Keywords

Case Presentation

A 66 year-old male with a past medical history notable for coronary artery disease on daily aspirin therapy presents to the emergency department with 1 week of epigastric pain and two episodes of dark, tarry stool over the past 24 h. He denies vomiting or bright red blood per rectum. The patient’s heart rate is 122 beats per minute and his blood pressure is 116/70 mm Hg in the seated position but falls to 94/60 mm Hg in the standing position. On exam, he is noted to have conjunctival pallor, tenderness in the epigastrium without peritoneal signs, and melena on rectal exam. Labs are notable for a blood urea nitrogen (BUN) of 40 mg/dL, serum creatinine of 0.71 mg/dL, hemoglobin of 6.8, platelet count of 246 × 109/L, and an INR of 1.1. Two large-bore peripheral intravenous lines are placed and volume resuscitation is initiated. Intravenous proton pump therapy is administered. The emergency department calls the on-call intensivist to discuss transfer to the intensive care unit.

Question

What is the next best step in the management of this patient?

Answer

Blood transfusion

This patient’s hemoglobin is below the transfusion threshold of 7 gm/dL for upper gastrointestinal bleeding (UGIB) , and thus transfusing 1 unit of packed red blood cells is the next best step in management. Research supports a restrictive transfusion strategy in patients presenting with UGIB who are not experiencing massive exsanguinating bleeding or cardiac ischemia. Adequate volume resuscitation and acid suppression with proton pump inhibitors are key aspects of management, and triage should be based on hemodynamic status, active blood loss, and validated prognostic scoring systems. Nasogastric lavage is no longer recommended for routine use.

Principles of Management

Fluid Resuscitation and Triage

In patients presenting with an acute UGIB, the initial key steps in management are placement of vascular access and adequate fluid resuscitation. Two large-bore peripheral venous lines (16 or 18 gauge) should be placed immediately for adequate venous access. Given that catheter length and diameter influence flow rate, shorter peripheral venous lines have been shown to have up to a 164% increase in flow rate compared to central vein catheters of the same gauge and are thus preferred [1]. A central venous catheter with a large diameter and short length is a reasonable option in cases where peripheral venous access is challenging. To restore intravascular volume, an isotonic fluid (e.g. normal saline or Lactated Ringer’s solution) should be administered.

All patients presenting with UGIB should receive close monitoring of their hemodynamics and respiratory status. Patients who are hemodynamically stable on presentation or stabilize after resuscitation can typically be managed on a telemetry ward. However, patients who remain hemodynamically unstable despite fluid resuscitation or have evidence of active bleeding should be transferred to an intensive care unit for further management [2].

Blood Transfusion

Transfusion of packed red blood cells for patients with UGIB can help replace ongoing blood loss as well as increase oxygen delivery to vital organs. In cases of exsanguinating bleeding, blood transfusion is indicated regardless of the patient’s initial hemoglobin level. It should be noted that the hemoglobin level on presentation in rapidly bleeding patients may not reflect the actual degree of blood loss, as additional time is required for equilibration. For patients with more stable bleeding, however, the benefits of blood transfusion must be weighed against the potential risks.

Traditionally, guidelines favored a liberal transfusion threshold of hemoglobin 9–10 g/dL for patients presenting with UGIB [3]. However, observational studies demonstrated improved patient outcomes with a lower hemoglobin threshold [4, 5]. To more directly address this question, Villanueva et al. conducted a randomized controlled trial of 921 patients with acute UGIB [6]. In the study, patients were assigned to a restrictive (transfusion for hemoglobin <7 g/dL) or a liberal (transfusion for hemoglobin <9 g/dL) blood transfusion strategy. Notable exclusion criteria included massive exsanguinating bleeding, a blood transfusion within the past 90 days, or a recent acute coronary syndrome or stroke/transient ischemic attack. Patients with portal hypertension, however, were included in the study sample. Consistent with prior observational data, the patients in the restrictive transfusion strategy had improved patient outcomes compared to patients in the liberal transfusion strategy. Specifically, the restrictive transfusion strategy was associated with lower mortality, decreased rebleeding rates, and fewer adverse events. A recent meta-analysis including 1965 patients found similar benefits with reduced morality and rebleeding rates with a restrictive strategy [7].

As a result, current guidelines recommend a restrictive transfusion strategy with a hemoglobin threshold of 7 g/dL for most patients with UGIB [8,9,10]. Notable exceptions include those with exsanguinating bleeding as well as those with unstable coronary disease, for whom a higher transfusion threshold (e.g. 9 or 10 g/dL) is often used. The current evidence supporting a higher hemoglobin target in patients with coronary disease is limited and expert guidelines are variable.

Proton Pump Inhibitors

Acid suppression with proton pump inhibitors (PPIs) plays a key role in the management of acute UGIB. Through irreversible inhibition of the hydrogen potassium ATPase, the pH of the stomach rises and clot stabilization is favored [11]. The benefit of PPIs is most clear in individuals presenting with high-risk features of peptic ulcer disease. The Forrest classification defines high-risk features based on endoscopic appearance (Fig. 67.1) [9]. Type I ulcers are actively bleeding (Ia: spurting; Ib: oozing) and type II ulcers have evidence of recent bleeding (IIa: visible vessel; IIb: adherent clot). In a randomized controlled trial, Lau et al. randomized 240 patients presenting with peptic ulcer bleeding and high-risk features on upper endoscopy to 72 h of IV PPI infusion vs. placebo [12]. All patients received endoscopic therapy to achieve hemostasis on the upper endoscopy. The patients who received the IV PPI infusion after endoscopic hemostasis had significantly lower rates of rebleeding, although there was no difference in mortality. A subsequent randomized controlled trial investigated the role of IV PPI prior to upper endoscopy in 638 patients with acute UGIB [13]. The authors found that IV PPI accelerated healing of peptic ulcer disease, which led to decreased need for endoscopic therapy and decreased length of stay.

Fig. 67.1
figure 1

Forrest classification . The risk of rebleeding, surgery and mortality based on the Forrest classification of the ulcer seen on endoscopy

Full size image

Although the above studies both used continuous IV PPI infusions in the treatment arms, more recent evidence suggests that intermittent IV PPI dosing is a comparable strategy. This was illustrated in a meta-analysis of 13 randomized controlled trials in which intermittent PPI dosing was not inferior to a continuous PPI infusion with regard to rebleeding, need for surgery or repeat intervention, or mortality [14]. Thus, intermittent IV PPI infusion is a reasonable strategy for patients presenting with acute UGIB. Following endoscopic hemostasis, patients with high-risk features should receive an additional 72 h of IV PPI, whereas patients with low-risk features (Forrest type IIIa: pigmented spot; IIIb: clean-based) may be transitioned immediately to oral PPI post-endoscopy.

Risk Stratification

Prognostic scores have been developed to help risk stratify patients presenting with acute UGIB and are part of guideline-based care. The Glasgow-Blatchford score (GBS) and the AIMS65 score are two validated scores that do not require endoscopic data and thus can be calculated at presentation [15, 16]. These scoring systems are summarized in Table 67.1.

Table 67.1 Components of the Glasgow-Blatchford and AIMS65 scores
Full size table

The GBS was developed and validated to predict those patients presenting with UGIB who would need an intervention during their hospitalization, as defined by the need for blood transfusion, endoscopic therapy, or surgery [15]. The score ranges from 0 to 23 based on clinical data available at presentation. Patients with a GBS ≤ 2 are classified as low-risk and subsequent studies have shown that these patients can be safely managed as outpatients [17, 18]. The AIMS65 score is comprised of 5 factors that each contribute one point to the total score: Albumin less than 3.0 g/dL, INR greater than 1.5, Mental status change, Systolic blood pressure ≤ 90 mm Hg, and age older than 65 years [16]. The AIMS65 score was developed and validated to predict in-hospital mortality, with scores of 0 or 1 indicating low mortality (0.3% and 1.2%, respectively) and scores of 4 and 5 noting higher mortality (22% and 32%, respectively). When compared head-to-head, the GBS performed best at predicting need for blood transfusion whereas the AIMS65 score was superior in predicting inpatient mortality [19].

Of note, the Rockall score is the most commonly used post-endoscopic risk stratification score [20]. This score consists of both clinical and endoscopic parameters (e.g. endoscopic diagnosis, stigmata of recent bleeding) and thus is not available on presentation.

Evidence Contour

Nasogastric Lavage

Traditionally, nasogastric lavage (NGL) was part of the initial management of patients presenting with UGIB, given its potential role in both diagnosis and prognosis. However, the use of NGL has fallen out favor as it can be difficult to interpret and has been cited by patients as the most painful of commonly performed procedures in the emergency department [21]. Further, NGL was not found to improve patient outcomes such as 30-day mortality or need for transfusion in a large retrospective study of 632 patients with UGIB [22]. Although NGL was previously recommended prior to upper endoscopy to improve gastric visualization, a recent randomized controlled trial showed no difference in achieving adequate visualization as compared to IV erythromycin infusion. For these reasons, current guidelines no longer recommend the routine use of NGL in the management of UGIB [9, 10].

Platelet Transfusion

It is widely accepted that for patients presenting with UGIB, the platelet level should be maintained above 50 × 109/L. However, some have proposed transfusing platelets for patients with normal platelet values but are also on anti-platelet therapy. To investigate this question, Zakko et al. conducted a retrospective case-control study of patients admitted with GI bleeding and platelet levels above 100 × 109/L who were also taking anti-platelet therapy [23]. 204 patients received platelet transfusion (i.e. cases) and 204 patients did not (i.e. controls). The authors found that the cases did not have lower rebleeding rates than the controls despite receiving platelet transfusion. In fact, the cases had higher mortality than the controls—although this was likely due to residual confounding, as the cases did have evidence of more severe bleeding than the controls. Regardless, this study found no evidence to support platelet transfusion for patients on anti-platelet therapy who present with UGIB and normal platelet values. Further research on appropriate indications for platelet transfusion is needed to guide management.

Intubation for Upper Endoscopy

In patients presenting with severe UGIB, prophylactic endotracheal intubation is often considered to prevent cardiopulmonary adverse events such as aspiration pneumonia from significant hematemesis. However, a recent retrospective study of 200 patients presenting with UGIB and hematemesis and/or melena with hemodynamic instability found higher rates of adverse cardiovascular events in patients who underwent prophylactic endotracheal intubation vs. a propensity-matched cohort of subjects who were not intubated [24]. Although this was a retrospective study with risk of residual bias, these findings highlight the need to carefully weigh the risks and benefits of prophylactic endotracheal intubation when considering airway protection for patients presenting with severe UGIB.

Timing of Upper Endoscopy

Upper endoscopy in acute UGIB has the potential to diagnose, risk stratify, and treat the culprit bleeding lesion. However, the decision of when to perform endoscopy remains an area of interest. The majority of evidence suggests that urgent endoscopy (<12 h) leads to increased detection of high-risk lesions and subsequent endoscopic therapy without any clear benefit in clinical outcomes, whereas early endoscopy (<24 h) decreases length of stay and possibly the need for surgical intervention.

In one of the few prospective randomized controlled trials investigating the timing of endoscopy in UGIB, Lee et al. randomized 110 patients to receive immediate endoscopy (within 1–2 h of presentation) vs. elective endoscopy (within 48 h of presentation) [25]. The authors found no significant differences between the two groups in terms of mortality, recurrent bleeding, transfusion requirement, need for surgery or need for repeat endoscopy. Subsequent retrospective studies similarly found no change in clinical outcomes with urgent endoscopy performed within 12 h of presentation, despite an increased detection of high-risk lesions and need for therapeutic endoscopy [26,27,28].

A multicenter retrospective study of 909 patients with UGIB found that early endoscopy (within 24 h) was associated with a significant reduction in length of stay when compared with elective endoscopy (after 24 h) [29]. When restricting the analysis to the high-risk patients who needed endoscopic therapy, there was also a significant decrease in the rate of rebleeding and need for surgery. However, no mortality benefit was noted in either analysis. A nationwide retrospective study of 2592 patients with peptic ulcer bleeding similarly found a significant reduction in length of stay and need for surgery with early (within 1 day) vs. delayed (after 1 day) endoscopy [30]. For these reasons, current guidelines recommend performing upper endoscopy within 24 h for most patients with UGIB [2].

Of note, recent studies have shown a potential harm in performing upper endoscopy too urgently in patients with severe UGIB [31, 32]. In a nationwide retrospective study, Laursen et al. found that in patients with hemodynamic stability and elevated American Society of Anesthesiologists (ASA) scores , higher mortality rates were seen when upper endoscopy was performed within 12 h of presentation [32]. In patients with hemodynamic instability (regardless of ASA score), endoscopy within 6 h was associated with increased mortality. These findings suggest that there is an optimal timing for when to perform endoscopy in UGIB, and the immediate management should focus on resuscitation and management of comorbidities prior to endoscopy.