In the JournalsPerspective

PPIs linked to increased risk for death

Researchers found a significant association between long-term use of proton pump inhibitors and risk for death in an observational study of U.S. veterans.

This adds to a growing list of health problems recently linked to PPI use, including chronic kidney disease, osteoporosis and bone fractures, listeriosis, Clostridium difficile infection and dementia, though data are conflicting.

“No matter how we sliced and diced the data from this large data set, we saw the same thing: There’s an increased risk of death among PPI users,” Ziyad Al-Aly, MD, an assistant professor of medicine at Washington University School of Medicine in St. Louis, said in a press release. “For example, when we compared patients taking H2 blockers with those taking PPIs for 1 to 2 years, we found those on PPIs had a 50% increased risk of dying over the next 5 years. People have the idea that PPIs are very safe because they are readily available, but there are real risks to taking these drugs, particularly for long periods of time.”

Al-Aly and colleagues used VA medical records to evaluate the association between PPI use and risk for all-cause mortality in a primary cohort of nearly 350,000 U.S. veterans who newly began taking PPIs (n = 275,933) or H2 blockers (n = 73,355) between October 2006 and September 2008.

Throughout a median follow-up of 5.71 years, the researchers observed a 25% increased risk for death associated with PPI use vs. H2 blockers (HR = 1.25; 95% CI; 1.23-1.28). This corresponds to one excess death for every 500 PPI users annually, which could translate into thousands of excess deaths per year given the widespread use of PPIs, according to investigators.

They also evaluated associations between PPI use vs. no PPI use in an additional cohort of more than 3 million veterans, and between PPI use vs. no PPI use and no H2 blocker use in more than 2.8 million.

They observed a higher risk for death when comparing PPI use vs. no PPI use (HR = 1.15; 95% CI, 1.14-1.15), and PPI use vs. no PPI and no H2 blockers (HR = 1.23; 95% CI, 1.22-1.24).

The risk for death linked to PPIs was also higher in those without GI conditions: PPIs vs. H2 blockers (HR = 1.24; 95% CI, 1.21-1.27), PPIs vs. no PPIs (HR = 1.19; 95% CI, 1.18-1.2), and PPIs vs. no PPIs and no H2 blockers (HR = 1.22; 95% CI, 1.21-1.23).

Finally, the researchers observed a steady increase in risk for death with a longer duration of exposure among new PPI users. Death risk between PPI and H2 blocker users was not significantly different after 30 days, but after 1 to 2 years, the risk was almost 50% higher in PPI users vs. H2 blocker users.

“A lot of times people get prescribed PPIs for a good medical reason, but then doctors don’t stop it and patients just keep getting refill after refill after refill,” Al-Aly said in the press release. “There needs to be periodic re-assessments as to whether people need to be on these. Most of the time, people aren’t going to need to be on PPIs for a year or two or three.”

The researchers acknowledged that the cohort included mostly older white men, which could limit the generalizability of their findings. They also noted that PPI users were older, sicker and had higher rates of diabetes, hypertension, and cardiovascular disease compared with H2 blocker users, but said the increased risk for death remained after controlling for age and illness.

The researchers concluded that limiting PPI use and length of exposure to medically necessary situations is warranted.

“PPIs save lives,” Al-Aly said in the press release. “If I needed a PPI, I absolutely would take it. But I wouldn’t take it willy-nilly if I didn’t need it. And I would want my doctor to be monitoring me carefully and take me off it the moment it was no longer needed.”

In light of the growing concern about the health risks of PPIs, the AGA recently issued 10 best practice recommendations for long-term PPI use, and emphasized that the benefits likely outweigh the risks when PPIs are prescribed appropriately. In response to the current study, the AGA has also released an interpretation of the data on its website, along with a guide for effectively communicating about the findings with patients.

“While a large data set was used, the causes of death were not reported. Nevertheless, the study is receiving a lot of attention and clinicians can expect to hear from many interested and worried patients and colleagues,” AGA wrote in the guide. – by Adam Leitenberger

Reference:

AGA. A Guide to Conversations About the Latest PPI Research Results. Accessed July 7, 2017. http://www.gastro.org/news_items/a-guide-to-conversations-about-the-latest-ppi-research-results

Disclosures: The researchers report no relevant financial disclosures.

Editor's note: This article was updated on July 7 with additional information from AGA.

Researchers found a significant association between long-term use of proton pump inhibitors and risk for death in an observational study of U.S. veterans.

This adds to a growing list of health problems recently linked to PPI use, including chronic kidney disease, osteoporosis and bone fractures, listeriosis, Clostridium difficile infection and dementia, though data are conflicting.

“No matter how we sliced and diced the data from this large data set, we saw the same thing: There’s an increased risk of death among PPI users,” Ziyad Al-Aly, MD, an assistant professor of medicine at Washington University School of Medicine in St. Louis, said in a press release. “For example, when we compared patients taking H2 blockers with those taking PPIs for 1 to 2 years, we found those on PPIs had a 50% increased risk of dying over the next 5 years. People have the idea that PPIs are very safe because they are readily available, but there are real risks to taking these drugs, particularly for long periods of time.”

Al-Aly and colleagues used VA medical records to evaluate the association between PPI use and risk for all-cause mortality in a primary cohort of nearly 350,000 U.S. veterans who newly began taking PPIs (n = 275,933) or H2 blockers (n = 73,355) between October 2006 and September 2008.

Throughout a median follow-up of 5.71 years, the researchers observed a 25% increased risk for death associated with PPI use vs. H2 blockers (HR = 1.25; 95% CI; 1.23-1.28). This corresponds to one excess death for every 500 PPI users annually, which could translate into thousands of excess deaths per year given the widespread use of PPIs, according to investigators.

They also evaluated associations between PPI use vs. no PPI use in an additional cohort of more than 3 million veterans, and between PPI use vs. no PPI use and no H2 blocker use in more than 2.8 million.

They observed a higher risk for death when comparing PPI use vs. no PPI use (HR = 1.15; 95% CI, 1.14-1.15), and PPI use vs. no PPI and no H2 blockers (HR = 1.23; 95% CI, 1.22-1.24).

The risk for death linked to PPIs was also higher in those without GI conditions: PPIs vs. H2 blockers (HR = 1.24; 95% CI, 1.21-1.27), PPIs vs. no PPIs (HR = 1.19; 95% CI, 1.18-1.2), and PPIs vs. no PPIs and no H2 blockers (HR = 1.22; 95% CI, 1.21-1.23).

PAGE BREAK

Finally, the researchers observed a steady increase in risk for death with a longer duration of exposure among new PPI users. Death risk between PPI and H2 blocker users was not significantly different after 30 days, but after 1 to 2 years, the risk was almost 50% higher in PPI users vs. H2 blocker users.

“A lot of times people get prescribed PPIs for a good medical reason, but then doctors don’t stop it and patients just keep getting refill after refill after refill,” Al-Aly said in the press release. “There needs to be periodic re-assessments as to whether people need to be on these. Most of the time, people aren’t going to need to be on PPIs for a year or two or three.”

The researchers acknowledged that the cohort included mostly older white men, which could limit the generalizability of their findings. They also noted that PPI users were older, sicker and had higher rates of diabetes, hypertension, and cardiovascular disease compared with H2 blocker users, but said the increased risk for death remained after controlling for age and illness.

The researchers concluded that limiting PPI use and length of exposure to medically necessary situations is warranted.

“PPIs save lives,” Al-Aly said in the press release. “If I needed a PPI, I absolutely would take it. But I wouldn’t take it willy-nilly if I didn’t need it. And I would want my doctor to be monitoring me carefully and take me off it the moment it was no longer needed.”

In light of the growing concern about the health risks of PPIs, the AGA recently issued 10 best practice recommendations for long-term PPI use, and emphasized that the benefits likely outweigh the risks when PPIs are prescribed appropriately. In response to the current study, the AGA has also released an interpretation of the data on its website, along with a guide for effectively communicating about the findings with patients.

“While a large data set was used, the causes of death were not reported. Nevertheless, the study is receiving a lot of attention and clinicians can expect to hear from many interested and worried patients and colleagues,” AGA wrote in the guide. – by Adam Leitenberger

Reference:

AGA. A Guide to Conversations About the Latest PPI Research Results. Accessed July 7, 2017. http://www.gastro.org/news_items/a-guide-to-conversations-about-the-latest-ppi-research-results

Disclosures: The researchers report no relevant financial disclosures.

Editor's note: This article was updated on July 7 with additional information from AGA.

    Perspective
    Daniel E. Freedberg

    Daniel E. Freedberg

    Déjà vu: another article appears warning about the potential risks for long-term proton-pump inhibitors. We’ve been here before and we’ll be here again.

    Like previous PPI studies, this study is retrospective, uses a dataset with millions of patient records, finds a weak association between PPIs and the outcome of interest, and lacks a clear mechanism to explain the association. Although the study was carefully conducted, these limitations make it difficult to draw conclusions.

    The key challenge is that PPI users differ markedly from non-users at the time they initiate PPIs. These baseline differences, rather than the PPI itself, are difficult or impossible to completely adjust for. PPI users should be expected to have higher baseline rates of obesity or smoking — variables not captured in this study — and in the degree of severity within disease categories (eg, more severe diabetes), which could bias the findings in a study like this. More fundamentally, a prescription for a PPI appears to reflect exposure to the health care system itself. The more doctors you see, the more likely you are to die (because you are already sick) and to be prescribed a PPI. However, that does not mean the PPI was the cause of your death.

    So, what do I tell my patients about long-term PPIs? If the indication for the PPI is strong, I tell them not to worry. If the indication is weak, I tell them to stop the PPI because polypharmacy is bad practice even if medications have no side effects. This study won’t change that message.

    • Daniel E. Freedberg, MD, MS
    • Assistant Professor of Medicine Columbia University Medical Center

    Disclosures: Freedberg reports no relevant financial disclosures.