Lessons from the Medical Field – Correlation isn't Causation

Recently, I read a Medscape article regarding an interesting topic – the failure of niacin to reduce cardiovascular risk. I don't normally try to aggregate and redistribute information from other websites, but this time, I'll make an exception, because it's highly informative for us.

Lessons from the Medical Field The HPS2-THRIVE study, recently published in the New England Journal of Medicine has called into stark relief the failure of niacin to reduce cardiovascular risk and its concerning increase in serious adverse events.¹ It looks like niacin's days as an antihyperlipidemic are numbered.

The results of this study lead Medscape contributor John Mandrola, MD to state the following five lessons learned, which I'll summarize:²

	1.	Association does not equal causation – Apparently this research, as well as others, have found that increased HDL levels (improved with niacin) does not necessarily correlate with cardiac health. Rather, it is a marker associated with cardiac health and not a cause of improved cardiac health.
	2.	Be careful with surrogate markers – Changing lab values doesn't necessarily mean you are improving patient health.

	3.	Pills do not confer heart-health – not even vitamins – Studies have repeatedly shown that vitamins do not reduce adverse cardiac events. Dr. Mandrola asserts that the reliance on pills distracts physicians from helping patients make healthy life choices.
	4.	Do no harm – We need to keep in mind that taking any medication has the potential for harm. For example, taking blood pressure medication increases the risk of falls due to hypotension.
	5.	Learn from mistakes – Physicians should learn from their mistakes, while trying to avoid repeating them.

Tonight's Premier Lecture is
The Next Generation of
Wound Care: Understanding Placental Membranes
Lawrence Lavery, DPM, MPH

Now you may or may not agree with the opinions expressed by Dr. Mandrola or his position on modern medical treatments, but I find his comments eye-opening for the podiatric community for several reasons. First, for those out there that feel the "medical doctors" (the MDs and DOs) have all the answers, think again. As with all things in medicine, appropriately designed and executed research remains the single best way to know if our patient care is safe and effective. "This works well in my hands" is not an appropriate level of evidence. We have moved beyond the "doctor knows best" stage of medicine. Ask yourself this question: Do my treatments follow the best evidence available, coupled with my experience, and my patients' needs? If your answer is yes, then you're practicing evidence-based medical practice. If not, then watch out. You may be the next niacin.

The lesson that correlation is not causation is one that all of us in the medical profession would do well to understand and heed. Take a look at the graph below from the website Spurious Correlations.³ According to the graph, U.S. spending on science increases the number of suicides. So, science leads to suicide?

I think it's pretty obvious that US spending on space and technology doesn't actually lead to increased suicides. However, if one were to take this somewhat tongue-in-cheek data literally, it would be simple to make this logical error. Consider, though, other fallacies of this nature, such as the vaccinations-causing-autism argument. Some have speculated that the increased incidence of autism is due to greater numbers of children being vaccinated. Add to this a falsified research study in the journal Lancet (subsequently retracted), and the damage is done. In reality, though, it is more likely that these two factors are not related at all and that the perceived increase in autism is more likely due to improved screening and diagnoses.

On a more mundane note we have to remain cognizant of this issue when reading and interpreting research studies. Since many of the orthopedic and podiatric studies are of questionable methodologic quality, it is easy to have a study that demonstrates a supposed cause when, in reality, only a simple correlation exists. One example that jumps out is the explanation for how total contact casts (TCC) heal diabetic neuropathic foot ulcers. The classic explanation is that the cast allows redistribution of weight from the foot to the leg, with researchers stating this as the reason for healing (i.e. cast correlates with healing). However, in reality, the improved healing rates are most likely caused by forcing patient compliance. In fact, a study by Katz and associates found no difference in healing rates between TCC and the instant TCC (a cast boot rendered irremovable).⁴ The authors explain this is a result of improved patient compliance. As this study demonstrates, there may be a correlation between the TCC and healing, but it is not the cause of that healing. Other situations like this abound in medicine.

Finally, the last lesson, learning from our mistakes, is perhaps the most important. No matter your level of training or years in practice, we as humans learn better from our mistakes than our successes. Because of this, we need to push to create an environment of safety and openness when it comes to medical mistakes. Physicians should be able to make errors (within reason) and learn from those errors without fear of retaliation. In a litigious country like the United States, it is anyone's guess how this will occur (more likely never to occur). In the case of niacin, I am awaiting the television commercials from lawyers asking for class action lawsuits against those companies that distributed the drug. We must be able to learn from our errors without retribution in order to harm less patients in the long term. For now, we must beware of the lessons learned from the unfortunate niacin situation.

Best wishes,

Jarrod Shapiro, DPM
PRESENT Practice Perfect Editor
[email protected]

###

References:

Landray M, et al. Effects of Extended-Release Niacin with Laropiprant in High-Risk Patients, N England J Med. 2014 Jul;371(3):203-212.
Mandrola M. Five Lessons From the Niacin Failure. Medscape. July 21, 2014. Last accessed July 30, 2014.
Spurious Correlations website. https://www.tylervigen.com/. Last accessed July 30, 2014.
Katz I, Harlana, Miranda-Palma B, et al. A Randomized Trial of Two Irremovable Off-Loading Devices in the Management of Plantar Neuropathic Diabetic Foot Ulcers. Diabetes Care. 2005 Mar; 28(3):555-559.