Myocardial infarction: one year mortality was decreased slightly with thrombolysis.

Clinical bottom line (level 2b)

  1. A tenth of patients with MI who were not given thrombolysis will be dead at one year, whereas just under a tenth of patients given thrombolysis were.
  2. Less than 1% of patients given thrombolysis had an intracranial haemorrhage associated with the therapy.
  3. A twentieth of patients given thrombolysis had a major bleed associated with therapy.
Selker et al: Annals of Internal Medicine 1997; 127 (7): 538-556
Expires March 2003

The study

Retrospective cohort study with objective outcomes, adjusted for confounding factors, validated in an independent set of patients.

Setting: 107 hospitals, USA

1648 patients (aged median 58 years, ?% male) acute myocardial infarction and ST-segment elevation of at least 1 mm (0.1 mV) in two or more leads on the ECG

Excluded if
  • >75 years old
  • onset of chest pain or other ischaemic symptoms not within 9 hours of presentation
  • systolic blood pressure >190 mmHg




Pooled logistic regression was used to construct the predictive instruments.

100% followed for 1 year
Outcomes studied:
  • one year mortality- no thrombolysis
  • one year mortality- thrombolysis
  • cardiac arrest- no thrombolysis
  • cardiac arrest- thrombolysis
  • thrombolysis-related intracranial haemorrhage
  • thrombolysis-related major bleed

  • Data was taken from 13 clinical trials and registries on patients with myocardial infarction, and entered into the Thrombolytic Predictive Instrument Database.
  • There were 3263 patients in the development data set (a third of the database). The patients were divided randomly.
  • A separate predictive instrument was developed for each of the outcomes: acute mortality (30 days) from acute infarction if given and not given thrombolysis; long-term (1 year) mortality if given and not given thrombolysis; cardiac arrest within 48 hours after first ECG if given and not given thrombolysis; intracranial haemorrhage with thrombolysis; bleeding necessitates transfusion with thrombolysis. Variable were entered onto the ECG. 30 day mortality predictive instrument: age <40, >75 or in between; systolic blood pressure <60 mmHg, >190 mmHg or in between; diabetes (1) or not (0); heart rate <70 beats/min (0), >120 beats/min (50) or in between (rate-70); St-segment elevation; Q-waves without ST elevation; location of MI; right bundle branch block (1); impact of thrombolysis. 1 year mortality predictive instrument: <60 years (60), >75 (age) or in between (75); heart rate <70 (0), >120 (50) or in between (rate-70); anterior MI (1); number of contiguous Q waves; right bundle branch block. Cardiac arrest instrument: age <35, >70 or in between; systolic BP <80 mmHg (80), >100 mmHg (100) or in between (systolic BP); size of MI; sum of ST elevation in all leads; corrected QT interval adjusted to time from onset of chest pain to ECG; impact of thrombolysis. Intracranial haemorrhage instrument: age <35 (0), >75 (40) or in between (age-35); excess pulse blood pressure. Major bleed instrument: age <30 (30), >75 (75) or in between (age); female (0) or male (1); history of hypertension (1); systolic blood pressure < 60 mmHg (60), >190 mmHg (190) or in between; heart rate with systolic blood pressure; sex with standard dose of thrombolytic agent adjusted for weight; standard dose of thrombolytic agents adjusted for weight.

The evidence

outcome time to outcome number of patients/total number %
(95% CI)
one year mortality- no thrombolysis 1 year 180/1648 10.9%
(9.42% to 12.4%)
one year mortality- thrombolysis 1 year 157/1648 9.50%
(8.10% to 10.0%)
cardiac arrest- no thrombolysis 1 year 89/1648 5.40%
(4.30% to 6.50%)
cardiac arrest- thrombolysis 1 year 48/1648 2.91%
(2.10% to 3.72%)
thrombolysis-related intracranial haemorrhage 1 year 10/1648 0.61%
(0.23% to 0.98%)
thrombolysis-related major bleed 1 year 82/1648 4.98%
(3.93% to 6.03%)

Comments

  1. The validation of a clincal prediction guide was performed on a split-sample analysis of the data. The model is complex, and not presented in this CAT.
  2. There were no clear categories of patients (ie. high risk/ low risk) determined by the predictive instrument, so it was not very effective at determining individual risk of death or cardiac arrest.

Citation

  1. Selker HP, Griffith JL, Beshansky JR, et al: Patient-specific predictions of outcomes in myocardial infarction for real-time emergency use: A thrombolytic predictive instrument. Annals of Internal Medicine 1997; 127 (7): 538-556
Contributor: Clare Wotton and Bob Phillips, January 2000
Reviewer:

Clinical Question.
Patient myocardial infarction
Intervention or Exposure prevelance
Outcome need for thrombolytics, death and side effects