Featured ArticleStatistical power and sample size calculations: A primer for pediatric surgeons☆,☆☆
Section snippets
Methods
We present a practical 5-step approach for the pediatric surgeon to determine the appropriate sample sizes needed to detect clinically and scientifically meaningful results based on sufficient statistical power. These steps can be applied to any prospective or retrospective research study scenario regardless of the nature of the data. The 5 steps are:
- 1.
Define the outcome variable of interest and comparison groups and formulate the clinical or scientific hypothesis. The outcome of interest may be
Discussion
We have demonstrated how to perform statistical power and sample size calculations with our 5-step approach using examples in pediatric surgery. Regardless of the research study, the nature of the outcome or exposure data, or the study design, our 5-step approach can be applied to perform power calculations correctly to determine the minimum required sample sizes to ensure sufficient power and meaningful interpretation of the study data.
It may be challenging to obtain sample sizes as large as
Conclusions
A properly performed and described power and sample size calculation improves the quality of manuscripts submitted to surgical journals, IRB proposals, and grant applications. The clinical hypotheses and the nature of the data guide the appropriate statistical test to be used in sample size calculations. Our article can be utilized as a practical guide for pediatric surgeons when designing a clinical study or laboratory investigation to ensure that the sample sizes will provide the sufficient
References (21)
- et al.
The power of a paired t-test with a covariate
Soc Sci Res
(2015) - et al.
Adjusting for multiple testing—when and how?
J Clin Epidemiol
(2001) Adjust for multiple comparisons? It's not that simple
Ann Thorac Surg
(2016)- et al.
How large a sample?
- et al.
Statistical power, sample size, and their reporting in randomized controlled trials
JAMA
(1994) - et al.
CONSORT compliance in surgical randomized trials: are we there yet? A systematic review
Ann Surg
(2013) - et al.
Trends in worldwide volume and methodological quality of surgical randomized controlled trials
Ann Surg
(2013) - et al.
Comparison of randomized controlled trial registry entries and content of reports in surgery journals
Ann Surg
(2013) - et al.
GPOWER: a general power analysis program
Behav Res Methods Instrum Comput
(1996) - et al.
G*Power 3: a flexible statistical power analysis for the social, behavioral, and biomedical sciences
Behav Res Methods
(2007)
Cited by (20)
Correlation of Strabismus Surgical Outcomes Graded by Goal-Determined Metric With Patient Satisfaction Survey
2024, American Journal of OphthalmologyStatistical power and sample size calculations for time-to-event analysis
2023, Journal of Thoracic and Cardiovascular SurgeryCitation Excerpt :We present a practical 5-step approach for cardiovascular and thoracic surgeons to determine the appropriate sample sizes needed to detect clinically and scientifically meaningful results based on sufficient statistical power.3
Qualitative features of esophageal fluorescence angiography and anastomotic outcomes in children
2023, Journal of Pediatric SurgeryCitation Excerpt :A two-tailed 5% alpha was considered statistically significant. A total sample size of 55 procedures (15 with poor anastomotic outcome and 40 with good outcome) provided 80% power for detecting a clinically meaningful difference in esophageal anastomotic perfusion (EAP) score between procedures with a PAO versus no PAO (mean difference of 2.7 and pooled standard deviation of 3.25; minimum detectable standardized difference of 0.86), based on Student's t-test assuming a two-tailed 5% alpha [29]. Statistical power calculations were performed using G*Power software (University of Dusseldorf, Germany).
Brainstem auditory physiology in children with listening difficulties
2023, Hearing ResearchHow many cauliflower seedlings are necessary to estimate experimental precision statistics reliably?
2023, Scientia HorticulturaeCitation Excerpt :Cochran (1977) proposed an optimum allocation of samples based on a cost function, and Desu and Raghavarao (1990) showed study power can be altered by sample size. Therefore, several studies have used the power analysis approach to define sample size, especially in medical research (Bujang and Adnan, 2016; Blaise et al., 2016; Staffa and Zurakowski, 2020). However, Rothman and Greenland (2018) highlighted the advantages of defining sample size based on precision rather than through power calculations.
Effects of Lacticaseibacillus rhamnosus GG supplementation, via food and non-food matrices, on children's health promotion: A scoping review
2022, Food Research InternationalCitation Excerpt :Another important aspect to be discussed is that statistical power relates to effect size in such a way that great group differences contribute to its increase (Boukrina et al., 2020). In this context, understanding considerations regarding sample size and statistical power will significantly contribute to improve the quality of published scientific articles (Staffa & Zurakowski, 2020). Nevertheless, Schnadower et al. (2018) verified that daily consumption of capsules containing L. rhamnosus GG (2 × 1010 cfu per 5 days) in a clinical trial involving 970 children with acute gastroenteritis showed no significant results.