What impact will implementation of ICD-10 have on

What impact will implementation of ICD-10 have on mortality statistics by cause in the elderly? Clare Griffiths Health and Care Division Office for National Statistics

Mortality Data • Collected for England Wales by ONS • from civil registration of deaths • info on cause of death comes from medical certificate of cause • certain cases referred to a coroner - info sent direct to registrar

Classification of cause of death • Using International Classification of Diseases • Since 1979 ICD-9 • Since 2001 ICD-10 • Data not yet published in ICD-10 • Classification important for assessing health of population • Change particularly affects elderly

ICD-10 for Mortality in England Wales • From January 2001 • more codes - detail • move conditions between chapters • change underlying cause of death selection rules • all these produce artefacts

Underlying cause of death • …’the disease or [external cause of] injury which initiated the train of morbid events leading directly to death…. ’ • For public health • preventing death and illness leading to death

Cause of death certificate direct cause due to Ia carcinomatosis b prostate cancer c other contributing cause II hypertension ………………………. . . Underlying cause of death - prostate cancer

General Principle and Rule 3 ICD-9 / ICD-10 General Principle “When one or more conditions is entered on the certificate, the condition entered alone on the lowest used line of Part I should be selected only if it could give rise to all the conditions entered above it” Rule 3 “If the condition selected by the General Principle or by Rule 1 or Rule 2 is obviously a direct consequence of another reported condition, whether in Part I or Part II, select this primary condition. ”

Cause of death I a bronchopneumonia b c II Alzheimer’s disease UCD bronchopneumonia Alzheimer’s disease ICD-9 ICD-10

Age standardised mortality rates in Women in England & Wales 1968 to 1999 by selected chapters of ICD-8 and ICD-9

Bridge coding • Identify artefacts due to classification change • independent coding to ICD-9 and ICD-10 • all deaths registered in 1999 • original text and amendments • routine coding procedures/ rules – US Software, clerical for inquests

Why bridge code? • Measure effect of change in classification • Allow adjustment of rates • Interpret time trends in cause-specific mortality in routine statistics

Bridge coding outputs • comparability ratios for each cause – eg IHD deaths in ICD-10/ IHD deaths in ICD-9 • stratified by age, sex? – do CRs vary significantly? – confidence limits • ‘adjusted rates’ for old trend data – ‘new’ baselines and targets

IHD mortality rates in women, E&W, 1999: by age, ICD-9, ICD 10 and comparability ratio

Cancer mortality rates in women, E&W, 1999: by age, ICD-9, ICD 10 and comparability ratio

Stroke mortality rates in men, E&W, 1999: by age, ICD-9, ICD 10 and comparability ratio

Pneumonia mortality rates in men, E&W, 1999: by age, ICD-9, ICD-10 and comparability ratio

Pneumonia ASMR women, E&W 1980 -1999 as published and adjusted for changes in classification

Other causes • Suicide is based on verdict and unlikely to change • Injuries (including falls): changes in the axis of classification

Effects and Plans • Cause-specific mortality rates will change • Comparability ratios available on the National Statistics website from Spring 2002 • Standard shortlists, Our Healthier Nation • New baselines and targets

Any questions? • Queries on ICD-10 & deaths data – ICD 10. mortality@ons. gov. uk • Queries on other ICD issues – who@ons. gov. uk • More details in Health Statistics Quarterly No 8 www. statistics. gov. uk/downloads/theme_health/HSQ 8 Book. pd f • Forthcoming in HSQ 13 February 2002