Evaluating Medical Literature: Patient Oriented Evidence vs. Disease Oriented Evidence

Data and Metrics Data Literacy

Results of medical studies are often reported in the news. You may have seen headlines like “What vegetable on your dinner plate is killing you?” or “New study shows Medicine X causes memory loss!” You are, very likely and very appropriately, highly skeptical of these headlines but it can be really tricky to sort out the wheat from the chaff (especially when there is so much chaff).

When evaluating medical studies, one way to help you sort it out and prevent wasted time is to look at the outcomes studied. Is this outcome actually meaningful to real people? This is called patient-oriented evidence, or POEs. Often, the outcome studied isn’t as helpful, something called disease-oriented evidence, or DOEs.

POEs are things like mortality, morbidity, and quality of life. Studies that look at POEs are asking questions like “Do people live longer, healthier or happier?”

DOEs are easier to study and are much more common. DOEs are things like imaging, lab or other test results, or physical exam findings. DOEs ask questions like “Is the cholesterol better? Is the blood sugar lower?”

Many times, people try to draw conclusions about patient-oriented outcomes from disease-oriented evidence and this doesn’t always fly. Let’s use cholesterol medications as an example. There are lots of drugs out there to treat high cholesterol. High cholesterol increases the risk of heart attack and stroke. The traditional teaching is the LDL is “bad cholesterol,” so it makes sense to lower that LDL to reduce heart attacks and stroke, right? Sadly, wrong! The class of medications called statins (like atorvastatin or simvastatin) reduces LDL AND lowers the risk of death, heart attack and strokes. Great! Niacin, another medicine used for cholesterol, lowers LDL but DOES NOT reduce the risk of death, heart attack or stroke. Less great! For a long time, people were prescribed niacin to reduce LDL with the belief that if we lowered LDL it would reduce heart attack and stroke. It just didn’t pan out that way when we actually studied if people live longer or have fewer heart attacks and strokes. Lowering LDL by itself doesn’t help people live longer or healthier. This is why it is so important to ask those POEs in our studies and not make assumptions.

One more example: Vitamin E and lung cancer. Lung cancer is unfortunately super common and is among the leading causes of cancer death around the world. Preventing lung cancer is incredibly important. It was hypothesized that antioxidants from supplements could help prevent lung cancer by lowering oxidative damage from free radicals to lung tissues. Vitamin E has been shown to be a good antioxidant (this is disease-oriented evidence). However, Vitamin E does not reduce lung cancer incidence or mortality, but it increases risk of strokes from bleeding in the brain (called hemorrhagic stroke). This DOE (being a good antioxidant) did not correlate with the POE we were looking for (reducing lung cancer).

When that splashy headline hits your news feed, take a quick moment to find out what the study was really looking at it. Is it a POE or a DOE? If it’s a DOE, feel free to ratchet up your skepticism. 😉

Stay safe. Stay well. Stay intellectually curious.

Those Nerdy Girls

Additional LinksPOEs vs DOEs: Farley Library Research Guide