An estimated 390,000 hospitalizations occur with a principal diagnosis of gastrointestinal bleeding each year in the United States, and an additional 600,000 occur with that as a secondary diagnosis. Patients hospitalized with upper GI bleeding are generally very ill, and have high overall mortality rates ranging from 4.5 to 8.2 percent. (JAMA 2012;307:1072.)
The evaluation of patients with suspected upper GI bleeding in many institutions obligates the use of nasogastric lavage to assist in confirming the diagnosis and guiding management. Many gastroenterologists consider nasogastric lavage a mandatory part of the pre-endoscopic workup of ED patients with suspected upper GI bleeding. Advocates of nasogastric lavage point to previous studies that suggest the presence of blood in nasogastric lavage can help determine the source of bleeding and predict active bleeding, high-risk lesions, and rates of recurrent hemorrhage, all of which can be used to help risk-stratify these patients. (Gastrointest Endosc 2004;59:172.)
On the other hand, nasogastric lavage, despite being minimally invasive, has been rated by patients as one of the most painful procedures performed in the emergency department. (Ann Emerg Med 1999;33:652.) It has also been associated with epistaxis, aspiration, pneumothorax, and perforation. (Ann Emerg Med 2010;55:365.) Other studies have suggested that lavage may actually worsen hemorrhage by disrupting established fibrin clots (J Surg Res 1980;28:204), and some have shown high false-negative rates (up to 18%) and inaccurate interpretation of the results up to 50 percent of the time. (Arch Intern Med 1990;150:1381.) The practice of nasogastric lavage use for upper GI bleeding varies widely, and has been a topic of deliberation for quite some time.
Impact of Nasogastric Lavage on Outcomes in Acute GI Bleeding
Huang ES, Karsan S, et al
This study retrospectively looked at patients presenting to the ED of West Los Angeles Veterans Affairs Medical Center between 1996 and 2005 with GI bleeding. Patients were included if they their primary reason for admission was GI bleeding, they were hospitalized for at least 24 hours, they received an inpatient GI consultation, and they had a complete set of relevant clinical data documented in the chart. Notably, patients were excluded if not hospitalized, if GI bleeding developed during hospitalization, or if they left against medical advice.
Detailed review of electronic records was used to determine if nasogastric lavage was performed in the ED. The study looked at length of hospitalization, transfusion requirements, use of endoscopy, time to endoscopy, performance of emergency surgery, and all-cause mortality within 30 days of index hospitalization, including inpatient mortality. A total of 632 patients met inclusion criteria, of which 60 percent received nasogastric lavage (case group) and 40 percent (control group) did not. Patients in each group were carefully assigned an illness severity and matched one-to-one to minimize bias and possible confounders; 193 subjects were in each group.
The study found no significant difference between the groups in transfusion requirement, length of hospital stay, need for surgery, or 30-day mortality. It was noted, however, that patients undergoing nasogastric lavage were more likely to undergo endoscopy (OR 1.71; 95% CI, 1.12–2.62, p=0.02) and to have this intervention performed sooner (hazard ratio 1.49; 95% CI, 1.09–2.04, p=0.012). The authors also found that bloody aspirates correlated with finding high-risk lesions on endoscopy among individuals who had nasogastric lavage (OR 2.69; 95% CI, 1.08–6.73).
This study had important limitations. The patient population was fairly homogenous, composed entirely of U.S. veterans, and the results may not be applicable to patients in EDs with more diverse populations. It was also not a randomized controlled trial. The investigators went to great lengths to minimize bias by assigning each patient an illness severity and matching patients in each group, but the results are still more likely to be influenced by confounders that a randomized trial would largely eliminate.
The American College of Gastroenterology conditionally said nasogastric lavage in patients with upper GI bleeding is not required for diagnosis, prognosis, visualization, or therapeutic effect. (Am J Gastroenterol 2012;107:345.) If the motivation to perform nasogastric lavage is to get your patient to endoscopy sooner, the data from this study could be used to support that practice. It is probably a much better goal, however, to minimize interventions that offer little help in ED decision-making and have no proven clinical benefit.
The data from this study support the college's guidelines, showing that foregoing nasogastric lavage has no effect on transfusion requirements, need for surgery, hospital stay, or mortality. Waving this time-consuming procedure might also permit you and your staff to focus on interventions more likely to benefit patients such as adequate resuscitation, correcting coagulopathy, early transfusion, serial HCTs, and assessing response to therapy. (Ann Emerg Med 2010;55:365.) Your patient will also appreciate your making his comfort a priority.
Click and Connect! Access the links in EMN by reading this issue on our website or in our iPad app, both available on www.EM-News.com.