Calls to curb unnecessary prescribing of proton-pump inhibitors (PPIs) in the hospital have been sounding for several years, but the thorny problem continues despite quality improvement efforts.
Now, the COVID-19 pandemic is complicating the issue further. On one hand, hospitalists are under more pressure than ever, with less time to focus on tasks outside of daily clinical care. At the same time, there is a growing body of literature raising questions about whether PPIs may be tied to an increased risk for SARS-CoV-2 infection or if they could worsen outcomes for patients hospitalized with COVID-19.
“The data raises concerns about the use of PPIs, especially in people with known COVID-19,” said Brennan Spiegel, MD, a gastroenterologist and director of health services research for Cedars-Sinai Medical Center in Los Angeles. “And yet I do see many people in the hospital, particularly those in the ICU, who are receiving intravenous PPIs.”
PPIs and COVID-19 risk
Dr. Spiegel and his colleagues kicked off debate about PPIs and COVID-19 with an August 2020 publication in the American Journal of Gastroenterology. Their study was aimed at testing whether patients on PPIs were more susceptible to SARS-CoV-2 infection, since PPIs have been associated with increased odds for enteric infections. The mechanism believed to be at work is PPI-induced hypochlorhydria, which breaks down the gut's natural defense against ingested bacteria and viruses.
The study was based on an online survey of more than 50,000 adults, asking about gastrointestinal (GI) symptoms, medications, and COVID-19 positivity. After multivariate regression analysis, the researchers found that individuals taking PPIs were more likely to report a positive COVID-19 test. The greatest risk was associated with taking PPIs twice daily (adjusted odds ratio [OR], 3.67). The study did not find any association between a positive COVID-19 test and use of a histamine-2 receptor antagonist.
“What we found that was most notable was it was the use of high-dose PPIs, twice a day rather than once a day, that increased the risk of acquiring COVID-19,” Dr. Spiegel said. “We did not look, in that study, at the impact on mortality or severity of disease, but since that time there have been many studies that have found an association between inpatient mortality, the need for intubation and mechanical ventilation, and other more severe measures of outcomes among people using PPIs.”
For example, a meta-analysis of nine studies, published as a letter by Gut in November 2020, found that current or regular PPI users were more likely to have severe outcomes of COVID-19 than nonusers (OR, 1.67). An October 2020 letter to the Journal of General Internal Medicine reported similar results from a meta-analysis, as well as significantly increased odds for the development of secondary infections in COVID-19 patients on PPIs (OR, 2.91).
In contrast, another retrospective analysis, published as a brief communication in Gastroenterology in November 2020, found that preadmission PPI use was not independently associated with the need for mechanical ventilation or death in COVID-19 patients. It also indicated that histamine-2 receptor antagonists, which some had thought might have a beneficial effect, were associated with increased risk for mechanical ventilation. “These findings caution against the modified use of these medications to improve COVID-19 related outcomes,” the researchers wrote.
An even more recent meta-analysis, published Feb. 17 by Gastroenterology, looked at nine studies and found there was no association between famotidine use and COVID-19 severity or death, and the data on PPI use and outcomes were mixed.
“Every one of those studies, ours included, is limited methodologically and every time someone publishes a new paper, there is appropriate skepticism around it, but at some point you've got to say, ‘Is there a kernel of truth in there?’” Dr. Spiegel said.
Taken together, these studies overall suggest that physicians should be especially judicious in their use of PPIs in patients hospitalized with COVID-19, Dr. Spiegel said. He is eagerly awaiting the results of a recently completed randomized controlled trial looking at the histamine-2 receptor antagonist famotidine as a potential inhibitor of COVID-19.
Until such randomized results are available, it's important to remember that there is no evidence at present showing a causal relationship between PPIs and COVID-19 infection or severity, noted Omar Kherad, MD, MPH, an internist with the University of Geneva in Switzerland and McGill University in Canada who has researched inpatient PPI use. “These results must be taken with caution,” he said.
Whether hospitalists are responding cautiously or just haven't seen the rapidly changing literature related to PPIs and COVID-19, practice regarding PPI use does not appear to have changed significantly due to the pandemic, according to Bradley Sharpe, MD, division chief of hospital medicine at University of California, San Francisco (UCSF), Health. “We just don't know,” he said. “We do not have specific guidance other than the usual indications for acid suppression in the hospital.”
COVID-19 is just the latest complication for hospitalists trying to figure out appropriate use of PPIs.
Existing guidelines from the American College of Gastroenterology indicate that PPIs are justified for the treatment of patients with gastroesophageal reflux disease (GERD), GI bleeding, and esophagitis but should only be used for prevention of ulcer in the case of risk factors such as age, concomitant use of NSAIDs or antiplatelet agents, or a history of ulcer, explained Dr. Kherad.
Several adverse events have been linked to the use of PPIs, but the quality of evidence behind these associations has been quite low, Dr. Kherad said, particularly around bone fracture, dementia, and community-acquired pneumonia. There is stronger evidence to show that PPIs are associated with vitamin B12 deficiency, Clostridioides difficile infection, and hypomagnesemia.
A 2013 Choosing Wisely recommendation from the Society of Hospital Medicine urged physicians to refrain from prescribing medications for stress ulcer prophylaxis to medical inpatients unless they are at high risk for GI complications. More recently, an international consensus group issued recommendations on use of PPIs in nonvariceal upper GI bleeding, which were published in Annals of Internal Medicine on Dec. 3, 2019. The group also recommended therapy for patients with previous ulcer bleeding who now require antiplatelet therapy, anticoagulant therapy, or cardiovascular prophylaxis.
Despite some guidance spelling out the clinical situations that call for PPI use, there is still a knowledge gap in many hospitals, Dr. Sharpe said. Recommendations on PPIs may involve a complex algorithm for decision making. “It's a complicated issue to distill it to something that is easy for the provider to remember,” he said. “So I think that's a challenge.”
Part of the problem is also that PPI use is so widespread, said Jennifer Austin Szwak, PharmD, a clinical pharmacist specialist at University of Chicago Medicine, who said use extends far beyond GERD, upper GI bleeds, and stress ulcer prophylaxis.
“We use it [for] those [reasons] but we also use it for a lot of different other reasons that aren't necessarily defined in guidelines,” Dr. Szwak said.
If a patient is admitted already on PPIs, it can be hard to even identify those reasons. In some cases, Dr. Szwak said, a hospitalist and pharmacist may have to trace prescribing decisions back over years to figure out why a patient was originally prescribed a PPI, possibly during a previous hospitalization. If clinicians cannot determine why patients are on a PPI, more likely than not they will stay on that medication, Dr. Szwak noted.
“Time becomes a big barrier,” she said. Often, reconciliation of the medication list happens at the time of discharge, and even if the hospitalist and the hospital pharmacist are asking the right questions, the patients may not have the answers, or the primary care clinician may not be available. “Sometimes there's just not time in the day to do all those things, for the pharmacist or hospitalist, but also for the primary care provider on the other side of that who may be busy and be unable to respond,” she said.
Another challenge in improving the appropriate inpatient prescribing of PPIs is quality improvement fatigue. Many institutions have implemented quality improvement programs, particularly education initiatives, around PPI use, Dr. Sharpe said, but the effects tend to wane over time.
“This may be a good example of the quality improvement principle that sustainability is hard,” Dr. Sharpe said. “Focusing on an intervention to improve quality often yields outcomes, yet as time goes on it's hard to sustain effort around one particular challenge because there are so many different challenges that come up.”
At UCSF, they have had some continued success by integrating a structured checklist into their electronic health record's progress note template, which includes a component on GI prophylaxis.
The goal is to prompt the prescriber to consider if a patient is currently on a PPI and if so, whether it's needed. Additionally, on the teaching service at UCSF, when housestaff or students present their patients, they go through the same checklist and explain whether the patient is taking a PPI and the indication for it.
Building high-quality, integrated decision support focused on PPI prescribing directly into the electronic health record to provide alerts at the point of care would be a big boost, Dr. Sharpe said, but that's likely still years down the road.