Routine use of acid-suppressive medication outside ICU questioned

April 15, 2011

Prophylactic acid suppression is not appropriate for most hospital patients who are not being treated in intensive care, according to researchers at Beth Israel Deaconess Medical Center, in Boston.

Key Points

Prophylactic acid suppression is not appropriate for most hospital patients who are not being treated in intensive care, according to researchers at Beth Israel Deaconess Medical Center, in Boston.

The researchers tracked nearly 80,000 patients for nosocomial gastrointestinal (GI) bleeding and acid suppression outside the intensive care unit (ICU). They concluded that the risks of nosocomial pneumonia and infection with Clostridium difficile or other organisms are greater than the potential benefits of proton pump inhibitors (PPIs) and histamine-2 receptor antagonists for most patients.

"Our data support guidelines that are already in place to limit the use of acid suppression in the general hospital population," said Shoshana J. Herzig, MD, MPH, lead investigator on the study and an instructor at Harvard Medical School and Beth Israel Deaconess Hospital.

"Overutilization of prophylactic acid suppression is a common and prevalent problem," she continued. "This is a pivotal study that not only supports the consensus guidelines on stress ulcer prophylaxis but also provides current data on just how low the risk of nosocomial GI bleeding is in non-ICU patients. This study should force institutions to re-examine stress ulcer prophylaxis practices."

The study was published in the February 14, 2011, issue of the Archives of Internal Medicine.

Thousands of patients

The researchers studied 78,394 non-ICU admissions at Beth Israel Deaconess Hospital occurring between January 2004 and December 2007 for evidence of nosocomial GI bleeding and exposure to acid-suppressive medications. Of those patients not critically ill, 59% were given acid-suppressive medication. Only 224 patients, 0.29%, developed nosocomial GI bleeding while hospitalized.

Use of acid-suppressive medication was associated with a 37% reduction in the odds of developing nosocomial GI bleeding. The researchers concluded that despite this protective effect and in light of the low overall incidence of developing nosocomial GI bleeding, 770 patients would need to be treated with acid-suppressive medication to prevent 1 episode of nosocomial GI bleeding and 834 to prevent 1 episode of clinically significant nosocomial GI bleeding.

These results should be considered in the context of previous studies, said the authors. One study found an association between acid-suppressive medication and hospital-acquired C. difficile infection, with a number needed to harm of 533. Another found a number needed to harm of 111 for hospital-acquired pneumonia.

Mission creep

The broad use of prophylactic acid suppression is a case of mission creep, Herzig said. When early studies found benefit in ICU patients, clinicians extrapolated the results to the broader hospital population.

"We thought that prophylactic acid suppression was probably not the right practice, but there were no real data outside the ICU," Herzig said. "There have been smaller studies in severely ill patients with risk factors for stress ulcers, but we are not aware of any other studies of prophylactic acid suppression and GI bleeding in a large population of patients who are not critically ill."

A recent in-house study conducted at Mission Hospital found that 66% of patients received prophylactic acid suppression. However, 54% had no risk factors for nosocomial bleeding and 53% did not have acid suppression discontinued upon their leaving the ICU, Michalets noted.

"We try to be evidence-based in our practice," Michalets said. "We want to decrease unnecessary utilization and prevent harm in our patients."

The next step

The next step at both institutions is clinical education to reduce the use of inappropriate acid suppression. That includes a combination of in-service training, one-on-one consultations, informal discussion, and adjustments to electronic order entry systems.

"Physicians have been pretty willing adopters of new practice patterns here, but they're not changing practice fast enough," Herzig said. "Clinical decision-support software may help spread the change."