Simple methods of determining confidence intervals for functions of estimates in published results