Goldacre Michael J, Duncan Marie E, Cook-Mozaffari Paula, Griffith Myfanwy
Unit of Health-Care Epidemiology, Department of Public Health, Institute of Health Sciences, University of Oxford, Oxford OX3 7LF.
J Public Health Med. 2003 Sep;25(3):249-53. doi: 10.1093/pubmed/fdg058.
Until recently, national coding and analysis of routine mortality statistics in most countries included only underlying cause of death. There were changes in the rules for selection and coding of underlying cause in England in 1984 and 1993. We report on trends in mortality rates in an English region from 1979 to 1998, comparing multiple-cause and underlying-cause coded rates, for individual diseases that were affected by coding changes. Among many others, these include pneumonia, venous thromboembolism, heart failure, respiratory distress syndrome, tuberculosis, diabetes, dementia, alcohol and drug abuse, epilepsy, multiple sclerosis, stroke, asthma, peptic ulcer, appendicitis, and cancers of the breast, colon and prostate. Comparisons over time of mortality rates based on underlying cause alone will be misleading when the time-period crosses years in which rules changed for selecting underlying cause.
