Background: Two areas of uncertainty about routine statistics for mortality after hospital admission for myocardial infarction (MI) or stroke are i) whether most deaths occur in the admission episode itself rather than after discharge, and ii) whether most deaths are certified on death certificates as, respectively, MI or stroke.
Methods: Use of linked hospital and mortality statistics to analyse the time, place and certified cause of death in people aged 35-74 after admission for MI or stroke.
Results: Of 7,964 deaths within a year of admission for MI, 5,686 (71.4%) occurred within 30 days of admission. Of these, 4,856 (85.4%) occurred during the initial hospital admission. Of 7,070 deaths within a year of admission for stroke, 4,905 (69.4%) were within 30 days, and 4,509 (91.9%) of these occurred during the initial admission. As expected, deaths at longer intervals than 30 days occurred mainly after discharge. Of deaths within 30 days of MI and stroke, 85.2% and 80.0%, respectively, were certified with MI or stroke as the underlying cause of death.
Conclusion: In-hospital death rates alone, calculated without record linkage to death certificates, would have identified most deaths that occurred within 30 days of admission. Nonetheless, linkage added to completeness of ascertainment even within this time period. Data without linkage are unreliable in identifying deaths at longer time intervals. Routine mortality statistics for MI and stroke, as the underlying cause, reliably included most deaths that occurred within 30 days of admission for each respective disease.