Ranitidine Ranitidine
Synopsis of Clinically Relevant Bleeding in 8 Patients
Pirenzepine
patients on mechanical ventilation receiving ranitidine developed pneumonias (28.6%) versus 3 out of 33 patients on mechanical ventilation under pirenzepine (9.1%), which is statistically significant at a level of P ,0.05. One reason for the striking difference from our results might be the different ranitidine and pirenzepine dosages used. Finally, the comparable pneumonia rates in our pirenzepine and control group patients are underlining the knowledge that H2-receptor antagonists with the dosage used do not increase the pneumonia risk of long-term ventilated patients in critical condition. In addition, confounding parameters like nasogastric tubes and tracheostomas are obviously of no statistical relevance in this context (Table IV). Moreover, Cook found no significant differences in the rates of ventilator-assisted pneumonia between ranitidine- and sucralfate-treated patients in a multicenter, randomized trial.20 In conclusion, we can demonstrate that in surgical critically ill patients on mechanical ventilation for $48 hours, stress ulcer prophylaxis with H2-receptor antagonists pneumonia rate is not adversely affected. Since the incidence of stress gastritis in our study is only 3.5% for patients on mechanical ventilation $48 hours in the placebo group, one might generally question prophylaxis. Of course, we are not allowed to draw any definite conclusion about this point, because the number of patients is too low for statistical analysis with respect to the outcome measure “clinically relevant bleeding.” Conceivably, improvements in the management of critically ill patients per se might have changed the likelihood of developing stress gastritis over the last decade.
12 21
35 13
Mechanical Ventilation (d) Days of bleeding postoperative Endoscopic findings
26 7
Aortic valve disease
20 8
Polytrauma
55
Head trauma
42
Coronary heart disease
53
Polytrauma Carcinoma of the floor of the mouth, liver cirrhosis 12 2 10 9
Placebo
Ileus
Pirenzepine
Diagnosis
Ranitidine
Placebo
Nasogastric tube .48 hours (n) Nasogastric tube .4 days (n) Tracheostoma (n)
Placebo
Nasogastric Tube Duration and Tracheostoma in Patients Mechanically Ventilated ≥48 Hours
Coronary heart disease
TABLE IV
Forrest Ib (duodenal ulcer) Survivor
STUDY OF STRESS ULCER PROPHYLAXIS/HANISCH ET AL
STUDY OF STRESS ULCER PROPHYLAXIS/HANISCH ET AL
colonization in nosocomial infections and endotoxemia: a prospective study in neurosurgical patients. J Infect Dis. 1989;160:414 – 421. 8. Karlstadt RG, Iberti TJ, Silverstein J, et al. Comparison of cimetidine and placebo for the prophylaxis of upper gastrointestinal bleeding due to stress-related gastric mucosal damage in the intensive care unit. J Intens Care Med. 1990;5:26 –32. 9. Apte NM, Karnad DR, Medhekar TP, et al. Gastric colonization and pneumonia in intubated critical ill patients receiving stress ulcer prophylaxis: a randomized, controlled trial. Crit Care Med. 1992;20:590 –593. 10. Metz CA, Livingston DH, Smith JS, et al. Impact of multiple risk factors and ranitidine prophylaxis on the development of stress-related upper gastrointestinal bleeding: a prospective, multicenter, double-blind, randomized trial. Crit Care Med. 1993;21: 1844 –1849. 11. Martin LF, Booth F, Karlstadt RG, et al. Continuous intravenous cimetidine decreases stress-related upper gastrointestinal hemorrhage without promoting pneumonia. Crit Care Med. 1993;21: 19 –30. 12. Ben-Menachem T, Fogel R, Patel RV, et al. Prophylaxis for stress-related gastric hemorrhage in the medical intensive care unit. Ann Intern Med. 1994;121:568 –575.
13. Daschner F, Reuschenbach K, Pfisterer I, et al. Der Einfluß von Streßulkusprophylaxe auf die Ha¨ufigkeit einer Beatmungspneumonie. Anaesthesist. 1987;36:9 –18. 14. Windolf J, Gottschalk S, Inglis R, Pannike A. Einsatz eines elektronischen Datenerfassungssystems in der chirurgischen Intensivmedizin. Unfallchirurgie. 1990;16:252–257. 15. Tryba M. Side effects of stress bleeding prophylaxis. Am J Med. 1989;86(suppl 6A):85–93. 16. Prod’hom G, Leuenberger P, Koerfer J, et al. Nosocomial pneumonia in mechanically ventilated patients receiving antacid. Ann Intern Med. 1994;120:653– 662. 17. Maier RV, Mitchell D, Gentilello L. Optimal therapy for stress gastritis. Ann Surg. 1994;220:353–363. 18. Pickworth KK, Falcone RE, Hoogeboom JE, Santanello SA. Occurrence of nosocomial pneumonia in mechanically ventilated trauma patients. Crit Care Med. 1993;12:1856 –1862. 19. Tryba M. Prevention of stress bleeding with ranitidine or pirenzepine and the risk of pneumonia. J Clin Anaesth. 1988;1:12– 20. 20. 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. NEJM. 1998;338: 791–797.
THE AMERICAN JOURNAL OF SURGERY® VOLUME 176 NOVEMBER 1998
457
H2-receptor antagonists are commonly used for stress ulcer prophylaxis on intensive care units. However, there is evidence that via the route of an elevated gastric pH, followed by bacterial overgrowth and subsequent tracheal aspiration, pneumonia could occur. In line with this assumption total gastrectomized patients should develop a very high incidence of pneumonia, which is actually not the case. We therefore formulated the hypothesis that stress ulcer prophylaxis with H2-receptor antagonists does not lead to an increased pneumonia rate. METHODS: A total of 158 patients with mechanical ventilation ≥48 hours of a surgical intensive care unit were randomized to the following groups: A, placebo (n 5 57); B, pirenzepine (3 3 10 mg intravenously, n 5 44); and C, ranitidine (3 3 50 mg intravenously, n 5 57). RESULTS: The pneumonia rate in ranitidine-, pirenzepine-, and placebo-treated patients is 10 of 57, 10 of 44, and 12 of 57, respectively. CONCLUSIONS: Pneumonia rate is not adversely affected by H2-receptor antagonists in stress ulcer prophylaxis. Am J Surg. 1998;176:453– 457. © 1998 by Excerpta Medica, Inc. BACKGROUND:
I
n the last 15 years a decline in the incidence of stress ulcer bleeding has been observed. One recent study reported even a virtual absence of stress-ulcer–related bleeding of patients in intensive care units (ICU) receiving prolonged mechanical ventilation without any prophylaxis.1 It is generally believed that this phenomenon is not exclusively related to the widespread use of H2-receptor antagonists in ICU, but has to be interpreted in association with multiple improvements in anesthesiological and surgical techniques. Surprisingly, the diminished rate of stress ulcer bleeding had no influence upon the overall mortality of patients in ICU.2 In this context the argument of an increasing effect of H2-receptor antagonists on the pneumonia rate, one of the leading causes of death in ICU
From the Department of Surgery (EWH, AE, FN), Division of General and Trauma Surgery (JW), Johann Wolfgang Goethe University, Frankfurt am Main, Germany. Requests for reprints should be addressed to E. W. Hanisch, MD, Department of Surgery, Johann Wolfgang Goethe University, Theodor-Stern-Kai 7, 60590 Frankfurt/Main, Germany. Manuscript submitted February 26, 1998 and accepted in revised form August 17, 1998.
© 1998 by Excerpta Medica, Inc. All rights reserved.
emerged.3,4 It was discussed that via the route of an elevated gastric pH, followed by bacterial overgrowth and subsequent tracheal aspiration, pneumonia would develop.5 This assumption, however, is subject to controversial discussions. In line with the above outlined hypothesis, total gastrectomized patients should develop a very high incidence of pneumonia, which is actually not the case. We therefore formulated the hypothesis that stress ulcer prophylaxis with H2-receptor antagonists does not lead to an increased pneumonia rate. Since only few double-blind, placebo-controlled trials do exist (for an overview see Table I) with quite differing pneumonia rates and results, patients were randomly assigned to a group with no stress ulcer prophylaxis. The trial considers also pirenzepine, which has only minor effects upon the gastric pH in comparison with H2receptor antagonists.
METHODS All patients referred to the intensive care unit of the surgical department of the Johann Wolfgang Goethe-University Frankfurt/Main were considered for the study. Exclusion criteria were patients with an active peptic ulcer disease and a concomittant ulcer medication; patients with upper gastrointestinal bleeding; patients ,18 years; transplanted patients (kidney, liver, heart); and patients with preexisting pneumonia and gastric resection. At the time of referral to the intensive care unit of the surgical department all patients were randomized to the following groups: A, placebo; B, pirenzepine (Gastrozepin, 3 3 10 mg intravenously, Thomae, Biberach, Germany); or C, ranitidine (Sostril, 3 3 50 mg intravenously, Cascan, Wiesbaden, Germany). The primary outcome measure was the pneumonia rate in patients under mechanical ventilation $48 hours. The definition of pneumonia was according to Daschner13: radiological signs of pneumonia and purulent tracheal secretion or positive microbiological findings in tracheal aspiration and temperature .38°C and leucocytosis .10,000 mm3. In all patients with radiological signs of pneumonia a bronchioalveolar lavage was routinely performed, but no quantitative cultures were done. Besides the primary study endpoint, pneumonia, clinically relevant stress bleeding was also documented. The definition of clinically relevant bleeding was bright red blood via gastric tube or melena combined with hemodynamic changes (systolic blood pressure ,100 mm Hg, tachycardia .100 beats per minute) and requirement of blood transfusion (fall in hemoglobin 0002-9610/98/$19.00 PII S0002-9610(98)00239-6
453
STUDY OF STRESS ULCER PROPHYLAXIS/HANISCH ET AL
TABLE I Placebo-controlled Studies: Effect on Pneumonia Rate n
Patients
Cheadle 1985 Reusser7 1989
200 40
Karlstadt8 1990 Apte9 1992 Metz10 1993 Martin11 1993 Ben-Menachem12 1994
87 34 167 131 300
Surgical Neurosurgical ventilation .48 hours Surgical/medical Tetanus, tracheotomized Severe head injury Surgical/medical Medical
6
Intervention
Pneumonia Rate
Cimetidine double blind Ranitidine/antacid
Placebo 1.5%, cimetidine 6.5% Placebo 38%, ranitidine 36%
Cimetidine double blind Ranitidine and/or antacid Ranitidine double blind Cimetidine double blind Cimetidine, sucralfate single blind
Placebo 6%, cimetidine 0% Placebo 50%, ranitidine 81% Placebo 19%, ranitidine 14% Placebo 9%, cimetidine 0% Placebo 6%, cimetidine 13%, sucralfate 12%
Figure. Trial profile.
.2 g/dL within 24 hours) and endoscopic identification of bleeding site and activity. Assuming a pneumonia rate of 25% in ranitidine-treated patients and a pneumonia rate of 15% in placebo-treated patients, the complete sample size was calculated to be at least 100 patients (a 5 0.05; b 5 0.20). A complete and balanced randomization schedule was generated by the Institute of Biomathematics of the University of Frankfurt. At the time of entering the ICU 454
patients were assigned to a consecutive study number, and the application of the blinded drug regimen was started. Blinding of drugs (placebo, ranitidine, pirenzepine) was achieved by dissolving them in glucose (100 mL; 5%; Braun, Melsungen, Germany) resulting in an equal infusion volume for each patient. Drugs were prepared in advance by a staff not involved in the treatment of patients according to the randomization schedule and identified by the above mentioned consecutive study number, ie, the
THE AMERICAN JOURNAL OF SURGERY® VOLUME 176 NOVEMBER 1998
STUDY OF STRESS ULCER PROPHYLAXIS/HANISCH ET AL
TABLE II General Patient Characteristics, Pneumonia, Stress Bleeding, and Mortality in Patients Mechanically Ventilated ≥48 Hours Ranitidine Patients (n) Age (years) Intensive care unit stay (days) APACHE II score Mechanical ventilation (days) Pneumonia (n)* Stress bleeding (n)† Mortality (n)
Pirenzepine
Placebo
57
44
55/22–88
53/18–86
58/22–88
57
9.7/2–95 19/2–30
9.9/2–39 21/12–34
12.6/2–58‡ 18/1–28
8.2/2–93 10 3 7
8/2–32 10 3 12
10.2/2–55§ 12 2 12
Data are presented as means and full range. * Chi square 5 3.55; DF 5 2; P 5 0.17. † Chi square 5 1.79; DF 5 2; P 5 0.41. ‡ Chi square 5 7.54; DF 5 2; P 5 0.02. § Chi square 5 8.54; DF 5 2; P 5 0.01.
code was located exclusively in this group during the trial. All blinded drugs (only identifiable by their consecutive study number) were given to the patients by the nurse staff of the ICU. Data documentation for each patient14 was performed prospectively and daily by different staff not involved in the treatment of patients and in the randomized assignment of drugs. Only one of the authors (FN) was a member of these two staff groups at different time periods. The complete trial ran between December 1, 1991, and December 30, 1992. Following several discussions of the study protocol at workshops of the “Project Group for Clinical Studies” of the German Surgical Society, the trial was approved by the Ethics Committee of the Johann Wolfgang Goethe University, Frankfurt/Main. Statistical Analysis Qualitative data were compared by the chi-square test; quantitative data were discriminated by the Kruskal-Wallis test.
RESULTS For the trial profile, see the Figure. Of 827 randomized patients, a total of 158 were mechanically ventilated $48 hours. The mean duration of the mechanical ventilation was 8.2, 8, and 10.2 days in patients receiving ranitidine, pirenzepine, and placebo, respectively. General characteristics, pneumonia, stress bleeding, and mortality of these patients are summarized in Table II. There is no statistical difference between the groups with respect to the pneumonia rate. Stress ulcer bleeding had no serious clinical sequelae in patients with ranitidine (blood transfusions, 2 units each) and pirenzepine (blood transfusion, 4 units each) except in 2 placebo patients. Both underwent cardiac surgery (aortic valve replacement; bypass surgery) with postoperative medication of coumarin, aspirin, and cortisone, respec-
tively. The first patient died owing to his ulcer bleeding, the second one received 12 units of blood. Bleeding had been controlled by endoscopic intervention. After the extension of exclusion criteria, no further bleeding was observed in placebo-treated patients. (The directly related death of a patient due to ulcer bleeding led the monitor to break the code. Since the patient belonged to the placebo group, exclusion criteria were extended by postoperative drugs like aspirin, cortisone, and coumarin.) Of 741 notrandomized patients, 22 developed pneumonia and 3 demonstrated clinical relevant bleeding. Table III summarizes the data of clinical relevant stress ulcer bleeding during the complete trial.
COMMENTS Prophylaxis has unequivocally decreased the formerly frightening stress ulcer bleeding of patients in ICU, although the general mortality of this population could not be reduced. This obvious discrepancy stimulated numerous investigations, and the knowledge of nosocomial infection as critical determinant in this context was established.15 According to this hypothesis, the pH-elevating effect of H2-receptor antagonists is prompting gastric bacterial growth, which in turn causes more pneumonias via an esophageal ascension and subsequent tracheal aspiration. Thus, giving sucralfate, which does not alter gastric pH, yields significantly less pneumonias.16 In this prospective randomized study, late pneumonias occurred to a significantly smaller extent in the sucralfate group than in the ranitidine group (5% versus 21%). In a recent study that compared the results of sucralfate versus maximal H2blocker infusion therapy plus antacids, the overall pneumonia rate was 27.5% in the H2-blocker antacid group and 29.8% in the sucralfate group, with a significant level of P 5 0.48. 17 This study implies a considerable reduction of charges/cost for the sucralfate approach. Contrary to the above-mentioned studies, in another randomized trial with trauma patients, no difference could be found.18 Our data, placebo-controlled, are corroborating these results (Table II). Furthermore, there are also placebo-controlled studies in long-term ventilated patients supporting the results of our study. Thus, Reusser et al7 randomized 40 neurosurgical patients requiring mechanical ventilation for .48 hours to either a control group receiving no stress ulcer prophylaxis or to a ranitidine/antacid group. There was no significant difference in the pneumonia rate of patients with and without stress ulcer prophylaxis. In line with these findings, Metz et al10 also could not find any difference in 167 placebo-controlled, randomized patients with severe head injury. On the contrary, Apte et al9 demonstrated for 34 randomized tracheotomized patients with tetanus, receiving intravenous ranitidine (50 mg, 6 hourly and/or antacids) or neither ranitidine nor antacids, significantly more pneumonias in the ranitidine group (81%) than in the control group (50%) were found. Moreover, patients treated with ranitidine developed pneumonia significantly earlier than patients in control groups. With respect to pirenzepine only a few trials are available, the largest one having been conducted by Tryba et al.19 In that study, 400 patients were randomized to a ranitidine (200 mg) or a pirenzepine (50 mg) group. Eight out of 28
THE AMERICAN JOURNAL OF SURGERY® VOLUME 176 NOVEMBER 1998
455
456
THE AMERICAN JOURNAL OF SURGERY® VOLUME 176 NOVEMBER 1998
Survivor
Survivor
Death at 31st day postoperative
Forrest Ia (duodenal ulcer) Death at 12th day postoperative Erosive gastritis Erosive gastritis
Survivor Death at 43rd day postoperative Mortality
Forrest IIb (angulus)
Forrest IIb (corpus) Survivor
Erosive gastritis
4 13 1 12 18 10 29 21
Pirenzepine Pirenzepine
13 24
Erosive gastritis
TABLE III
Ranitidine
REFERENCES 1. Zandstra DF, Stoutenbeek CP. The virtual absence of stressulceration related bleeding in ICU patients receiving prolonged mechanical ventilation without any prophylaxis. A prospective cohort study. Intens Care Med. 1994;20:335–340. 2. Schuster DP. Stress ulcer prophylaxis: in whom? With what? Crit Care Med. 1993;21:4 – 6. 3. Tryba M, Zevounou F, Wruck G. Streßblutungen und postoperative Pneumonien bei Intensivpatienten unter Ranitidin oder Pirenzepin. Dtsch Med Wschr. 1988;113:930 –936. 4. Driks MR, Craven DE, Celli BR, et al. Nosocomial pneumonia in intubated patients given sucralfate as compared with antacids or histamine type 2 blockers. NEJM. 1987;317:1376 –1382. 5. Cook DJ, Laine LA, Guyatt GH, Raffin ThA. Nosocomial pneumonia and the role of gastric pH. A meta-analysis. Chest. 1991;11:7–13. 6. Cheadle WG, Vitale GC, Mackie CR, Cuschieri A. Prophylactic postoperative nasogastric decompression. Ann Surg. 1985;202: 361–366. 7. Reusser P, Zimmerli W, Scheidegger D, et al. Role of gastric
Ranitidine Ranitidine
Synopsis of Clinically Relevant Bleeding in 8 Patients
Pirenzepine
patients on mechanical ventilation receiving ranitidine developed pneumonias (28.6%) versus 3 out of 33 patients on mechanical ventilation under pirenzepine (9.1%), which is statistically significant at a level of P ,0.05. One reason for the striking difference from our results might be the different ranitidine and pirenzepine dosages used. Finally, the comparable pneumonia rates in our pirenzepine and control group patients are underlining the knowledge that H2-receptor antagonists with the dosage used do not increase the pneumonia risk of long-term ventilated patients in critical condition. In addition, confounding parameters like nasogastric tubes and tracheostomas are obviously of no statistical relevance in this context (Table IV). Moreover, Cook found no significant differences in the rates of ventilator-assisted pneumonia between ranitidine- and sucralfate-treated patients in a multicenter, randomized trial.20 In conclusion, we can demonstrate that in surgical critically ill patients on mechanical ventilation for $48 hours, stress ulcer prophylaxis with H2-receptor antagonists pneumonia rate is not adversely affected. Since the incidence of stress gastritis in our study is only 3.5% for patients on mechanical ventilation $48 hours in the placebo group, one might generally question prophylaxis. Of course, we are not allowed to draw any definite conclusion about this point, because the number of patients is too low for statistical analysis with respect to the outcome measure “clinically relevant bleeding.” Conceivably, improvements in the management of critically ill patients per se might have changed the likelihood of developing stress gastritis over the last decade.
12 21
35 13
Mechanical Ventilation (d) Days of bleeding postoperative Endoscopic findings
26 7
Aortic valve disease
20 8
Polytrauma
55
Head trauma
42
Coronary heart disease
53
Polytrauma Carcinoma of the floor of the mouth, liver cirrhosis 12 2 10 9
Placebo
Ileus
Pirenzepine
Diagnosis
Ranitidine
Placebo
Nasogastric tube .48 hours (n) Nasogastric tube .4 days (n) Tracheostoma (n)
Placebo
Nasogastric Tube Duration and Tracheostoma in Patients Mechanically Ventilated ≥48 Hours
Coronary heart disease
TABLE IV
Forrest Ib (duodenal ulcer) Survivor
STUDY OF STRESS ULCER PROPHYLAXIS/HANISCH ET AL
STUDY OF STRESS ULCER PROPHYLAXIS/HANISCH ET AL
colonization in nosocomial infections and endotoxemia: a prospective study in neurosurgical patients. J Infect Dis. 1989;160:414 – 421. 8. Karlstadt RG, Iberti TJ, Silverstein J, et al. Comparison of cimetidine and placebo for the prophylaxis of upper gastrointestinal bleeding due to stress-related gastric mucosal damage in the intensive care unit. J Intens Care Med. 1990;5:26 –32. 9. Apte NM, Karnad DR, Medhekar TP, et al. Gastric colonization and pneumonia in intubated critical ill patients receiving stress ulcer prophylaxis: a randomized, controlled trial. Crit Care Med. 1992;20:590 –593. 10. Metz CA, Livingston DH, Smith JS, et al. Impact of multiple risk factors and ranitidine prophylaxis on the development of stress-related upper gastrointestinal bleeding: a prospective, multicenter, double-blind, randomized trial. Crit Care Med. 1993;21: 1844 –1849. 11. Martin LF, Booth F, Karlstadt RG, et al. Continuous intravenous cimetidine decreases stress-related upper gastrointestinal hemorrhage without promoting pneumonia. Crit Care Med. 1993;21: 19 –30. 12. Ben-Menachem T, Fogel R, Patel RV, et al. Prophylaxis for stress-related gastric hemorrhage in the medical intensive care unit. Ann Intern Med. 1994;121:568 –575.
13. Daschner F, Reuschenbach K, Pfisterer I, et al. Der Einfluß von Streßulkusprophylaxe auf die Ha¨ufigkeit einer Beatmungspneumonie. Anaesthesist. 1987;36:9 –18. 14. Windolf J, Gottschalk S, Inglis R, Pannike A. Einsatz eines elektronischen Datenerfassungssystems in der chirurgischen Intensivmedizin. Unfallchirurgie. 1990;16:252–257. 15. Tryba M. Side effects of stress bleeding prophylaxis. Am J Med. 1989;86(suppl 6A):85–93. 16. Prod’hom G, Leuenberger P, Koerfer J, et al. Nosocomial pneumonia in mechanically ventilated patients receiving antacid. Ann Intern Med. 1994;120:653– 662. 17. Maier RV, Mitchell D, Gentilello L. Optimal therapy for stress gastritis. Ann Surg. 1994;220:353–363. 18. Pickworth KK, Falcone RE, Hoogeboom JE, Santanello SA. Occurrence of nosocomial pneumonia in mechanically ventilated trauma patients. Crit Care Med. 1993;12:1856 –1862. 19. Tryba M. Prevention of stress bleeding with ranitidine or pirenzepine and the risk of pneumonia. J Clin Anaesth. 1988;1:12– 20. 20. 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. NEJM. 1998;338: 791–797.
THE AMERICAN JOURNAL OF SURGERY® VOLUME 176 NOVEMBER 1998
457