Assessing the reporting of categorised quantitative variables in observational epidemiological studies