Correcting for misclassification and selection effects in estimating net survival in clinical trials