Background: Statistical significance between test scores can sometimes be obtained even when the actual difference between the scores is very small. Because of this researchers have recommended that clinical significance also be calculated when reporting results of interventions. The calculation of clinical significance between scores on a particular test takes into account the reliability of the test, as well as normative data from clinical and non-clinical samples. Jacobson and Truax (1991) describe the calculation of the Reliable Change Index (RCI) as a method of calculating this clinical significance.
Result: The RCI for the Edinburgh Postnatal Depression Scale (EPDS) is calculated to be four points. This is the size of difference between two scores which is needed for a clinician to be 95% confident that this change reflects a real change in the individual's mood, and is not likely to be due to measurement error. Using the accepted EPDS cut-off score of 12/13 for major depression (English-speaking women), examples are given of the clinical significance classification of individual's pre-post EPDS scores. A further example of reporting both statistical significance and clinical significance on the EPDS scores of 23 women is given, highlighting the improved information that is obtained when this clinical significance is calculated.
Clinical implications: Statistical significance allows the interpretation of whether women 'on average' improve more in one condition than another. Clinical significance allows the further interpretation of how many women show improvement in each condition. This improvement can be categorised into those who recover, those who improve but are still in the distressed range, those who show no reliable change in their EPDS score, and those who deteriorate. Calculating clinical significance, using the reliable change index and cut-off classification, therefore adds to our understanding of the impact of postnatal depression treatment studies.