Differences in Study Results Are Seldom Due to Whether Study Design Is Observational or a Randomized Clinical Trial

(p. A17) The health system would be less burdened if more patients were treated before they require hospitalization, and there are promising therapeutic options that patients can administer themselves at home. This was the subject of a Nov. 19 [2020] hearing before the Senate Homeland Security and Governmental Affairs Committee.

Testimony from the hearing underscored an important issue: Too many doctors have interpreted the term “evidence-based medicine” to mean that the evidence for a treatment must be certain and definitive before it can be given to patients. Because accusing a physician of not being “evidence based” can be a career-damaging allegation, fear of straying from the pack has prevailed, favoring inertia and inaction amid uncertainty about Covid-19 treatments.

For diseases with established treatment options, holding out for certainty may be prudent. But when options are limited and there are safe treatments with evidence for effectiveness, holding out for certainty can be catastrophic. Requiring a high degree of certainty during a crisis may elevate the augustness of medical organizations and appease the sensibilities of medical professionals, but it does nothing for patients who need help.

The penchant for certainty is visible in the frequently updated treatment guidelines for Covid-19 from the National Institutes of Health. These guidelines were developed by scientists around the country, but because of a mentality that is biased toward virtually irrefutable evidence, no distinction is made for treatments with evidence for effectiveness that falls below the mark of certainty. This framework almost certainly has contributed to many avoidable deaths during this pandemic.

. . .

While some health officials dismiss nonrandomized studies, the Cochrane organization, an international leader in evidence-based medicine, published a review of several hundred studies showing that randomized clinical trials and nonrandomized studies of treatments generally yield similar findings. Modern epidemiologic and statistical methods can usually overcome biases inherent in nonrandomized study designs.

For the full commentary, see:

Joseph A. Ladapo. “Too Much Caution Is Killing Covid Patients.” The Wall Street Journal (Wednesday, Nov. 25, 2020): A17.

(Note: ellipsis, and bracketed year, added.)

(Note: the online version of the commentary has the date November 24, 2020, and has the same title as the print version.)

The Cochrane organization review mentioned above is:

Anglemyer, Andrew, Hacsi T. Horvath, and Lisa Bero. “Healthcare Outcomes Assessed with Observational Study Designs Compared with Those Assessed in Randomized Trials.” In Cochrane Database of Systematic Reviews, 2014.

Leave a Reply

Your email address will not be published. Required fields are marked *