A public health specialist aims to bring about a policy on attendance to pre-natal clinics. So, he wants to compare the rates of prematurity in infants born to women who attend pre-natal clinics (hypothesized to be P1=0.25) with non-attenders (hypothesized to be P2=0.40). What sample size will he need if he assumes to be 0.01, power at 80% and response rate at 80%?