Statistical vs. Clinical Significance Statistical

Statistical vs. Clinical Significance

Statistical and clinical significance are both important terms within nursing research. Although some researchers tend to use the terms interchangeably, they are not the same concept. Indeed, some researchers are against favoring the one over the other. According to Kain (2005), the importance of statistical significance lies in the fact of academic validity, or at least the appearance of such validity. The reason for this is the calculation of probability (P). The boundary of 0.05 determines whether a study in question will enjoy further research. Indeed, the author also states that the 0.05 probability level is also used to determine whether a drug can be marketed as "effective."

Statistical significance is concerned with quantification. The probability levels of certain effects are measured. However, clearly this is not sufficient to determine the true effect of a drug or procedure upon a population. When physicians misinterpret statistically significant results as indicating clinical significance, problems begin to arise. With a large sample, for example, there could be a statistically significant effect between groups, but the actual impact of treatment is minimal. Hence statistical significance does not necessarily indicate that there would be clinical significance as well.

Clinical significance is concerned with the practical clinical effects of a treatment upon a population. If a significant proportion of the population responds well to the treatment, the clinical significance is likely to be high. Davidson (1994) notes that significantly different effects between groups are indicated as a relative risk reduction. The impact of therapy upon a group is referred to as "number needed to treat." This refers to the percentage of patients in a group who responded to a treatment. This is the number of patients who need to be treated for one person to show improvement within a randomized sample. The p value therefore determines the likelihood of the difference indicated being due to chance. Number needed to treat (NNT) on the other represents the possible clinical impact of the study. In other words, p values correspond to statistical significance, while NNT corresponds to clinical significance. In clinical trials, statistical validity reflects the theoretical basis of the study, with hypotheses being formulated and quantified in terms of likelihood. Clinical significance is concerned with the practical outcome of trials, and with the results of actual treatment and how this relates to the hypotheses that are proven or void.

2.

In nursing practice, both statistical and clinical significance play an important role in research. In practice, however, it is clinical significance that should have the greatest impact upon nursing practice. Clinical significance provides actual data from research conducted to determine such effects. It concerns the outcome of trials, while statistical significance is more concerned with determining new research and the likelihood of success before trials have been conducted.

Indeed, Davidson notes that an advantage of NNT is the format of its results -- resulting from clinical trials upon human subjects, they can be easily related to the human patients that nurses work with. A focus on NNT help clinicians to make more critical decisions when it comes to administering therapy.