Citation

Choi IC, Park JW, Jung JY, Kim DK, Kwak YH, Suh D, Lee SU. J. Korean Med. Sci. 2020; 35(16): e147.

Affiliation

Department of Emergency Medicine, Samsung Medical Center, Seoul, Korea.

Copyright

(Copyright © 2020, Korean Academy of Medical Science)

DOI

10.3346/jkms.2020.35.e147

PMID

32329261

Abstract

On behalf of myself and my co-authors, we appreciate Dr. Ali for his interest in our article and important comments on our article's interpretation. In this study, rock climbing equipment was the only statistically significant risk factor for significant kids café-related injuries as a result of multivariable logistic regression. Because the confidence interval (CI) does not contain the null of 1.0, rock climbing equipment is a statistically significant factor by definition. However, the CI was very wide (aOR, 14.94; 95% CI, 1.51–147.72). In other words, we are 95% confident that the true population OR is between 1.51 and 147.72. The width of the interval implies the amount of variability in the estimate and is a reflection of the sample size from which the estimate was drawn. In principle, the larger the sample size, the less variability and the narrower the CI, leading to a more precise estimate. Conversely, with smaller sample sizes, the CI is generally wider, indicative of a more variable and less precise estimate.1 In this study, we concluded that rock climbing equipment is risky for significant injuries in kids cafés based on the results of statistical analysis. We agree with Dr. Ali's opinion that OR interpretation must be prudently made keeping in view the CI. In this study, small sample size leads to the wide CI and a less precise estimate. We did not conclude that rock climbing equipment is about 15 fold riskier than the reference factor because this is a less precise estimate because of the small sample size. However, we could conclude that rock climbing equipment is risky for significant injuries in kids cafés because the CI did not contain the null value. Further studies with a larger sample size may be needed to support our conclusion and make a meaningful clinical inference.

Language: en