Johns Hopkins Study finds 41% of studies prone to bias

Advertisement

According to a release from Johns Hopkins Children’s Center, lead author Michael Crocetti, M.D., M.P.H., with colleagues, have performed a review of 146 trials and have found that these pediatric clinical trials are prone to bias.

Bias, in this sense, is defined as anything that produces variations in research findings. These variations are unexpected and are usually controlled for to eliminate them. The recent study by Johns Hopkins Children’s Center reviewed clinical trials in pediatrics to see how well the trials controlled for potential bias which can affect outcomes. If pediatricians then rely on the biased outcomes, that could pose problems.

41% of these studies had improper randomization techniques or those techniques were not described adequately for transparency. Of all the trails, the ones that most clearly showed this potential for bias were industry-funded trails. These were six times more likely to have biased randomization than government-funded or non-profit funded trials.

Further, 57% of the trials reviewed showed improper use of masking techniques that ensure the researchers nor the participants would know who was getting a real treatment versus a placebo. This is important as researchers may see only what they wish they see in the results if they know ahead of time a particular participant has had a treatment.

Advertisement

Dr. Crocetti recommends that physicians and other readers of these studies use critical reading skills. Ask questions about the process the findings were derived from. Are bias-eliminating techniques adequately described and used? Does the data match the findings? These are common sense questions that doctors should use before implementing recommended treatments.

The study reviewed clinical trials using the Cochrane Collaboration tool. This tool helps identify potential bias along six aspects, including Sequence generation; Allocation concealment; Blinding of participants; personnel and outcome; Incomplete outcome data; Selective outcome reporting; and Other sources of bias. All of these areas may produce bias in a clinical trial.

Dr. Crocetti is an Associate Professor and specializes in General Pediatrics and Adolescent Medicine.

Study co-authors are Roberta Scherer, Ph.D., (Johns Hopkins Bloomberg School of Public Health) and Diane Amin, PA-C, of Hopkins.

Hopkin’s Children Center treats about 90,000 children a year and offers a comprehensive pediatric medical program. For more information, please visit: http://www.hopkinschildrens.org

Advertisement