Sample and setting
A calculation of the required sample size was based on the power analysis with respect to Visual Analogue Scale pain, because this is an intervention study concerning postoperative pain management. A mean of 3.5 and standard deviation (SD) of 2.4 were expected. A clinically-significant difference of 30% and statistically-significant level of 0.05 were expected. With a power of 80%, a sample size of 83 patients for each group (n = 166) was calculated as being appropriate (Burns & Grove, 2005). A pilot study was conducted with 10 patients. After the pilot study, changes were made to the pain assessment so that patients assessed both pain intensity and pain distress while resting, during deep breathing, and when shifting position.