More Evidence Relating Gastric Acid Suppression to HAP

John G. Bartlett, MD


August 25, 2009

Although gastric acid-suppressing agents are used in about half of all hospitalized patients and their use has been associated with increased risk for pneumonia, the data are from relatively small studies. This viewpoint discusses a large study designed to address the risk for hospital-acquired pneumonia (HAP) due to acid suppressive therapy in hospital patients not receiving intensive care.

Acid-Suppressive Medication Use and the Risk for Hospital-Acquired Pneumonia

Herzig SJ, Howell MD, Ngo LH, et al
JAMA. 2009;301:2120-2128

Article Summary

Background. Acid-suppressive medications are given to 40% to 70% of medical inpatients, and it has been suggested that this practice promotes HAP as a result of reducing the "gastric barrier" to infection.

Purpose. To determine the association between gastric acid suppression and the risk for HAP.

Method. Records were reviewed for Beth Israel Deaconess Medical Center in Boston, Massachusetts, from January 2004 to December 2007. Case patients were those hospitalized over 3 days; intensive care unit patients were excluded. The diagnosis of HAP was based on ICD-9-CM coding for bacterial pneumonia divided into aspiration or nonaspiration pneumonia. Medication records were reviewed for dispensing either a proton pump inhibitor or an H2 receptor antagonist (ie, gastric acid suppression).

Results. The final cohort included 63,878 admissions. Of these, 52% received acid-suppressive medications, and HAP developed in 2219 (3.5%). The analysis showed a statistically significant risk was associated with acid suppression with an odds ratio for HAP of 1.3, which indicates a 30% increase. These data and those for the subsets of HAP that were or were not attributed to aspiration are shown in the Table.

Table. Frequency of HAP With and Without Gastric Acid Suppression

Outcome Acid Suppression
n = 32,922
No Acid Suppression
n = 30,956
Odds Ratio
Hospital-acquired pneumonia 1609 (4.9%) 610 (2.0%) 1.3
• Aspiration pneumonia 61 (1.1%) 112 (0.4%) 1.4
• Nonaspiration pneumonia 1262 (3.8%) 501 (1.6%) 1.2

The authors also did a propensity-matched analysis with 16,396 admissions with acid suppression for comparison with 16,396 without. In this analysis also, the odds ratio for HAP was 1.3 (95% confidence interval [CI], 1.1-1.4). Because the confidence interval does not include 1, the result is statistically significant.

Conclusion. The authors conclude that acid-suppressive medication use was associated with a 30% increase in the probability of HAP.


Comment. The use of acid suppression has long been the subject of controversy with respect to its risk of promoting HAP. The presumed mechanism is gastric acid neutralization with bacterial colonization of the stomach that then becomes the source of bacterial pneumonia due to aspiration. The distinction between "aspiration pneumonia" and "nonaspiration pneumonia" is not readily made on the basis of clinical observation.

"Microaspiration" is a major source of all pneumonia, which may be caused by bacteria that reside in either the upper airways or the stomach. Also of note is the fact that patients who are hospitalized often have upper airway colonization with gram-negative bacteria , not as a result of being in the hospital, but as a result of being sick. This presumably accounts for the high rates of gram-negative bacteria in HAP. It should be noted that this issue has been addressed previously, but those results are not as convincing; however, the very large sample size in this report gives it substantial statistical power.

The authors note that about half of their patients (and about half of all patients who are hospitalized in the United States) received acid-suppression; however, they note that about 50% are new initiations and that up to 70% of this use is for indications that are not well supported in the literature, most frequently stress ulcer prevention. They also point out that acid suppression has been implicated in promoting both Clostridium difficile-associated disease in all patients and ventilator-associated pneumonia in those receiving mechanical ventilation.