Introduction

A t-test is an appropriate test for normally distributed continuous variables (like pulse and systolic blood pressure).

A chi squared test is an appropriate test for dichotomous variables (like alive or dead).

The authors have presented confidence intervals of the mean difference. A confidence interval of a difference that does not cross from positive to negative indicates statistical significance.

Confidence intervals are more useful than p-values because, as well as indicating whether there is statistical significance, they indicate the range of uncertainty.

Clinical Significance

A result can be statistically significant without being of clinical significance. For instance, is a statistically significant reduction in pain score of 3/100 of clinical importance? When looking at study outcomes, it is necessary to always consider the clinical significance of the data.