Difference of the means | Standard deviation | Power |
|---|
1.0 | 3.0 | 99% |
3.5 | 95% |
4.0 | 89% |
4.5 | 81% |
1.1 | 3.0 | >99% |
3.5 | 98% |
4.0 | 94% |
4.5 | 87% |
1.2 | 3.0 | >99% |
3.5 | 99% |
4.0 | 97% |
4.5 | 92% |
1.3 | 3.0 | >99% |
3.5 | >99% |
4.0 | 99% |
4.5 | 96% |
- The primary hypothesis for power calculation implied that the mean number of prescribed drugs per patient would increase during hospital stay. Relying on preliminary investigations into the distribution pattern of prescriptions at admission to Essen University Hospital (mean number of prescriptions ± standard deviation: 7 ± 3) and assuming correlations of r = 0.2 to r = 0.7 between the mean prescription numbers at admission and discharge, the standard deviation of the difference between them was deduced to take on values between 3.0 and 4.5. The difference between the means itself was assumed to be at least 1. Based on these assumptions the primary hypothesis was tested in varying scenarios using matched-pair signed-rank test with a significance level of 5%. With a set sample size of n = 180 a study power of more than 80% was achieved in each tested scenario. To account for a drop-out rate of up to 10% the needed number of participants was determined to be 200.
