Important statistical techniques in medical research

Statistical techniques in medical education (Credit: Sean/Flickr)

Main Point:

More elaborate statistical training is important for clinicians so that they can do research, and elaborate/interpret the medical findings better, according to a recent study.

Published in:

PLoS ONE

Study Further:

Researchers, in the present study, worked on the statistical techniques used in one of the most prestigious journals, Journal of the American Medical Association (JAMA), and found that the frequency and complexity of statistical reporting have increased in the past two decades. However, researchers have reported that very few internal medicine residents (<50% of 277) have proper knowledge of statistics and they are unable to properly interpret the statistical results. Moreover, there is significant lack of knowledge in advanced statistics such as Kaplan Meier and regression analysis.

They found that almost all of the papers published in the journal in the last two decades have presented some form of statistical techniques and complex statistical techniques are present in nearly half of the papers published in JAMA in the last decade.

This finding clearly shows that medical schools and post-graduate training programs must have to impart statistical training especially some of the complex statistical techniques such as sensitivity analysis, survival analysis, Wilcoxon Rank and multi-level modeling as the use of these techniques have been increased since 1990.

Following table shows the list of important statistical techniques:

Statistical measures and methods in JAMA articles published in 1990, 2000, and 2010 (Credit: Lauren D. Arnold et al./PLoS ONE)

Research Suggestion:

Researchers wrote, “One journal alone does not serve as a “gold standard” by which to judge medical education needs.” So, you can do the same study with other journals as many of the related studies are older or have smaller number of samples.

One study presented that longer research papers can result in more citations, so you can work on the related concept that whether there is a correlation between the complicated statistical techniques in the paper and the chances of more citations.

Reference:

Arnold LD, Braganza M, Salih R, Colditz GA (2013). Statistical Trends in the Journal of the American Medical Association and Implications for Training across the Continuum of Medical Education PLoS ONE DOI: 10.1371/journal.pone.0077301

Usman Zafar Paracha

Usman Zafar Paracha is Assistant Professor, Pharmaceutics, in Hajvery University, Lahore, Pakistan.