These are proven facts. Well there at least is a statistical correlation which proves this….right? There has been a lot of information in the press about the possibility of developing a heart attack while on prilosec or nexium therapy. If you look at this screen shot on this topic, you can see the hype surrounding it.
When this study came out, my facebook account and email was flooded with messages from patients and friends asking me about this study and whether they should stop taking these medications.
I am in favor of minimizing unnecessary drugs for my patients. If there is not a clinical indication to be on a treatment, then we should try and get you off of it. Proton pump inhibitors or PPIs have helped thousands of patients with treatment of bleeding ulcers, reflux symptoms or indigestion, and overall made people feel better.
So should you continue taking these medications in light of this study?
Here is what I have to say:
What did the study ACTUALLY find
The risk of developing a heart attack was 16%. This sounds scary – but when you look at the data itself it is not as impressive as it may seem. The rate of a heart attack or myocardial infarction (MI) would require treatment of 4000 patients with PPIs to produce just one additional MI. That is a lot of people that need treatment for the possibility of one heart attack.
The study methods involve taking data from thousands of patients, mixing it up together, and using a computer model to suggest an association. 2.9 MILLION patients were data mined – which means they were put into this computer to find an association. This is analogous to taking the population of Dallas, seeing which patients had heart-attacks, and finding most of them had taken a bubble-bath in the evening of their heart-attack, therefore concluding bubble-baths cause an MI.
Studies of this type cannot prove causality and certainly do not define a mechanism.
In the study, the patients who were on PPI therapy may be sicker than those who were not
No one really knows why those patients who were on PPI therapy had this association. Maybe they were sicker overall? Maybe they had other risk factors for MI that were not addressed such as obesity, diabetes, smoking – all risk factors for both heartburn and a heart attack?
The point is that there are numerous queries about this patient population that we just don’t know. This raises the question as to whether there were confounding factors involved, limitations the authors of this study acknowledge themselves.
Drug dosages were not analyzed
Also of contention is the study authors could not stratify which drug dosages are at higher risk. Is it all doses of PPI or only the high-doses? Is it patients who take it everyday or as needed?
Duration of medication use was also not addressed – if I only take a PPI for 3 months, is my risk higher versus taking it for one year? What if i stop my medications for a couple of weeks and restart again – does my risk change?
Additionally, not all of the drugs within the class of PPIs had this association; two of the five drugs studied did NOT show a correlation. Why was that?
These are all questions in which the study could not answer, leaving me to wonder how accurate the data might be.
All patients, regardless of PPI therapy, need to improve their overall health
Regardless if you have heartburn or reflux, or are on PPI therapy, living a heart-healthy lifestyle is best for your overall health. This means exercising, eating right, routine health maintenance, checking your lipid profile, etc. Mitigating risk factors for a heart attack will not only improve your overall health, but will also help with your reflux symptoms.
There is nothing in this study which would change my clinical practice
For me, if I feel like my patients clinically need a PPI, I would still recommend this to them. If I feel like they do not need a PPI, I would recommend them to stop it, solely based on the fact I truly believe the less medications the better, and NOT based on this study.
Very commonly in the medical field, studies come out that seem to show a correlation and immediately a causative effect is assumed. This is a false fallacy. Don’t forget the Nicholas Cage effect; correlation does NOT necessarily equal causation.