The International Classification of Diseases (ICD) is revised approximately every 10 years. New revisions can introduce discontinuities in time series of mortality statistics, presenting challenges for analyzing mortality across ICD revisions. Starting with 1999 data, causes of death will be classified according to the Tenth Revision of the ICD (ICD-10) replacing ICD-9 in use since 1979. The extent of these discontinuities is measured using a "comparability ratio," described in an earlier presentation in this session. This paper presents statistical methods for analyzing differences in cause-specific death rates between revisions. Formulas for computing relative standard errors, confidence intervals and hypotheses tests are discussed. Consideration is given to crude and age-adjusted death rates as well as to cases involving small numbers of events. The presentation will enable public health analysts to recognize the methodological challenges to analyzing death rates across revisions of the ICD; to understand the role of comparability studies and comparability ratios in dealing with these challenges; and to use statistical methods to analyze death rates across ICD revisions under a variety of circumstances.
Learning Objectives: Participants will understand how to apply comparability ratios that permit quantitative comparisons of death rates across revisions of the International Classification of Diseases
Keywords: Death, Mortality
Presenting author's disclosure statement:
Organization/institution whose products or services will be discussed: None
I do not have any significant financial interest/arrangement or affiliation with any organization/institution whose products or services are being discussed in this session.