The sample size is estimated to detect the difference between the rate of infection in normal (p1=40%) and the expected rate in the pPROM mothers (p2=60%) according to Karat et al.(2006), with a 95% level of confidence (error = 5%), and a study power of 80% (βerror=20%). Using the
for the difference between two proportions (Schlesselman, 1982). Accordingly, the estimated sample size is 107 subjects per group. After adjustment for a dropout rate of20%, the sample size will be 135 women per group.