Dig Dis Sci (2007) 52:1685–1690

DOI 10.1007/s10620-006-9684-1

ORIGINAL ARTICLE

Intravenous Versus High-Dose Oral Proton Pump Inhibitor

Therapy After Endoscopic Hemostasis of High-Risk Lesions in
Patients with Acute Nonvariceal Upper Gastrointestinal Bleeding
Sanjay Murthy · Leila Keyvani · Shauna Leeson ·
Laura E Targownik

Received: 12 August 2006 / Accepted: 16 November 2006 / Published online: 24 March 2007


C Springer Science+Business Media, LLC 2007

Abstract Intravenous proton pump inhibitors (IV PPIs) de- Introduction

crease rebleeding following endoscopic hemostasis of bleed-
ing peptic ulcers. Oral PPIs may be equally efficacious More than 40% of acute nonvariceal upper gastrointestinal
and may significantly reduce health care costs. This study bleeding (ANVUGIB) cases in North America may be at-
aimed to compare outcomes in patients receiving oral ver- tributed to bleeding peptic ulcers [1]. Such patients who
sus IV PPI therapy following endoscopic hemostasis in pa- also have evidence of high-risk ulcer stigmata on upper en-
tients with acute nonvariceal upper gastrointestinal bleeding doscopy, including active bleeding, a nonbleeding visible
(ANVUGIB). We performed a retrospective review of all vessel (NBVV), or an adherent clot, carry anywhere from
patients who received PPI therapy following endoscopic 10% to 100% risk of rebleeding, which is strongly associ-
hemostasis for ANVUGIB. The primary outcome was the ated with the need for surgical management and subsequent
adverse gastrointestinal event rate. One hundred sixty-two mortality [2–4]. While performance of endoscopic hemosta-
patients met the entry criteria (72 oral PPIs, 90 IV PPIs). sis is associated with a significant decrease in the rate of
The difference in the rate of adverse gastrointestinal events recurrent bleeding events in this setting, these patients still
between the two groups was 1% (P = 0.85). Postendo- carry a 10% to 25% risk of early rebleeding [3, 5].
scopic IV PPI use was associated with an odds ratio of 1.01 The presence of gastric acid is a necessary cofactor in the
for developing an adverse outcome versus oral PPIs (95% CI: pathogenesis of recurrent bleeding in this setting. In vitro
0.44–2.33). We conclude that oral PPIs are probably equiv- studies have demonstrated that activation of both extrinsic
alent to IV PPIs for preventing rebleeding in ANVUGIB and intrinsic coagulation cascades, as well as platelet aggre-
patients. gation, is significantly impaired when the pH is < 6, thereby
potentially leading to poor clot stabilization [6]. Clinical
Keywords Acute nonvariceal upper gastrointestinal studies have also confirmed that a mean intragastric pH < 6
bleeding . Proton pump inhibitor . Endoscopic hemostasis . is predictive of rebleeding within 72 hr of endoscopic inter-
Recurrent bleeding . Oral therapy vention for ANVUGIB [7].
Not surprisingly, proton pump inhibitors (PPIs) are very
effective in further decreasing rebleeding rates in patients
S. Murthy
Department of Internal Medicine, University of Manitoba, who have undergone endoscopic hemostasis of bleeding pep-
Winnipeg, Manitoba, Canada tic ulcers. In a landmark study, subjects undergoing endo-
scopic hemostasis for acute upper gastrointestinal bleeding
L. Keyvani · S. Leeson · L. E. Targownik secondary to high-risk peptic ulcer disease who were given
Section of Gastroenterology, University of Manitoba,
Manitoba, Canada continuous intravenous omeprazole for 72 hr had a signifi-
cantly decreased risk of recurrent bleeding events compared
L. E. Targownik (
) to similar patients who received no active medical therapy
804E-715 McDermot Avenue, Winnipeg, [5]. This finding was recently confirmed in a randomized
Manitoba, Canada R3E 3P4
e-mail: [email protected] control trial comparing 72-hr continuous intravenous (IV)
pantoprazole with placebo in a similar clinical setting [8].

Springer
1686 Dig Dis Sci (2007) 52:1685–1690

However, the use of IV PPIs adds significantly to the cost heart rate); results of relevant laboratory testing (hemoglobin
of patient care in hospital. Moreover, these agents are often level, platelet count, INR); time to performance of upper
used even when they are not indicated, further adding to hos- endoscopy; endoscopic findings; endoscopic interventions;
pital costs [9]. Recent studies have shown that high-dose oral pre- and postendoscopic antacid therapy (including time to
PPI therapy may also be effective in decreasing rebleeding initiation of antacid therapy postendoscopy); recurrent he-
rates in patients with acute gastrointestinal bleeding due to matemesis, melena, or hematochezia postendoscopy; num-
high-risk peptic ulcer disease, and the magnitude of bene- ber and timing of blood transfusions postendoscopy; need
fit appears similar to what has been demonstrated with IV for repeat endoscopic hemostasis; need for surgery to con-
PPIs [10, 11]. Given the significant cost savings over their trol recurrent bleeding; occurrence of in-hospital death; and
IV counterparts, oral PPIs would be an attractive choice of need for readmission to hospital within 30 days of discharge
therapy in this situation provided that they have a similar due to recurrent bleeding.
efficacy to IV PPIs. However, no studies have directly com-
pared oral and IV PPI therapy in this setting. Therefore, we Comparator groups and outcomes
sought to compare outcomes between these two therapeutic
modalities in subjects presenting to hospital with acute upper Patients were separated based on whether they received high-
gastrointestinal bleeding who required endoscopic hemosta- dose oral or IV PPIs within 8 hr of endoscopic hemostasis.
sis for control of active bleeding or prevention of recurrent For the IV PPI group, patients receiving either continuous
bleeding. infusion or intermittent bolus IV PPIs were considered eli-
gible. High-dose oral PPI was defined as receipt of at least
40 mg of omeprazole per day or an equipotent dose of an
Methods alternative PPI. Patients who received both IV and high-dose
oral treatment within the first 8 hr following performance of
Subject identification and study criteria endoscopy were included in the IV PPI group.
The primary outcome of the study was the difference
The medical records of all consecutive patients presenting in the combined adverse event rates in patients receiving
to one of two tertiary care hospitals in Winnipeg, Canada, postendoscopic oral versus IV PPI therapy. Adverse gas-
between 1999 and 2004 with a primary discharge diagno- trointestinal events included in-hospital rebleeding, surgery
sis consistent with ANVUGIB were reviewed. In order to for control of rebleeding, in-hospital death from any cause,
maximize the likelihood of identifying potentially eligible and readmission to hospital within 30 days of discharge for
subjects, we used a list of ICD-9 codes that has been vali- ANVUGIB. We defined rebleeding as hematemesis follow-
dated and used extensively for other studies of ANVUGIB ing initial endoscopy, documented melena or hematochezia
[12]. following endoscopy if associated with a drop in hemoglobin
Included subjects had to present to hospital with evidence of more than 20 g/L from the first postendoscopy hemoglobin
of ANVUGIB, defined as hematemesis, melena, or hema- measurement, recurrent melena after documentation of non-
tochezia, with identification of a suitable nonvariceal upper melenic stool, or the presence of active bleeding or an ulcer
gastrointestinal lesion on endoscopy (and with no definite with high-risk stigmata on a repeat upper endoscopy for any
bleeding source identified on colonoscopy or sigmoidoscopy reason within the same hospital admission. Other outcomes
in the case of hematochezia) and had to have undergone en- of interest included the rate each of the aforementioned ad-
doscopic hemostasis within 24 hr of presentation (injection verse events individually, the incidence of high-risk stigmata
of epinephrine, electrocautery, or use of an endoclip) fol- at the time of endoscopy, and the overall length of hospital
lowed by either intravenous or high-dose oral PPI therapy stay of study patients, as well as the average number of blood
within 8 hr of performance of endoscopy. We excluded pa- transfusions overall and after 24 hr postendoscopy.
tients whose bleeding developed while in hospital for another
indication, as well as those transferred from other facilities Data integrity and statistics
more than 6 hr after their initial presentation.
One of the investigators (L.E.T.) manually checked the com-
Data collection pleted data abstraction forms to assess the provided infor-
mation for inconsistencies. An automated check was then
Three research assistants (S.M., L.K., S.N.) independently performed after data entry had been completed for each sub-
reviewed all of the charts for relevant data. The following ject file. Missing or inconsistent data were resolved through
information was collected for each patient: demographic a review of the patients’ medical records. One-sided Stu-
data; relevant comorbidities; relevant medications on presen- dent t-tests and Fisher’s exact test were used to compare
tation; entrance hemodynamic parameters (blood pressure, means and proportions of appropriate data. Nonnormally

Springer
Dig Dis Sci (2007) 52:1685–1690 1687

distributed data were analyzed with the Wilcoxon rank-sum The baseline characteristics of the 162 enrolled subjects
test. A one-sided P value of < 0.05 was considered statisti- are presented in Table 1. Overall, 72 patients received at least
cally significant. We chose to perform one-sided testing as one dose of intravenous PPIs following performance of en-
there is no physiologic reason to anticipate that IV adminis- doscopy, whereas 90 subjects received high-dose oral PPIs
tration of PPIs would be therapeutically inferior to oral PPI only. Of the patients in the IV PPI group, 91% received an
dosing. 80-mg bolus of pantoprazole, followed by an 8 mg/hr pan-
Multiple logistic regression analysis was performed toprazole infusion. The median duration of IV PPI therapy
using NCSS 2004 to account for the effect of potentially following endoscopy was 61.4 hr; 77% of subjects received
confounding variables, including age, gender, receipt of IV PPIs for at least 24 hr following performance of endo-
pre-endoscopic PPI, clinical Rockall score ≥ 5, presenting scopic hemostasis, while 38% received an IV PPI infusion
systolic blood pressure < 100 mm Hg, and presence of for > 72 hr. Subjects who received only oral PPIs following
arterial bleeding or nonbleeding visible vessel on endoscopy. endoscopy were more often female and were more likely to
We estimated that the risk of adverse events would be 15% have received pre-endoscopic PPI therapy, while those who
in the IV PPI group and 25% in the oral PPI group. Based received postendoscopic IV PPI therapy were significantly
on this, we calculated that 394 subjects would be required more likely to have a high-risk lesion visualized at endoscopy
to detect a statistically significant one-sided difference and had a shorter mean time to endoscopy (although the pro-
between the two groups with 95% confidence at a power of portion of patients undergoing endoscopy within 6 hr was
0.8, assuming equal sizes of the comparator groups. not statistically different between the two groups). Nonethe-
less, the mean postendoscopic Rockall scores were similar
between the two groups, suggesting that both groups were
Results at an equivalent baseline risk of rebleeding and mortality.
All other observed characteristics were statistically similar
One-hundred sixty-two of 552 patients who underwent en- between the two groups.
doscopy for ANVUGIB during the 5-year study period met Comparison of adverse outcome rates in patients receiv-
the inclusion criteria for the study. Of the 390 excluded pa- ing IV versus oral PPIs after endoscopic hemostasis is pre-
tients, 351 did not undergo endoscopic hemostasis and an- sented in Fig. 1. Both groups had a statistically similar risk
other 39 did not receive PPIs within 8 hr postendoscopy. of developing an adverse outcome during their hospital stay
The statistical power to detect a hypothesized 10% lower (21% in the IV PPI group vs. 22% in the oral PPI group;
incidence in overall adverse events in the IV PPI group with P > 0.2). Subjects receiving postendoscopic IV PPIs were
95% confidence was 48.2%, given the sample sizes of our significantly more likely to require an operative interven-
comparator groups. tion for control of bleeding than those receiving oral PPIs

Table 1 Baseline
characteristics of enrolled Postendoscopic Postendoscopic oral
subjects IV PPI (n = 72) PPI (n = 90) P value

Mean age (yr) 62.3 65.8 > 0.2

% female 23.6 46.7 0.002
Mean SBP (mm Hg) 123 126 > 0.2
Mean DBP (mm Hg) 72 69 > 0.2
Mean HR (bpm) 93 95 > 0.2
% with SBP ≤ 100 mm Hg 13 16 > 0.2
% with HR > 100 33 41 > 0.2
Hgb on presentation (g/L) 91 89 > 0.2
Urea/creatinine ratio 158 163 > 0.2
Postendoscopic Rockall score
Mean ± SD 5.3 ± 1.8 5.1 ± 1.8 > 0.2
% with score ≥ 5 63 61 > 0.2
% gastric or duodenal ulcers/erosions 81 70 0.15
% high-risk endoscopic stigmata 94 86 > 0.2
% with active bleeding/NBVV 67 48 0.025
% receiving pre-endoscopic PPI 35 54 0.017
% receiving pre-endoscopic IV PPIs 32 2 < 0.001
Median time to endoscopy (hr) IQR 10.5 (4.5–14.7) 18.8 (4.9–20.5) 0.045
% undergoing endoscopy at ≤ 6 hr 38 32 > 0.2

Springer
1688 Dig Dis Sci (2007) 52:1685–1690

Fig. 1 Risks of adverse

gastrointestinal outcomes
between comparator groups

(11% vs. 2%; P = 0.023), but when subjects who underwent been equivalent between the two groups, based on the sim-
surgery within 24 hr of endoscopy were excluded, the dif- ilarity in their mean Rockall scores. Interestingly, the IV
ference in surgical rates was no longer significantly different PPI group actually showed an increased risk of requiring
(6% vs. 2%; P > 0.2). There was no significant difference postendoscopic surgery to control bleeding; however, this
between the groups in the need for late blood transfusions difference was not significant beyond 24 hr, suggesting that
(31% vs. 25%; P > 0.2) or in the median time to hospital more patients in the IV PPI group may have simply required
discharge (5 days vs. 4 days; P > 0.2). surgery to control bleeding that could not be controlled at
After adjusting for multiple potential confounders, the use endoscopy, likely even before the clot-stabilizing benefit of
of postendoscopic IV PPIs was associated with a nonsignif- PPIs had a chance to take effect.
icant odds ratio of 1.01 for developing an adverse outcome Although the current standard of care in North America
compared to oral PPIs (95% CI: 0.44–2.33). A subgroup is to administer IV PPI therapy for a period of at least 72 hr
analyses of 91 patients with either active arterial bleeding following endoscopic intervention for ANVUGIB with high-
or a nonbleeding visible vessel on the initial endoscopy also risk stigmata, our findings suggest that high-dose oral PPI
showed comparable adverse outcome rates between the com- therapy is equally effective in minimizing the rate of adverse
parator groups (21% IV PPI vs. 25% oral PPI). outcomes in this setting. The use of oral PPIs in this situation
would lead to a significant decrease in both medication and
hospital costs, the latter related to the potential earlier dis-
Discussion charge from hospital of less sick patients who are otherwise
kept in hospital for a minimum of 72 hr for continuous IV
To our knowledge, this is the first study directly comparing PPI therapy.
oral and IV PPIs in patients who have undergone endoscopic It is not surprising to find that oral PPIs are comparable
hemostasis for high-risk upper gastrointestinal bleeding le- to IV PPIs in terms of overall efficacy in preventing rebleed-
sions. We determined that patients receiving high-dose oral ing after endoscopic intervention for ANVUGIB. Oral PPIs
PPIs following the performance of endoscopy were no more have been shown to reach maximal plasma concentration
likely to develop recurrent adverse gastrointestinal events 2–3 hr after ingestion [13] and studies have demonstrated
than patients who received IV PPIs. The noninferiority of that the bioavailability of oral PPIs is high [14]. Also, a
oral PPIs remained even after adjusting for known potential single dose of IV administered PPI on intragastric pH is
confounders. Although the IV PPI group had a significantly not significantly different from the same dose administered
higher proportion of patients with high-risk ulcer lesions orally [15]. Furthermore, the presence of blood in the upper
(active bleeding/NBVV), this did not appear to affect the gastrointestinal tract promotes activation of proton pumps,
equivalence observed between these two treatment groups which is necessary for their subsequent irreversible deacti-
following regression analysis. On the whole, the baseline vation by a PPI. The subsequent continuous infusion of IV
risk of rebleeding (i.e., prior to PPI therapy) should have PPIs would therefore only be of benefit for newly generated

Springer
Dig Dis Sci (2007) 52:1685–1690 1689

proton pumps. As the mean half-life recovery time of proton In an effort to maximize the power of our study, we did
pumps in healthy volunteers following a single dose of PPI is not impose strict exclusion criteria on various factors, in-
anywhere between 25 and 50 hr [16, 17], twice daily dosing cluding the type of lesion observed on endoscopy (i.e., pep-
or oral PPIs may be sufficient to inhibit any newly generated tic ulcer, AVM, gastric cancer, etc.), presence or absence of
proton pumps. documentation of the high-risk stigmata observed, method
The ability of oral PPIs to suppress acid secretion to a sim- of endoscopic hemostasis, or time to development of an ad-
ilar extent as their bolus IV counterparts has been shown in verse event postendoscopy. All patients who underwent en-
several clinical studies [15, 18]. More recently, Metz and col- doscopic hemostasis were included in the study under the
leagues showed a similar acid suppression effect (in terms assumption that the endoscopist believed the patient was
of percentage of time that intragastric pH is > 6) of oral at high risk of rebleeding, irrespective of whether the ex-
lansoprazole given every 6 hr and continuous infusion of act nature of the high-risk lesion documented. Furthermore,
lansoprazole following a bolus dose [19]. This is in contrast although most studies evaluating postendoscopic PPI ther-
to earlier studies showing superior 24-hr acid suppression apy in the present setting have only included patients with
with infusional PPI therapy over either oral or bolus injec- peptic ulcer disease, we may reasonably extrapolate the ef-
tion therapy [20, 21], again using the outcome criterion of fect of PPI therapy on clot stabilization would not differ for
percentage of time that intragastric pH is > 6. non-ulcer-based upper gastrointestinal lesions. Despite this
The notion that maintenance of intragastric pH > 6 results effort, the power of our study was still quite low, and it is
in lower rebleeding rates is based on early in vitro studies by therefore quite possible that we simply failed to capture a
Green and colleagues [6] showing that clot stabilization and true significant difference in outcomes between these two
platelet aggregation are impaired at a pH of < 6. However, therapeutic modalities.
a study conducted by Tseng and colleagues [22] evaluating In summary, the results of this study suggest that oral PPI
this relationship in patients who undergo endoscopic inter- therapy may be an effective alternative to IV PPI therapy
vention for acute peptic ulcer bleeds with high-risk stigmata in patients who present to hospital with ANVUGIB and un-
failed to show any significant effect of the percentage of time dergo endoscopic hemostasis of a high-risk lesion. Future
that intragastric pH was > 6 on the development of recurrent well-designed, prospective, randomized trials are required
bleeding. It may be that the effects of acid suppression on clot to confirm our findings as a step toward implementing cost-
stabilization in vivo differ from what occurs in vitro or that effective use of PPI therapy in patient care.
some other pH parameter is required to better characterize
the relationship between acid suppression and adverse clini- Acknowledgments Dr. Targownik is supported by the Rudy Falk
cal outcomes in this setting. While this relationship needs to Clinician Scientist Award. This study was partially funded through an
unrestricted grant provided by Altana Pharmaceuticals Canada, Inc.
be studied in more detail in the clinical setting, our findings
suggest that the net effect of acid suppressive therapy on ad-
verse event rates does not differ between oral and continuous References
IV PPI therapy.
There are a number of limitations to our study. First, its 1. Boonpongmanee S, Fleischer DE, Pezzullo JC, Collier K, Mayoral
retrospective design precluded a standardized approach to W, Al-Kawas F, Chutkan R, Lewis JH, Tio TL, Benjamin SB (2004)
patient care and recording of observations. For example, as The frequency of peptic ulcer as a cause of upper-GI bleeding is
exaggerated. Gastrointest Endosc 59(7):788–794
both endoscopy and patient management would have been 2. Savides TJ, Jensen DM (2000) Therapeutic endoscopy for non-
undertaken by different gastroenterologists and general sur- variceal gastrointestinal bleeding. Gastroenterol Clin North Am
geons over a number of years, there may have been con- 29(2):465–487
siderable interobserver variability with respect to reported 3. Adler DG, Leighton JA, Davila RE, Hirota WK, Jacobson BC,
Qureshi WA, Rajan E, Zuckerman MJ, Fanelli RD, Hambrick RD,
endoscopic findings and the decision to pursue hemostasis. Baron T, Faigel DO, ASGE (2004) ASGE guideline: the role of en-
Furthermore, there may have been a channeling bias regard- doscopy in acute non-variceal upper-GI hemorrhage. Gastrointest
ing the administration of IV or oral PPIs, if IV PPIs were Endosc 60(4):497–504
believed to be more effective and thereby administered more 4. Bleau BL, Gostout CJ, Sherman KE, Shaw MJ, Harford WV, Keate
RF, Bracy WP, Fleischer DE (2002) Recurrent bleeding from peptic
often to patients with higher-risk lesions or poorer prognoses. ulcer associated with adherent clot: a randomized study comparing
In order to minimize the potential effects of such biases on endoscopic treatment with medical therapy. Gastrointest Endosc
our results, we performed a multivariate logistic regression 56(1):1–6
to adjust for obvious confounders and still determined that 5. Lau JY, Sung JJ, Lee KK, Yung MY, Wong SK, Wu JC, Chan
FK, Ng EK, You JH, Lee CW, Chan AC, Chung SC (2000) Effect
use of IV PPIs was not predictive of an adverse event free of intravenous omeprazole on recurrent bleeding after endoscopic
outcome. However, we accept that there may be unmeasured treatment of bleeding peptic ulcers. N Engl J Med 343(5):310–316
confounders that may not be easily accounted for. 6. Green FW Jr., Kaplan MM, Curtis LE, Levine PH (1978) Effect
of acid and pepsin on blood coagulation and platelet aggregation.

Springer
1690 Dig Dis Sci (2007) 52:1685–1690

A possible contributor prolonged gastroduodenal mucosal hemor- tion to healthy male subjects. Eur J Clin Pharmacol 44(6):575–
rhage. Gastroenterology 74(1):38–43 578
7. Hsieh YH, Lin HJ, Tseng GY, Perng CL, Wang K, Lo WC, Chang 15. Hartmann M, Ehrlich A, Fuder H, Luhmann R, Emeklibas S,
FY, Lee SD (2004) Poor responders to intravenous omeprazole Timmer W, Wurst W, Lucker PW (1998) Equipotent inhibition
in patients with peptic ulcer bleeding. Hepatogastroenterology of gastric acid secretion by equal doses of oral or intravenous
51(55):316–319 pantoprazole. Aliment Pharmacol Ther 12(10):1027–1032
8. Zargar SA, Javid G, Khan BA, Yattoo GN, Shah AH, Gulzar GM, 16. Shin JM, Homerin M, Domagala F, Ficheux H, Sachs G (2006)
Sodhi JS, Mujeeb SA, Khan MA, Shah NA, Shafi HM (2006) Pan- Characterization of the inhibitory activity of tenatoprazole on the
toprazole infusion as adjuvant therapy to endoscopic treatment in gastric H + ,K + -ATPase in vitro and in vivo. Biochem Pharmacol
patients with peptic ulcer bleeding: prospective randomized con- 71(6):837–849
trolled trial. J Gastroenterol Hepatol 21(4):716–721 17. Metz DC, Ferron GM, Paul J, Turner MB, Soffer E, Pisegna JR,
9. Enns R, Andrews CN, Fishman M, Hahn M, Atkinson K, Kwan Bochenek WJ (2002) Proton pump activation in stimulated parietal
P, Levy A (2004) Description of prescribing practices in patients cells is regulated by gastric acid secretory capacity: a human study.
with upper gastrointestinal bleeding receiving intravenous proton J Clin Pharmacol 42(5):512–519
pump inhibitors: a multicentre evaluation. Can J Gastroenterol 18. Shin JM, Sachs G (2002) Restoration of acid secretion fol-
18(9):567–571 lowing treatment with proton pump inhibitors. Gastroenterology
10. Kaviani MJ, Hashemi MR, Kazemifar AR, Roozitalab S, 123(5):1588–1597
Mostaghni AA, Merat S, Alizadeh-Naini M, Yarmohammadi H 19. Metz DC, Amer F, Hunt B, Vakily M, Kukulka MJ, Samra N
(2003) Effect of oral omeprazole in reducing re-bleeding in bleed- (2006) Lansoprazole regimens that sustain intragastric pH > 6.0:
ing peptic ulcers: a prospective, double-blind, randomized, clinical an evaluation of intermittent oral and continuous intravenous infu-
trial. Aliment Pharmacol Ther 17(2):211–216 sion dosages. Aliment Pharmacol Ther 23(7):985–995
11. Javid G, Masoodi I, Zargar SA, Khan BA, Yatoo GN, Shah AH, 20. Netzer P, Gaia C, Sandoz M, Huluk T, Gut A, Halter F, Husler
Gulzar GM, Sodhi JS (2001) Omeprazole as adjuvant therapy to J, Inauen W (1999) Effect of repeated injection and continuous
endoscopic combination injection sclerotherapy for treating bleed- infusion of omeprazole and ranitidine on intragastric pH over 72
ing peptic ulcer. Am J Med 111(4):280–284 hours. Am J Gastroenterol 94(2):351–357
12. Cooper GS, Chak A, Lloyd LE, Yurchick PJ, Harper DL, Rosenthal 21. Brunner G, Luna P, Hartmann M, Wurst W (1996) Optimizing the
GE (2000) The accuracy of diagnosis and procedural codes for intragastric pH as a supportive therapy in upper GI bleeding. Yale
patients with upper GI hemorrhage. Gastrointest Endosc 51(4; Pt J Bio Med 69:225–231
1):423–426 22. Tseng GY, Lin HJ, Lin HY, Perng CL, Lee FY, Lo WC, Chang
13. Huber R, Hartmann M, Bliesath H, Luhmann R, Steinijans VW, FY, Lee SD (1999) The influence of intravenous omeprazole
Zech K (1996) Pharmacokinetics of pantoprazole in man. Int J Clin on intragastric pH and outcomes in patients with peptic ulcer
Pharmacol Ther 34(5):185–194 bleeding after successful endoscopic therapy—a prospective ran-
14. Pue MA, Laroche J, Meineke I, de Mey C (1993) Pharmacokinetics domized comparative trial. Hepatogastroenterology 46(28):2183–
of pantoprazole following single intravenous and oral administra- 2188

Springer