Estimation after blinded sample size reassessment