Another ICU Intervention That May Not Work as Advertised

Aaron B. Holley, MD


January 24, 2019

Deborah J. Cook, MD, from McMaster University in Hamilton, Ontario, Canada, has been publishing on gastrointestinal (GI) prophylaxis in the intensive care unit (ICU) for decades. I still cite her article on risk factors for bleeding when teaching on ICU rounds.[1] When I began in the ICU 20 years ago, before proton-pump inhibitors (PPIs) were as ubiquitous as they are now, Dr Cook drove the debate over using antihistamines (H2-blockers) versus sucralfate.[2,3] Once PPIs penetrated all markets, she compared their efficacy to using H2-blockers.[4,5]

At one time I was careful to risk-stratify all my ICU patients prior to initiating GI prophylaxis. Using an article published in the Annals of Internal Medicine,[6] which drew heavily from Dr Cook's work, I provided GI prophylaxis to anyone on prolonged mechanical ventilation or with coagulopathy. In the absence of those factors, I refrained out of fear that the ventilator-associated pneumonia (VAP) risk was too high.

Subsequently, my use of PPIs gradually, and somewhat insidiously, expanded, and I paid less attention to risk factors. This occurred as sucralfate fell out of favor at the bedside because it was difficult to administer, and PPIs proved superior to H2-blockers for reducing bleeding and equivocal in causing VAP.[4,5] In addition, the Surviving Sepsis Guidelines listed PPIs as the preferred choice for prophylaxis.[7] Meanwhile, the risk for Clostridium difficile infection (CDI) due to PPI use became better understood.[8]

Where are we now? Lucky for us we have Dr Cook to keep us current. Earlier this year she published an excellent review on this topic in the New England Journal of Medicine (NEJM).[9] She covers meta-analyses, randomized controlled trials, and observational studies that address the prevention of GI bleeding in the ICU. Recent meta-analyses seem to support PPIs over H2-blockers for efficacy, but the risk for nosocomial infections, mainly VAP and CDI, is real. Unfortunately, quantifying the risk-versus-benefit for an individual patient is challenging. Dr Cook notes in her conclusion that the benefits from acid suppression might be smaller than originally thought, and overprescription is rampant. She ends her review with a call for defining which populations are best served by prophylactic acid suppression therapy.

Enter the recent randomized controlled trial published in NEJM,[10] which compared prophylactic acid suppression with pantoprazole versus placebo in patients at increased risk for GI bleeding. There was no difference in the primary outcome of 90-day mortality and no difference in the predefined composite outcome of clinically important events during admission (GI bleeding, CDI, myocardial ischemia, or pneumonia). When breaking the composite into its individual components, the percentage of patients with GI bleeding was 4.2% versus 2.5% in the PPI and placebo groups, respectively, and the percentage of patients with CDI or pneumonia was identical in both groups. Unfortunately, due to trial design, no P values are available for the individual clinically important events, only the composite. So it remains possible that there is a small reduction in GI bleeding, and this might mean that PPIs are beneficial, given equivalence in nosocomial infection rates.[11]

I am less optimistic. Study entry criteria required that patients have one of the following risk factors for GI bleeding: shock, use of anticoagulant agents, renal-replacement therapy, mechanical ventilation (expected to last > 24 hours), any history of liver disease, or any history of ongoing coagulopathy. A look at the patient characteristics at baseline (Table 1 in the study) confirms that these patients were pretty sick, and many would have had multiple risk factors for bleeding. You really cannot stack the deck any better. Although the difference in GI bleeding may be important, there is no overall difference in the primary or composite outcomes. The vast majority of patients I have prophylaxed with PPIs probably did not benefit from them, and we still have not identified a population that does. Perhaps said population does not exist.

Follow Medscape on Facebook, Twitter, and Instagram.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.