The article by Gardner and Altman emphasizes the importance of using confidence intervals (CIs) over P values in medical research. They argue that the overemphasis on hypothesis testing and statistical significance has led to a shift away from interpreting the magnitude of differences between groups, which is more clinically relevant. Confidence intervals provide a range of values that are plausible for the population parameter, offering more informative and precise estimates than simple dichotomies of significant or non-significant results. The authors suggest that major findings in medical studies should be presented with full statistical information, including sample estimates, confidence intervals, test statistics, and P values. They recommend using 95% CIs for most studies, as they balance precision and confidence. The article also provides methods for calculating CIs for means, proportions, and their differences, and discusses the limitations of using P values alone. Confidence intervals are illustrated through examples, such as comparing systolic blood pressure between diabetic and non-diabetic patients, and comparing treatment responses. The article concludes by advocating for the use of confidence intervals as the standard method for presenting statistical results in medical research, emphasizing their ability to provide a more comprehensive and clinically useful assessment of study findings.The article by Gardner and Altman emphasizes the importance of using confidence intervals (CIs) over P values in medical research. They argue that the overemphasis on hypothesis testing and statistical significance has led to a shift away from interpreting the magnitude of differences between groups, which is more clinically relevant. Confidence intervals provide a range of values that are plausible for the population parameter, offering more informative and precise estimates than simple dichotomies of significant or non-significant results. The authors suggest that major findings in medical studies should be presented with full statistical information, including sample estimates, confidence intervals, test statistics, and P values. They recommend using 95% CIs for most studies, as they balance precision and confidence. The article also provides methods for calculating CIs for means, proportions, and their differences, and discusses the limitations of using P values alone. Confidence intervals are illustrated through examples, such as comparing systolic blood pressure between diabetic and non-diabetic patients, and comparing treatment responses. The article concludes by advocating for the use of confidence intervals as the standard method for presenting statistical results in medical research, emphasizing their ability to provide a more comprehensive and clinically useful assessment of study findings.