By Arzo Hamidi, PharmD, BCCCP
Clinical Pharmacy Specialist, Adult Critical Care, Rush University Medical Center, Chicago
SYNOPSIS: In previously critically ill adults, the use of a proton pump inhibitor without an indication for more than eight weeks increased morbidity and mortality.
SOURCE: Palmowski L, von Busch A, Unterberg M, et al. Timely cessation of proton pump inhibitors in critically ill patients impacts morbidity and mortality: A propensity score-matched cohort study. Crit Care Med 2024;52:190-199.
In a retrospective cohort study using health claims data from Germany, data were reviewed for more than 11,000 patients. The cohort included critically ill adults who did not receive a proton pump inhibitor (PPI) for a period of more than one month in the year prior to the hospital admission, who then were initiated on PPI treatment during their intensive care unit (ICU) stay, and subsequently did not have a coded indication for PPI therapy beyond eight weeks after hospital discharge. The patients’ diagnoses and preexisting conditions were identified for potentially inappropriate therapy. Patients then were allocated into two groups: group 1 did not receive any PPI therapy after discharge, and group 2 received PPI after hospital discharge without any identifiable indication.
In the patients studied, almost half of them received a PPI for more than one year. There were no differences at baseline between groups in terms of ICU or hospital type where the PPI was initiated. Those who received a PPI after discharge without an indication (group 2) had a 27% increased risk of pneumonia (odds ratio [OR], 1.27; 95% confidence interval [CI], 1.15-1.39; P < 0.001) compared to patients who did not receive a PPI after discharge. In group 2 vs. group 1, there also was a 26% greater risk of developing chronic renal failure (OR, 1.26; 95% CI, 1.26-1.41; P < 0.001), but no difference in the risk of acute interstitial nephritis. Compared to group 1, group 2 had a 30% increased risk of vitamin B12 deficiency (OR, 1.3; 95% CI, 1.13-1.49; P < 0.001), a 2.1-fold risk of hypomagnesemia (OR, 2.1; 95% CI, 1.24-3.6; P < 0.001), and a 55% increased risk of hypocalcemia (OR, 1.55; 95% CI, 1.22-1.96; P < 0.001). There also were increased risks of esophageal, pancreatic, and colon cancers associated with PPI therapy. The rehospitalization rate of patients in group 2 was 73% within one year. After two years, 72.5% of patients in the continued PPI group survived compared to 74.7% in the group who did not receive PPI post-ICU admission (adjusted hazard ratio, 1.17; 95% CI, 1.08-1.27; P < 0.001).
COMMENTARY
This study shows an increased risk of morbidity and mortality from a large group of previously critically ill patients who received PPI therapy beyond eight weeks after hospital discharge without an indication. While these were retrospective data, the authors used regression analysis to determine the relevant factors to implement in the propensity score matching. This allowed for a robust comparison of patients between groups. The risk associations seen in these patients have been reported previously in the literature. The adverse events seen also are biologically plausible since PPI use can change bacterial flora and decrease gastric acid secretion.
However, the true likelihood of PPI therapy causing such adverse effects is controversial. Despite the robust design and statistical analyses, association does not necessarily equate to causation in retrospective, observational studies from a health claims database. Additional diagnoses could have contributed to patient outcomes in the following year. Missing data or incorrect coding also could skew data as well. Furthermore, PPI adherence cannot be assessed from retrospective data to know if all patients were actively taking the medication they had filled from the pharmacy.
Presumably, the most common indication for PPI use in the ICU was stress ulcer prophylaxis. Since the authors did not collect data on patients who received histamine-2 receptor antagonists after ICU discharge, it is unclear if there are increased risks with other similar medications.
One of the biggest takeaways from this study is to always review and understand the indications of any medications a patient is receiving. De-escalation with decreased risk factors or discontinuation when an indication no longer is present are both important action items to take with any medication to avoid the risk of adverse effects. Future studies should evaluate the risk of additional medications post ICU discharge.
