From the Journals

Data point to optimal window for endoscopy in sicker patients with peptic ulcer bleeding


 

The timing of endoscopy may make the difference between life and death in sicker patients with peptic ulcer bleeding, according to an analysis of more than 12,000 patients treated in Denmark.

Patients who were hemodynamically stable but had a higher level of comorbidity were about half as likely to die during their hospital stay if they underwent endoscopy within 12-36 hours of presentation as compared with sooner or later, results showed (Gastrointest Endosc. 2016 Sep 10. doi: 10.1016/j.gie.2016.08.049). And hemodynamically unstable patients had a roughly one-fourth reduction in the odds of death if they underwent the procedure within 6-24 hours.

“Although caution should be applied when interpreting these data, the current recommendation of endoscopy within 0-24 hours may not be optimal for all patients,” wrote the investigators, who were led by Stig B. Laursen, PhD, department of medical gastroenterology, Odense (Denmark) University Hospital.

“Our data may suggest that in patients with major comorbidities, the first few hours of hospital admission might be best used for optimising treatment of comorbidities, which may include correction of severe anaemia, reversal of anticoagulants, and investigation for possible infection that requires rapid treatment with antibiotics,” they elaborate. “Likewise, in patients with hemodynamic instability, endoscopy between 6 and 24 hours from time of admission to hospital allows time for optimal resuscitation and initiating treatment of comorbid diseases before endoscopy. However, these data should not lead to delayed endoscopy in patients with severe hemodynamic instability not responding to intensive resuscitation.”

The investigators analyzed data from 12,601 consecutive patients with peptic ulcer bleeding admitted between January 2005 and September 2013 to Danish hospitals, where all patients had access to 24-hour endoscopy. Time to endoscopy was assessed from hospital admission, defined as arrival in the emergency department, or from symptom onset in patients who developed bleeding when already hospitalized.

For analyses, the patients were stratified by hemodynamic status (a marker for the severity of bleeding) and by American Society of Anesthesiologists score (a marker for the extent of comorbidity).

The timing of endoscopy did not significantly influence in-hospital or 30-day mortality in hemodynamically stable patients with an American Society of Anesthesiologists score of 1-2 as a whole, Dr. Laursen and his colleagues report. Subgroup analyses suggested a reduction of in-hospital mortality when it was done between 0 and 24 hours in those patients whose bleeding began outside the hospital (adjusted odds ratio, 0.48).

In contrast, analyses revealed a U-shaped association between timing and mortality for hemodynamically stable patients with an American Society of Anesthesiologists score of 3-5. For this group, in-hospital mortality was significantly lower when endoscopy was performed within 12-36 hours as compared with times outside this window (adjusted OR, 0.48), and 30-day mortality tended to be lower as well.

Similarly, timing appeared to influence outcome for hemodynamically unstable patients, having both systolic blood pressure below 100 mm Hg and heart rate above 100 beats/min. For this group, performance of endoscopy within 6-24 hours was associated with significantly lower in-hospital mortality (adjusted OR, 0.73) and also 30-day mortality (adjusted OR, 0.66). Patients’ American Society of Anesthesiologists score did not appear to play a role here.

The study’s findings may have been affected by unmeasured and unknown confounders, acknowledge the investigators, who declared that they have no competing interests related to the research.

“Although a well-powered randomized controlled trial represents the best way to account for these problems, randomizing patients with [peptic ulcer bleeding] to early versus late endoscopy will be very difficult, including from an ethical and methodological point of view,” they note.

Next Article: