A new study study suggests drugs like ibuprofen increase the risk of sudden death by 30%. What does

All medical interventions provide both benefit and risk. It is up to an individual patient to ascertain whether the benefit sufficiently exceeds the risk to make the intervention worthwhile. For larger health care systems and insurance carriers, the risks and benefits become exaggerated over an entire population and the cost of the intervention must be factored into what value it achieves. For instance, if a certain drug helps an individual prevent a heart attack in 1/1000 people who take it, but gives nuisance side effects like belly pain and head ache, then that patient might decide not to take the drug given its tiny chance of helping. But for an insurance company, if a million of their clients take the drug, and thus 1000 prevent a heart attack, such a drug may be a good investment depending on its cost. Similarly, if a drug helps someone feel better in 300/1000 instances, but there is a 3/1000 chance of having a bad outcome from the drug, then a patient may weigh the risks and benefits and decide to take the drug despite the small risk. However, for the insurance company, the cost of the bad outcome may be very high, and they may discourage use of the drug. This is the difference between individual health outcomes and population health, and both have to be considered with assessing the value of an intervention.

Two recent studies illustrate this, and both of the studies, which flooded the news sources in the past week, were colored by the fact that their risk and benefit were inflated by relative numbers. Only by breaking their results down to absolute risks and benefit, and then put into a format that is comprehensible, can we grasp their impact on both individual patients and health care systems. In this blog, we will talk about anti-inflammatory drugs, many of which are over the counter. In the next blog we will discuss the novel cholesterol drug Repatha, which is claimed to reduce heart attack and stroke by 20% without significant side effects but comes with a price tag of $14,000 a year.

With the narcotic epidemic dominating the airways, many doctors are advocating taking patients off narcotics for their pain and using other drugs, such as Tylenol and anti-inflammatories (called NSAIDS) like ibuprofen. While Tylenol is typically a very safe pain medicine, in many cases it simply does not work. And NSAIDs do provide more pain relief but they can be harmful. A study that came out this week from Denmark hit home about the dangers of NSAIDs. The study found that the use of these drugs increased the chance of dying from cardiac arrest (the heart suddenly stops) by 30%, a shocking result that has resulted in many doctors and health care advocates suggesting that NSAIDs should be more heavily restricted.

While the study results are frightening, their conclusions are relayed in deceptive relative numbers. The bigger question is of all people who take these drugs, how many, out of 1000, will die of cardiac arrest? The article does not answer the question, but it does supply clues. In Denmark, 338 people die of cardiac arrest in a year who were taking NSAIDs 30 days before their death. Denmark is a country of 5.6 million people, and it is estimated that about half the people in Denmark get at least one prescription for NSAIDs. That means that 2.8 million (minus 338) took NSAIDs without dying, or one person has a cardiac arrest out of 10,000 people who take NSAIDs, too small a number to even portray with a BRCT. Also, we cannot be sure that there is cause and effect. For instance, perhaps people with accelerating heart disease who are getting more chest pain use these drugs to help the pain, and it is not the drugs that kill them but rather the underlying heart disease. While the study tries to account for this, it cannot possibly eliminate people who have heart disease at the time of their cardiac arrest but did not have documented heart disease prior.

The bottom line is that the risk of dying of cardiac arrest by taking these drugs is tiny, far lower than the exaggerated relative risk of 30% would indicate. In fact, according to stories in the media, many doctors and organizations want to restrict NSAID use due to this study, something that would seem rash in light of the absolute data. However, that is not to say that such drugs are not without other risks. A BMJ study from September 2016 showed that people who use some NSAIDs have a 20% higher rate of congestive heart failure (CHF) than non-users, though again with much smaller absolute risk, approximately 1-2/1000 additional hospital stays for CHF than non-users, especially among people with known prior CHF. A Lancet article from 2013 demonstrates other risks. In this meta-analysis of many studies, they divided people into two groups: those with a high risk or pre-existing heart disease, and those with lower risk of heart disease. As the BRCT shows, there was an increase of 3/1000 heart attacks or strokes in people using NSAIDs (higher in Celebrex and ibuprofen), and an increase of 16/1000 additional serious bleeds with drugs such as Motrin and Naprosyn (lower with Celebrex) compared to nonusers, as shown in the second BRCT.

Thus, for both individuals and insurance companies, NSAIDS need to be seriously evaluated if there is pre-existing heart disease, bleeding risk, or even kidney disease. Again, although the risks for the individual are small and should be weighed against the pain relief obtained, the risk for insurance companies may be much larger and could prompt a suggestion for alternative treatments. These discussions are best held not by using the exaggerated numbers published in recent studies, but by have patient-centered discussions using BRCTs like the ones published in this blog.