Skip to main content

Table 2 Power for effect size given fixed sample size

From: Impact of nationwide enhanced implementation of best practices in pancreatic cancer care (PACAP-1): a multicenter stepped-wedge cluster randomized controlled trial

Population

N

p0

p1

RD

ICC

Power

Interpretation

25-month study duration (including 5.8-week wash-in period)

 DPCG

2142

0.65

0.550

− 0.100

0.0184

0.8

80% power for true reduction of 10.0%

 DPCG

2142

0.65

0.535

− 0.115

0.0184

0.9

90% power for true reduction of 11.5%

 All NL

4769

0.78

0.714

− 0.066

0.0368

0.8

80% power for true reduction of 6.6%

 All NL

4769

0.78

0.704

− 0.076

0.0368

0.9

90% power for true reduction of 7.6%

 All NL

4769

0.78

0.722

− 0.058

0.0092

0.8

80% power for true reduction of 5.8%

 All NL

4769

0.78

0.712

− 0.068

0.0092

0.9

90% power for true reduction of 6.8%

  1. N sample size, p0 current 1-year mortality, p1 expected 1-year mortality, RD risk difference, ICC intra-cluster correlation coefficient, CAC cluster autocorrelation, DPCG Dutch Pancreatic Cancer Group, NL the Netherlands