April 25, 2010

What is High Quality Research?

Patients (and health professionals) are inundated with information overload regarding "research studies" that prove this or prove that on TV, magazines, radio, etc.

Who do you trust? How can you determine if a given study is worthwhile while dismissing others?

Well, it unfortunately takes some homework...

1) Determine WHO did the research. Clearly, research published by Harvard University is more "trustworthy" than research done by XYZ (XYZ being an institution you have never heard of before).

2) What journal was it published in? Not all journals are created equal. First, make sure the journal is peer-reviewed. What that means is that experts in a given field read and determine if the research is worthwhile before publishing it. Some journals do not have this vetting process and pretty much publish anything that crosses their desk. Second, the more reputable the journal, the more rigorous the review process before the research is published.

Highly regarded peer-reviewed journals include:
• New England Journal of Medicine
• Annals of (insert medical specialty)
• Archives of (insert medical speciality)
• Journals produced by national medical organizations such as Otolaryngology-Head & Neck Surgery published the American Academy of Otolaryngology-Head & Neck Surgery

3) What kind of research is it? A randomized, prospective, placebo-controlled, double-blinded research is the best quality. Anything less has the potential to provide conclusions that may not be entirely correct. Let's talk about each of these research study components:

Randomized: Subjects are selected at random. For example, if you are doing a study on pain control, the results may be different if you only select patients who are healthy without any medical problems versus patients who have chronic pain or recently underwent surgery.

Prospective: The study starts when you begin administering the intervention or drug versus looking at historical data (hindsight is 20/20, but not a good thing when you are trying to prove something without bias).

Placebo-Controlled: This means that there is a separate "control" group which is given a sham intervention or a sugar pill not meant to help/improve the problem being studied. Why is this necessary? Because over 20% of the time, patients given this placebo report improvement... which means a given intervention or drug must demonstrate improvement above and beyond this placebo effect to be considered authentic.

Double-Blinded: This term means that both the patient and doctor have no idea whether the intervention is a placebo or the actual drug/intervention. Bias may occur if not double-blinded. For example, what if the doctor is being paid millions of dollars for a positive outcome of the research. With this kind of monetary influence, the results may become skewed towards a positive outcome.

Unfortunately, it is the rare research that incorporates all of these elements for one reason or another. Most research studies have only a few, but not all these elements. The less elements a given research contains, the less rigorous it is considered and the results considered less valid.

In conclusion, not all research studies are created equal. Some are highly trust-worthy and results are beyond question valid. Other research studies are almost worthless.

Don't automatically trust what you hear which is what our mothers' have told us all along. Check the info out yourself or ask where's the proof (and than validate it).
Fauquier blog
Fauquier ENT

Dr. Christopher Chang is a private practice otolaryngology, head & neck surgeon specializing in the treatment of problems related to the ear, nose, and throat. Located in Warrenton, VA about 45 minutes west of Washington DC, he also provides inhalant allergy testing/treatment, hearing tests, and dispenses hearing aids.


Banner Map

Pediatric Neck Masses

Adult Neck Mass Workup