Issues in reporting epidemiological studies: can't we all agree on confidence intervals?