Breaking data: Today we released a new study of 53K Americans revealing that use of proton pump inhibitors (eg. Prilosec, Prevacid, Nexium) is associated w/ a 2-4x risk of #COVID19. In this thread I explain why we did this study, what we found, what it may mean, & key limitations 
2. Here is a link to the full peer-reviewed study published online in #AmJGastro: journals.lww.com/ajg/Documents/…. Also, the @AmCollegeGastro released this fact sheet with answers to FAQs: webfiles.gi.org/links/media/AC…. Both documents are more effective than tweets to convey the full results.
3. Let's start with biological plausibility. There's a reason we have stomach acid; it helps sterilize food by killing pathogens. We've known for years that acid can rapidly inactivate certain viruses. For example, this 2004 study shows pH<5 inactivates SARS-CoV-1.
4. This is important because #coronavirus sheds into saliva in >90% of patients (see below), meaning it can be easily swallowed. If it encounters a stomach with low acid, as might occur when taking proton pump inhibitors (PPIs), then a larger viral load may enter the intestines.
5. What happens next is key: recent evidence shows that #SARSCoV2 can rapidly invade and replicate within the cells lining the intestinal tract, as demonstrated in the study, below. It does this by hijacking the ACE-2 receptor which is widely present throughout the gut.
6. There is also evidence that ~50% of patients with #COVID19 have evidence of viral RNA in their stool. I was part of a study early on in the pandemic, below, revealing stool positivity and high prevalence of diarrhea, suggesting the virus infects the gut & clears slowly.
7. Could this all mean that strong acid-blockers like PPIs increase risk for #COVID19? Well before the pandemic, it was already established from meta-analysis that PPIs can increase the risk of GI infections (see study in pic), so this isn't a new idea. What's new is COVID.
9. Before discussing the results: We know that people using PPIs might be different from those who aren't. They might be older, or have more illnesses, etc. So, we adjusted our analyses for potential confounders, including demographic, socioeconomic, and comorbidity variables.
10. We also worried that people might have started a PPI *because* they had #COVID19, possibly to treat COVID-related gut symptoms. For that reason, we didn't count among PPI users those who only started the med after their diagnosis; that would have inflated the effect of PPIs.
11. We found PPIs were associated with a *dose-response* risk of #COVID19, meaning higher the dose, higher the risk. Twice daily PPI=3.7x risk. Once daily or less=2.2x risk. The effect persisted for long vs. short-term use. In contrast, less powerful H2RAs revealed no risk.
12. In addition to PPI use, we found that males, current smokers, non-Hispanic blacks, & Latinxs were significantly more likely to report #COVID19. I mention these findings because they are consistent with previous research and thus support the generalizability of our dataset.
13. Before discussing the significance of these findings, it's important to note limitations. This is a large, hypothesis-driven study, but it's not a prospective, randomized trial. There is always risk of residual confounding. Our study shows an association, but *not* causation.
14. The takeaway is that PPI use, particularly the common but non-approved twice-daily dose, may increase risk of #COVID19. Always worth considering whether twice-daily is needed, particularly for those especially vulnerable to severe disease (eg. elderly or comorbid patients)
15. Moreover, meta-analysis reveals that twice-daily PPI does not offer benefits over once-daily dosing for acid reflux (pic), although individual results vary. As with any med, lowest effective dose should be used. H2RAs may also be considered as an alternative (when possible).
16. For patients using PPIs who are concerned about #COVID19, best way to reduce personal risk is to practice regular hand washing, social distancing, and masking. It's worth emphasizing that these practices will have a much greater impact on personal risk than PPI dosing.
17. If nothing else, the discussion about PPI risk could save lives simply by reminding users to closely adhere with public health guidelines given the potential increased risk conferred by PPIs. If you're on a PPI, take special care to do the right things that matter most.
18. If you are on PPIs, this study does *not* mean you must stop. As always, the decision about whether, when, and how to modify PPIs dosing should be based on a thoughtful assessment of the risk-benefit ratio for individual patients made in partnership with your clinician.
19. Finally, some may have heard about other research that famotidine (Pepcid), an H2RA, may be *protective* against #COVID19. It's thought that the Pepcid effect, if real, is unrelated to acid suppression & results from antiviral properties of famotidine. We'll see in time.
