Gastrointestinal Endoscopy
Volume 70, Issue 2 , Pages 236-239, August 2009

Endoscopy for acute upper GI tract hemorrhage: sooner is better

VA Greater Los Angeles Healthcare System, David Geffen School of Medicine at UCLA, CURE Digestive Diseases Research Center, UCLA/VA Center for Outcomes Research and Education, Los Angeles, California, USA

Article Outline

Abbreviation: UGIH, upper GI tract hemorrhage

 

Now is the time for ≤24-hour endoscopy to become as universal as ≤1-hour “door-to-needle” times for acute myocardial infarction.

It is now standard-of-care to perform endoscopy in patients presenting with acute upper GI tract hemorrhage (UGIH). But this was not always the case. In 1980, Graham1 concluded that the natural history of acute UGIH “precludes endoscopy from having a significant effect on patient management.” The following year, Peterson et al2 published a widely cited, randomized, controlled trial in the New England Journal of Medicine comparing routine endoscopy to no routine endoscopy and concluded that “endoscopy should not be a routine procedure in patients with upper gastrointestinal tract bleeding.” Because endoscopy could not provide definitive hemostasis in those earlier years, it is not surprising that merely viewing the gastric mucosa failed to make a significant difference in the clinical course of patients.

It took another decade of technological advances and accrual of clinical trial data for endoscopy to gain its rightful place as a vital intervention for acute UGIH. In a landmark meta-analysis published in the Journal of the American Medical Association in 1990, Sacks et al3 demonstrated that endoscopic hemostasis is superior to no endoscopic hemostasis. The analysis definitively revealed that hemostasis reduced the risk of continued bleeding, minimized recurrent hemorrhage, lowered the rate of emergency surgery, and provided a 30% risk reduction for overall mortality, particularly in patients with active spurting ulcers. These findings were corroborated by a second meta-analysis in 1992 by Cook et al,4 which included additional studies and further demonstrated that endoscopy with hemostasis improves outcomes and saves lives. In short, the age of routine endoscopy had finally arrived after years of trial and error in developing and evolving endoscopic technology.

Once endoscopy was proven to be effective, the next step was to define its optimal timing in the management of acute UGIH. As a general principle, effectiveness and timing are often inseparable for both medical and procedural interventions. For example, it is not adequate to simply provide coronary angioplasty for patients with myocardial infarction—it must be performed within 1 hour of presentation.5 Similarly, although tissue plasminogen activator has proved effective in acute ischemic stroke, it must be initiated within hours of onset to maximize outcomes.6 The question for gastroenterologists was whether these analogies applied to endoscopy. That is, when it came to endoscopy for acute UGIH, was sooner better?

Some authors initially raised concerns about rapid endoscopy. In particular, there were various reports in the mid 1990s that emergent endoscopy was associated with more complications than delayed endoscopy.7, 8, 9 For example, Yen et al8 assessed the safety of early endoscopy and found that it resulted in an increased risk of oxygen desaturation. However, this study used early endoscopy disproportionately in a high-risk group, and endoscopy was performed without sedation—a rare occurrence in modern practice. In a second study, Hill et al9 documented that 80% of patients undergoing emergency endoscopy had oxygen desaturation without supplemental oxygen and found that this desaturation rate far surpassed that of elective endoscopy. However, this study was not directed solely at UGIH, and the comparison groups were not comparable in many important ways. Nonetheless, these studies raised the specter of iatrogenic complications from early endoscopy and focused the question on whether the potential risks of urgent endoscopy, in particular, might outweigh the benefits. In other words, just because endoscopy was proven to be safe and effective for acute UGIH, it did not follow that earlier endoscopy was more effective or safer than later endoscopy—so long as endoscopy was ultimately performed during the hospital stay.

Moreover, a policy of early endoscopy could be difficult and expensive to implement in the real world. The associated costs of operating a 24-hour endoscopy service are substantial, as this not only requires support of additional physicians, nurses, and equipment technicians, but also requires extensive facility investments, particularly for institutions not well equipped to perform endoscopy in the emergency department. Early endoscopy could further escalate costs by detecting stigmata of recent hemorrhage that might otherwise have resolved by the time of delayed endoscopy, thereby leading to potentially unnecessary interventions. In addition, if endoscopy is performed too rapidly, particularly in the setting of active exsanguination, then it might yield an enormous and obstructing “placental” clot that prevents identification or treatment of culprit lesions. This requires repeat endoscopy, adds risk, and escalates costs. In short, there are reasons to believe that routine early endoscopy could be expensive, potentially risky, and in some systems altogether untenable.

In light of these concerns, and given the previous lack of consensus regarding the optimal timing of endoscopy in acute UGIH, we performed a systematic review of the literature in 2001 to address the following questions: (1) Does early endoscopy allow for safe and prompt discharge of low-risk patients with acute nonvariceal UGIH? (2) Does early versus delayed endoscopy improve patient outcomes in high-risk patients with acute nonvariceal UGIH? and (3) Does early versus delayed endoscopy reduce resource utilization for all patients with acute nonvariceal UGIH?10 We found 23 studies at the time that addressed the question of endoscopy timing and its relationship to outcomes. The highest-quality study examining outcomes in low-risk patients found no significant complications at 1-month follow-up for any outpatients managed with early endoscopy.11 The largest randomized trial of high-risk patients at the time showed no mortality benefit but a significant decrease in transfusion requirements with early endoscopy.12 Seven of the 8 studies examining the effect of early endoscopy on length of stay demonstrated a significant reduction compared with delayed endoscopy.11, 12, 13, 14, 15, 16, 17, 18 However, most of the studies we included suffered from 1 or more potentially significant methodological shortcomings.

Nonetheless, for all 3 clinical questions, the weight of the evidence favored early endoscopy. For low-risk patients, the evidence clearly supported the claim that early endoscopy promotes safe patient disposition. For high-risk patients, there was benefit from early versus delayed endoscopy in many clinical outcomes, including transfusion requirements, recurrence of bleeding, and need for emergency surgery. Furthermore, although there was no evidence that early versus delayed endoscopy decreased mortality, there was also no evidence that the practice resulted in patient harm. Regarding resource utilization, the evidence suggested that early endoscopy significantly reduced length of stay compared with delayed endoscopy for all risk groups with nonvariceal UGIH, without evidence of cost shifting to the outpatient setting. In short, despite previous concerns to the contrary, there was convergent validity that sooner was indeed better. This conclusion was subsequently highlighted in a major international consensus conference in 2003,19 solidifying early endoscopy (defined as within the first 24 hours) as a vital component of high-quality care in acute UGIH. This recommendation will remain in the updated international consensus guideline to be published in 2009.

In this issue of Gastrointestinal Endoscopy, Cooper et al20 provide more data that early endoscopy improves outcomes in acute UGIH. The authors point out that all previous studies in this area were conducted at individual hospitals or metropolitan areas but not at a national level. In their current study, in order to obtain population-based data about the prevalence and outcomes of this practice, the authors measured the impact of early endoscopy in a large cohort of Medicare beneficiaries with peptic ulcer hemorrhage. Using a 5% random sample of inpatient and outpatient Medicare claims from 2004, the authors conducted a cross-sectional analysis to evaluate the bivariate and multivariate relationship between early endoscopy (≤24 hours) and 3 outcomes: 30-day mortality, upper GI surgery, and length of stay. The study was restricted to patients older than 66 years, which includes the vast majority of Medicare beneficiaries. Because the study relied on administrative data sets, all definitions for case findings, comorbidities, and outcomes were based on ICD 9 codes.

The authors identified 2592 episodes of bleeding peptic ulcers in their 5% random sample, including 414 individuals who were managed as outpatients. Of this group, 71.5% had undergone endoscopy within the first 24 hours after initial presentation, as determined by the timing of claims data. The authors were unable to split this value any thinner, because there were no data about the time to endoscopy within the initial 24-hour window. This may not be important, in any event, as recent data indicate that outcomes remain constant regardless of when endoscopy is performed within the 24-hour window, so long as endoscopy occurs sometime during this critical period.21 After adjusting for potential confounders, the authors found that early endoscopy was independently associated with a dramatic reduction in length of stay (-1.95 days, 95% CI, -2.60 to 1.29 days) and a lower likelihood of surgery (odds ratio 0.37, 95% CI, 0.21-0.66) but no improvement in 30-day mortality.

There are some important limitations with this study. In particular, claims databases, although powerful in their scope and size, are incapable of fully adjusting for systematic differences between groups. In particular, patients who undergo early endoscopy are surely a very different group from those who undergo delayed endoscopy. In a recent audit at the West Los Angeles VA Medical Center, we found that, compared to patients undergoing delayed endoscopy, those with early endoscopy had a lower initial blood pressure, lower platelet count, higher transfusion requirement, and were more likely to be managed in an intensive care unit and have underlying varices (unpublished data). That is, patients undergoing early endoscopy had more severe illness compared to those undergoing delayed endoscopy. This is similar to findings by da Silveira et al,22 who documented that early endoscopy was more commonly performed in patients with gross blood on rectal examination, again suggesting a higher severity of bleeding. In the current study, Cooper et al were unable to collect this level of granular data regarding bleeding severity, such as vital signs, physical examination findings, nasogastric lavage results, key laboratory test values, or initial site of triage. Thus, it is possible that the negative relationship between early endoscopy and mortality is heavily confounded by greater bleeding severity in the early endoscopy group. It is notable that the early endoscopy group did have a numerically lower, albeit statistically insignificant, risk of mortality in this study (12% reduction in the odds of mortality with early vs delayed endoscopy). It remains untested whether this relationship might become statistically significant after results are adjusted for inevitable confounders. There have been randomized controlled trials, which should balance all measured and unmeasured confounders, but their relatively small sample size could yield a type II error for an endpoint like mortality. Future research should measure the independent effect of early endoscopy on mortality while adjusting for granular clinical variables—an effort that will require a highly detailed and comprehensive data collection approach.

In contrast to the negative relationship between early endoscopy and mortality, Cooper et al found a dramatic impact of early endoscopy on length of stay, as noted previously. Specifically, provision of early endoscopy reduced length of stay by nearly 2 hospital days. Because most “delayed” endoscopy still occurs within 48 hours (albeit not in the first 24 hours), it is difficult to conceive how a 24-hour head start on the front end leads to a 48-hour advantage on the back end. Moreover, if patients undergoing early endoscopy are systematically sicker than patients undergoing delayed endoscopy, it might be possible that length of stay would be no different or only modestly shorter after adjusting for differences in multivariable regression analysis. In our own local audit, using Poisson regression techniques, we found that early endoscopy independently reduces length of stay by approximately 8 hours—a small yet statistically significant effect. In a randomized controlled trial of early versus delayed endoscopy, Lee et al11 found a significant 1-day reduction in length of stay. Hussain et al13 compared length of stay for both low-risk and high-risk patients before and after a management protocol for peptic ulcer bleeding was initiated. They found a reduction of 1.24 days per low-risk patient and 1.01 days per high-risk patient with the use of early endoscopy. In 4 previous observational studies,14, 15, 16, 17 including 2 from the current authors,14, 16 there was a range of 1 to 2 days saved per patient. Finally, of 2 studies that specifically examined high-risk patient subgroups, 1 showed a decreased length of stay of 2 days,14 whereas 1 demonstrated an insignificant trend in shorter length of stay with early endoscopy.17 Overall, it seems clear that early endoscopy, on average, shortens length of stay. The large reduction in length of stay observed in this study (-1.95 days) is large and dramatic but is still within the range of other published reports, including previous reports from the current authors.14, 16 It may be possible that the impact of early endoscopy on length of stay is more pronounced in the elderly than in younger populations, which would explain why the impact in Medicare beneficiaries in this study is large compared to that of some other studies.

The finding that 70% of cases now receive early endoscopy is, in itself, an important and interesting result. Previous studies have generally shown lower rates of early endoscopy, indicating that the practice has been inconsistent. In fact, the decision to perform early versus delayed endoscopy has been independently associated with sex,23 insurance status,16 time of day,22 and day of the week.16 However, it now appears that the message is penetrating and that more patients than ever before are receiving early endoscopy—a sign of high-quality care. Yet despite this positive outlook, roughly 1 in 3 patients (in 2004, at least) in the large cohort underwent delayed endoscopy. Future research should investigate whether facility-based knowledge, attitudes, or beliefs predict the local rate of early endoscopy. If so, future educational intervention might be developed and deployed to target modifiable factors that drive facility-level adoption of early endoscopy. That is, the propensity to adopt early endoscopy may be driven as much or more by local resources, policies, or institutional culture as by personal knowledge and beliefs.

Although minor variations in early endoscopy rates are to be expected, large variations denote continued disconnects between theory and practice. In other areas of medicine, wide practice variations may have several explanations, including the need for more evidence to determine the best course of action, the possibility that multiple approaches may be equally effective for a clinical scenario, or the need for existing evidence to be more effectively consolidated into guidelines and disseminated into practice.24, 25, 26 In the case of early endoscopy for acute UGIH, none of these excuses apply. We now know, with near certainty, that sooner is better.10 The evidence has been effectively summarized and disseminated.19 Now is the time for 24-hour endoscopy to become as universal as 1-hour “door-to-needle” times for acute myocardial infarction. Timing matters, and for universal early endoscopy in acute UGIH, the time has finally come.

Back to Article Outline

Disclosure 

The author disclosed financial relationships relevant to this publication: B. M. R. Spiegel is supported by a Veterans Affairs Health Services Research and Development Career Development Award (RCD 03-179-2), the CURE Digestive Disease Research Center (NIH 2P30 DK 041301-17), and NIH Center grant 1 R24 AT002681-NCCAM as well as grants from AstraZeneca, TAP, Amgen, Bristol Myers Squibb, and Takeda, and is on the advisory boards of AstraZeneca, Takeda, and Prometheus.

Back to Article Outline

References 

  1. Graham DY. Limited value of early endoscopy in the management of acute upper gastrointestinal bleeding: prospective controlled trial. Am J Surg. 1980;140:284–290
  2. Peterson WL, Barnett CC, Smith HJ, et al. Routine early endoscopy in upper-gastrointestinal-tract bleeding: a randomized, controlled trial. N Engl J Med. 1981;304:925–929
  3. Sacks HS, Chalmers TC, Blum AL, et al. Endoscopic hemostasis. An effective therapy for bleeding peptic ulcers. JAMA. 1990;264:494–499
  4. Cook DJ, Guyatt GH, Salena BJ, et al. Endoscopic therapy for acute nonvariceal upper gastrointestinal hemorrhage: a meta-analysis. Gastroenterology. 1992;102:139–148
  5. Berger PB, Ellis SG, Holmes DR, et al. Relationship between delay in performing direct coronary angioplasty and early clinical outcome in patients with acute myocardial infarction: results from the global use of strategies to open occluded arteries in Acute Coronary Syndromes (GUSTO-IIb) trial. Circulation. 1999;100:14–20
  6. Tissue plasminogen activator for acute ischemic stroke. The National Institute of Neurological Disorders and Stroke rt-PA Stroke Study Group. N Engl J Med. 1995;333:1581–1587
  7. Iber FL, Livak A, Kruss DM. Apnea and cardiopulmonary arrest during and after endoscopy. J Clin Gastroenterol. 1992;14:109–113
  8. Yen D, Hu SC, Chen LS, et al. Arterial oxygen desaturation during emergent nonsedated upper gastrointestinal endoscopy in the emergency department. Am J Emerg Med. 1997;15:644–647
  9. Hill DB, Stokes BD, Gilinsky NH. Arterial oxygen saturation during emergency esophagogastroduodenoscopy: the effects of nasal oxygen. J Clin Gastroenterol. 1994;18:284–286
  10. Spiegel BM, Vakil NB, Ofman JJ. Endoscopy for acute nonvariceal upper gastrointestinal tract hemorrhage: is sooner better? A systematic review. Arch Intern Med. 2001;161:1393–1404
  11. Lee JG, Turnipseed S, Romano PS, et al. Endoscopy-based triage significantly reduces hospitalization rates and costs of treating upper GI bleeding: a randomized controlled trial. Gastrointest Endosc. 1999;50:755–761
  12. Lin HJ, Wang K, Perng CL, et al. Early or delayed endoscopy for patients with peptic ulcer bleeding: a prospective randomized study. J Clin Gastroenterol. 1996;22:267–271
  13. Hussain I, Farrar WH, Sofian EJ, et al. A cost effective strategy for managing GI bleeding secondary to peptic ulcer disease. [abstract] Gastroenterology. 1995;108:A18
  14. Cooper GS, Chak A, Way LE, et al. Early endoscopy in upper gastrointestinal hemorrhage: associations with recurrent bleeding, surgery, and length of hospital stay. Gastrointest Endosc. 1999;49:145–152
  15. Lindenauer PK, Roll F, Gebretsadik T, et al. Endoscopy improves outcome and reduces length of stay in elderly patients with acute upper gastrointestinal bleeding: results of a statewide population based analysis. [abstract] Am J Gastroenterol. 1999;116:A76
  16. Cooper GS, Chak A, Connors AF, et al. The effectiveness of early endoscopy for upper gastrointestinal hemorrhage: a community-based analysis. Med Care. 1998;36:462–474
  17. Rockall TA, Logan RF, Devlin HB, et al. Selection of patients for early discharge or outpatient care after acute upper gastrointestinal haemorrhage. National audit of acute upper gastrointestinal haemorrhage. Lancet. 1996;347:1138–1140
  18. Hay JA, Maldonado L, Weingarten SR, et al. Prospective evaluation of a clinical guideline recommending hospital length of stay in upper gastrointestinal tract hemorrhage. JAMA. 1997;278:2151–2156
  19. Barkun A, Bardou M, Marshall JK. Consensus recommendations for managing patients with nonvariceal upper gastrointestinal bleeding. Ann Intern Med. 2003;139:843–857
  20. Cooper GS, Kou TD, Wong RCK. Use and impact of early endoscopy in elderly patients with peptic ulcer hemorrhage: a population-based analysis. Gastrointest Endosc. 2009;70:229–235
  21. Targownik LE, Murthy S, Keyvani L, et al. The role of rapid endoscopy for high-risk patients with acute nonvariceal upper gastrointestinal bleeding. Can J Gastroenterol. 2007;21:425–429
  22. da Silveira EB, Lam E, Martel M, et al. The importance of process issues as predictors of time to endoscopy in patients with acute upper-GI bleeding using the RUGBE data. Gastrointest Endosc. 2006;64:299–309
  23. Terdiman JP, Lindenauer PK, Gebretsadik T, et al. Decision to perform early upper endoscopy in the elderly with upper gastrointestinal bleeding is not based on clinical factors that predict mortality. [abstract] Am J Gastroenterol. 1998;93:A100
  24. Brook RH, McGlynn EA, Shekelle PG. Defining and measuring quality of care: a perspective from US researchers. Int J Qual Health Care. 2000;12:281–295
  25. Brook RH, Lohr KN. Efficacy, effectiveness, variations, and quality: boundary-crossing research. Med Care. 1985;23:710–722
  26. Wennberg JE. Practice variations and health care reform: connecting the dots. Health Aff (Millwood). 2004;Suppl Web Exclusives:VAR140-4

 The opinions and assertions contained herein are the sole views of the author and are not to be construed as official or as reflecting the views of the Department of Veterans Affairs.

PII: S0016-5107(08)03132-5

doi:10.1016/j.gie.2008.12.053

Gastrointestinal Endoscopy
Volume 70, Issue 2 , Pages 236-239, August 2009

Article Tools