Dynamic evaluation of short-term prognosis after myocardial infarction

Authors

  • Philip Hougaard,

    1. Statistical Research Unit and Institute of Mathematical Statistics, University of Copenhagen, Institute of Mathematical Statistics, Universitetsparken 5, DK-2100 Copenhagen Ø, Denmark
    Search for more papers by this author
  • Erling Birk Madsen

    1. Department of Cardiology, Glostrup Hospital, Department of Medicine B, Rigshospitalet, Blegdamsvej 9, DK-2100 Copenhagen Ø, Denmark
    Search for more papers by this author

Abstract

The risk of cardiac arrest, cardiogenic shock or death in the first 44 days after myocardial infarction is evaluated using a version of the Cox regression model. The (time-dependent) covariates include complications that have occurred, making it possible to give an individual assessment of prognosis that can be updated each day.

Low risk patients can be discharged from hospital after a few days, whereas high risk patients must remain in hospital until their risk has fallen to an acceptable level. For the former group of patients it is possible to reduce the number of days in hospital without increased mortality and for the latter group mortality can be reduced by longer hospitalization.

Ancillary