Reply to Yap et al.

We appreciate the opportunity to express our comments in relationship to the letter from Dr. Yap and colleagues. In their retrospective audit, the combination of epinephrine plus argon plasma coagulation (APC) followed by the administration of intravenous pantoprazole for 48 h resulted in a lower rebleeding rate and a decreased length of hospital stay in comparison with a historical group of patients managed with epinephrine injection alone plus pantoprazole. No significant effects were seen with regard to transfusion, surgery, and case fatality rates.

It is proper to point out that the results of case-control studies are usually open to criticism due to the various aspects of bias and confounding implicit in their design. Excellent resources elsewhere address the strengths and weaknesses of this particular type of study design and the appropriate methods of analyzing data obtained with it [1]. We have no intention of belaboring this point, but several issues in the report by Yap et al. need to be addressed.

Selection bias is difficult to completely eliminate from their results, given the lack of information regarding the exact mechanism for selecting the control patients.

The authors make reference to the administration of Helicobacter pylori treatment in some patients, and they state that there were no statistical differences between the two groups with regard to ulcer size or location; but no information is provided regarding the actual etiology of the bleeding lesions (e. g., Helicobacter pylori infection, nonsteroidal anti-inflammatory drugs, anastomotic ulcers).

Similarly, the distribution of some important prognostic factors within the two groups, which are known to be associated with independent prediction of the outcome after endoscopic therapy, is not reported. In addition to ulcer size and active bleeding, some of these clinical parameters also include hemodynamic status and hemoglobin level on presentation [2] [3] [4] [5] [6] [7]. Independent predictors of outcome can strongly affect the overall results if not accounted for using multivariate regression analysis or some other appropriate statistical technique [1].

With regard to the administration of endoscopic treatment, there are no details on the volume of epinephrine injections in the two groups. Available data suggest that injecting a large volume of epinephrine can reduce the rate of rebleeding in comparison with lower volumes of epinephrine in patients with high-risk peptic ulcers [8]. Similarly, there are no clear data regarding the exact dosage and route of pantoprazole administration in the control patients. Differences between the two groups with regard to the total dose and bioavailability of pantoprazole could potentially have influenced the results.

Finally, the results of their statistical analysis are in contradiction with our analysis of the presented data. Using the StatsDirect statistical program, version 2.2.8 (StatsDirect Ltd., Cheshire, United Kingdom) for calculations, the risk difference in rebleeding rates between controls (27 %) and cases (12 %) is 15 %; 95 % confidence interval (CI) - 0.05 % to 35 % (uncorrected chi-squared test 2.4, P = 0.1217; Yates-corrected chi-squared test 1.532, P = 0.215; P = 0.1378 by Fisher’s exact test). Using the odds ratio as a measure of effect for this case-control study yields a point estimate of 0.37 (Fisher’s exact test 95 % CI, 0.08 to 1.55; P = 0.215) for the experimental group in comparison with the control group. Thus, the risk difference in event rates does not significantly differ between the two groups (the number needed to treat for epinephrine injection plus argon plasma coagulation treatment is seven; the number needed to harm is 22.5; the number needed to benefit is three [9]).

The issue of whether combined endoscopic therapy (injection plus thermocoagulation) is superior to injection endoscopic therapy alone is of extreme clinical relevance. Although it remains to be confirmed in large-scale randomized studies, analysis of full published trials suggests that combination treatment is probably superior to injection treatment alone [10]. The role and impact of APC on the outcome of peptic ulcer bleeding in comparison with other contact-coagulation hemostatic techniques are largely undefined and remain contentious at present [11].

References

• 1 Rothman K J, Greenland S. Case-control studies. In: Rothman KJ, Greenland S, (eds). Modern epidemiology, 2nd ed.  Philadelphia; Lippincott-Raven 1998: 93-114
• 2 Choudari C P, Rajgopal C, Elton R A, Palmer K R. Failures of endoscopic therapy for bleeding peptic ulcer: an analysis of risk factors.  Am J Gastroenterol. 1994;  89 1968-1972
• 3 Brullet E, Calvet X, Campo R. et al . Factors predicting failure of endoscopic injection therapy in bleeding duodenal ulcer.  Gastrointest Endosc. 1996;  43 111-116
• 4 Brullet E, Campo R, Calvet X. et al . Factors related to the failure of endoscopic injection therapy for bleeding gastric ulcer.  Gut. 1996;  39 155-158
• 5 Thomopoulos K C, Mitropoulos J A, Katsakoulis E C. et al . Factors associated with failure of endoscopic injection haemostasis in bleeding peptic ulcers.  Scand J Gastroenterol. 2001;  36 664-668
• 6 Villanueva C, Balanzo J, Espinos J C. et al . Prediction of therapeutic failure in patients with bleeding peptic ulcer treated with endoscopic injection.  Dig Dis Sci. 1993;  38 2062-2070
• 7 Wong S K, Yu L M, Lau J Y. et al . Prediction of therapeutic failure after adrenaline injection plus heater probe treatment in patients with bleeding peptic ulcer.  Gut. 2002;  50 322-325
• 8 Lin H J, Hsieh Y H, Tseng G Y. et al . A prospective, randomized trial of large- versus small-volume endoscopic injection of epinephrine for peptic ulcer bleeding.  Gastrointest Endosc. 2002;  55 615-619
• 9 Altman D G. Confidence intervals for the number needed to treat.  BMJ Br Med J. 1998;  317 1309-1312
• 10 Rollhauser C, Fleischer D E. Current status of endoscopic therapy for ulcer bleeding.  Baillière’s Best Pract Res Clin Gastroenterol. 2000;  14 391-410
• 11 Laine L. Endoscopic therapy for bleeding ulcers: room for improvement?.  Gastrointest Endosc. 2003;  57 557-560