Editorial

PPI Risk-Benefits: A Practical Approach to Patient Discussions

Edward V. Loftus Jr.

In looking at this month’s cover story, one of the big takeaways for me is that much of the data on which current proton pump inhibitor risks are based comes from “outcomes research studies” using very large administrative databases, usually insurance company databases. In using large databases, you’re not often getting the full story.

In these databases, physicians submit diagnoses, office visits or link prescriptions. You can construct a story from this information, but if you think about it, a lot of that accuracy lies with the accuracy of the billing providers’ reports. There are tens of thousands of patients in them, but they’re messy. It’s like the famous saying from John Godfrey Saxe (often misattributed to Otto von Bismark): “Laws, like sausages, cease to inspire respect in proportion as we know how they are made.”

In this case, you don’t necessarily want to look too closely at how outcomes research is done with these large databases. Unless you do a validation study where you test your algorithm using these billing codes and validated that against actual medical records, you’re never quite sure how accurate these algorithms, and therefore overall study results, are.

Sheer Volume ORs

When some of these huge databases show an odds ratio of 1.2, it is statistically significant because of the sheer number of patients. But is that clinically significant? How accurate is it really?

In our cover story, Michael F. Vaezi, MD, PhD, brought up issues with all of the confounders and biases that can occur in these retrospective observational studies. Maybe what is driving the association is not a direct association between PPIs and the various outcomes. It could be that the exposure and outcome are indirectly linked to another variable that’s to blame.

When you see these low odds ratios of less than two or three, you should take the result with a grain of salt and we have to improve our communication of that to our patients. We must improve our translation of mathematical language to one our patients can understand. When we tell patients, “Your risk of X is 20% higher than the general population,” that often gets misinterpreted by the patient as, “Wow, I have a 20% risk of that side effect.” Relative risks always sound scarier than absolute risks so consider discussing absolute risks. Or if you discuss the relative risks, remind them of the baseline risks instead of just saying “elevated risk.”

Ultimately, the patient must decide. If your patient has a good reason to be on the PPI in the first place — such as bad GERD that led to esophagitis — then the benefit they’re getting from the PPI far outweighs the risk.

Origin of Prescription

There are valid points that these drugs probably are overprescribed. A diagnosis of non-ulcer dyspepsia or a stress ulcer might warrant a PPI prescription, but time passes and nobody bothers to review the chart and discontinue the medication. Sometimes patients are left on these drugs months or years longer than they need.

Whenever a patient asks me if he or she should stop taking a PPI, I ask, “What was the reason you went on it in the first place?” If they can’t remember, there may be good reason to at least attempt weaning from the PPI. You don’t want to rely on a medication unless you have to.

One of the problems occurs when you try to stop a PPI cold turkey; there’s often a rebound hyper secretion of acid as they are coming off the medication. That rebound convinces the patient the PPI is necessary. The patient says, “But I got reflux right away!”

I tell them to try a gradual approach. If they’re on PPI twice a day, they should titrate down to once daily for a day and back to twice the next day, alternating between the two doses for a few weeks. If that is successful, then drop it down to daily. If they are down to daily, try every other day. Don’t stop cold turkey.

Physicians can also suggest trying H2 blockers as a backup, such as ranitidine or famotidine, because these medications are not tagged with all the controversy of the PPIs.

Lifestyle, Weight

William D. Chey, MD, brought up lifestyle modifications in the cover story. There are so many classic things we can tell people to avoid such as carbonated beverages or too much caffeine, chocolate or alcohol and not eating right before you go to bed. Those are all important and shouldn’t be ignored.

Additionally, the cover story suggests obesity and overweight are driving a lot of acid reflux and probably driving a lot of PPI use. We need to get back to the basics and fundamentals when we talk to patients about reflux management.

We as physicians too easily pull the trigger on medications because that is what we do. We have to remind ourselves and the patients of these other factors.

As a tip, though, I never put it on the patient; I use myself as an example. I say, “Whenever I start getting heartburn again, it’s a signal for me that I need to work on losing 5 or 10 pounds.” I don’t want to sound accusatory about a patient’s weight, but it’s also true. When I do gain weight, one of the first places I feel it is in my esophagus. The reflux is a reminder from our bodies about how lifestyle management affects our GI health.

Pitching it to the patient in those terms makes it more realistic and palatable. Talk to your patients in numbers and ideas that are easy to grasp and we can best treat reflux, using PPIs when most effective but hopefully not more than necessary.

How do you talk to your patients about PPIs? What tips do you have for non-pharmaceutical management of reflux? As always, let me know your thoughts in our Twitter conversations through @EdwardLoftus2 and @HealioGastro.

Disclosure: Loftus reports consulting for Janssen, Takeda, AbbVie, UCB Pharma, Amgen, Pfizer, Celgene, Gilead, Eli Lilly, CVS Caremark, Celltrion Healthcare, and Napo Pharma; and research support from Janssen, Takeda, AbbVie, UCB Pharma, Amgen, Pfizer, Celgene, Gilead, Robarts Clinical Trials, MedImmune, Allergan, Genentech, and Seres Therapeutics.

Edward V. Loftus Jr.

In looking at this month’s cover story, one of the big takeaways for me is that much of the data on which current proton pump inhibitor risks are based comes from “outcomes research studies” using very large administrative databases, usually insurance company databases. In using large databases, you’re not often getting the full story.

In these databases, physicians submit diagnoses, office visits or link prescriptions. You can construct a story from this information, but if you think about it, a lot of that accuracy lies with the accuracy of the billing providers’ reports. There are tens of thousands of patients in them, but they’re messy. It’s like the famous saying from John Godfrey Saxe (often misattributed to Otto von Bismark): “Laws, like sausages, cease to inspire respect in proportion as we know how they are made.”

In this case, you don’t necessarily want to look too closely at how outcomes research is done with these large databases. Unless you do a validation study where you test your algorithm using these billing codes and validated that against actual medical records, you’re never quite sure how accurate these algorithms, and therefore overall study results, are.

Sheer Volume ORs

When some of these huge databases show an odds ratio of 1.2, it is statistically significant because of the sheer number of patients. But is that clinically significant? How accurate is it really?

In our cover story, Michael F. Vaezi, MD, PhD, brought up issues with all of the confounders and biases that can occur in these retrospective observational studies. Maybe what is driving the association is not a direct association between PPIs and the various outcomes. It could be that the exposure and outcome are indirectly linked to another variable that’s to blame.

When you see these low odds ratios of less than two or three, you should take the result with a grain of salt and we have to improve our communication of that to our patients. We must improve our translation of mathematical language to one our patients can understand. When we tell patients, “Your risk of X is 20% higher than the general population,” that often gets misinterpreted by the patient as, “Wow, I have a 20% risk of that side effect.” Relative risks always sound scarier than absolute risks so consider discussing absolute risks. Or if you discuss the relative risks, remind them of the baseline risks instead of just saying “elevated risk.”

PAGE BREAK

Ultimately, the patient must decide. If your patient has a good reason to be on the PPI in the first place — such as bad GERD that led to esophagitis — then the benefit they’re getting from the PPI far outweighs the risk.

Origin of Prescription

There are valid points that these drugs probably are overprescribed. A diagnosis of non-ulcer dyspepsia or a stress ulcer might warrant a PPI prescription, but time passes and nobody bothers to review the chart and discontinue the medication. Sometimes patients are left on these drugs months or years longer than they need.

Whenever a patient asks me if he or she should stop taking a PPI, I ask, “What was the reason you went on it in the first place?” If they can’t remember, there may be good reason to at least attempt weaning from the PPI. You don’t want to rely on a medication unless you have to.

One of the problems occurs when you try to stop a PPI cold turkey; there’s often a rebound hyper secretion of acid as they are coming off the medication. That rebound convinces the patient the PPI is necessary. The patient says, “But I got reflux right away!”

I tell them to try a gradual approach. If they’re on PPI twice a day, they should titrate down to once daily for a day and back to twice the next day, alternating between the two doses for a few weeks. If that is successful, then drop it down to daily. If they are down to daily, try every other day. Don’t stop cold turkey.

Physicians can also suggest trying H2 blockers as a backup, such as ranitidine or famotidine, because these medications are not tagged with all the controversy of the PPIs.

Lifestyle, Weight

William D. Chey, MD, brought up lifestyle modifications in the cover story. There are so many classic things we can tell people to avoid such as carbonated beverages or too much caffeine, chocolate or alcohol and not eating right before you go to bed. Those are all important and shouldn’t be ignored.

Additionally, the cover story suggests obesity and overweight are driving a lot of acid reflux and probably driving a lot of PPI use. We need to get back to the basics and fundamentals when we talk to patients about reflux management.

We as physicians too easily pull the trigger on medications because that is what we do. We have to remind ourselves and the patients of these other factors.

PAGE BREAK

As a tip, though, I never put it on the patient; I use myself as an example. I say, “Whenever I start getting heartburn again, it’s a signal for me that I need to work on losing 5 or 10 pounds.” I don’t want to sound accusatory about a patient’s weight, but it’s also true. When I do gain weight, one of the first places I feel it is in my esophagus. The reflux is a reminder from our bodies about how lifestyle management affects our GI health.

Pitching it to the patient in those terms makes it more realistic and palatable. Talk to your patients in numbers and ideas that are easy to grasp and we can best treat reflux, using PPIs when most effective but hopefully not more than necessary.

How do you talk to your patients about PPIs? What tips do you have for non-pharmaceutical management of reflux? As always, let me know your thoughts in our Twitter conversations through @EdwardLoftus2 and @HealioGastro.

Disclosure: Loftus reports consulting for Janssen, Takeda, AbbVie, UCB Pharma, Amgen, Pfizer, Celgene, Gilead, Eli Lilly, CVS Caremark, Celltrion Healthcare, and Napo Pharma; and research support from Janssen, Takeda, AbbVie, UCB Pharma, Amgen, Pfizer, Celgene, Gilead, Robarts Clinical Trials, MedImmune, Allergan, Genentech, and Seres Therapeutics.