Chest pain: clinical features suggestive of MI and ECG risk factors increased the risk of death.

Clinical bottom line (level 1b)

  1. A quarter of patients with symptoms suggestive of myocardial infarction were dead within 5 years.
  2. Patients were at increased risk of dying if:
    • signs of acute ischaemia on ECG (NNF = 3 for 5 years) , or evidence of old ischaemia on ECG (NNF = 4 for 5 years)
    • history of smoking (NNF = 6 for 5 years) , diabetes (NNF = 6 for 5 years) , congestive heart failure (NNF = 7 for 5 years) hypertension (NNF = 19 for 5 years)
    • presentation with acute severe heart failure (NNF = 6 for 5 years)
  3. Patients with normal ECGs (NNF = 8 for 5 years) , or who were female (NNF = 14 for 5 years) were less likely to die.
Herlitz et al: Journal of Internal Medicine 1998; 243: 41-48
Expires March 2003

The study

Inception cohort study with objective outcomes, adjusted for confounding factors, not validated in an independent set of patients.

Setting: emergency department, university hospital, Sweden

5241 patients (aged ~50% aged 49-75 years, 55% male) with chest pain or symptoms suggestive of myocardial infarction

Factors studied:
  • death
  • smoker
  • diabetes
  • acute severe heart failure on presentation
  • history of congestive heart failure
  • history of MI
  • unspecific symptoms on presentation
  • hypertension
  • age (increase per year added)
  • female
  • ECG on admission: signs of acute ischaemia
  • pathological but no sign of acute ischaemia
  • presence of q-waves
  • normal




    • Multivariate analysis was used to adjust for confounding factors.

    98% followed for 5 years
    Outcomes studied:
  • death

    • risk factors:
      • clinical features on presentation
      • ECG: normal- pathological signs; pathological: old infarctions, bundle branch block, non-specific ST changes; acute ischaemia- ST elevation 2 mm or more in leads V1-V4 or 1 mm or more in aVL, aVF, I, II, III, V5-V6; ST depression > 1mm, T-wave inversion Q wave 2 mm or more deep in at least 2 leads)
    • outcome: death (from Swedish National Registry of Death: 39% autopsied)

    The evidence

    outcome time to outcome number of patients/total number %
    (95% CI)
    death 5 years 1345/5241 26%
    (24% to 27%)

    prognostic factor for
    death     		time to outcome adjusted RR
    (95% CI)    		 NNF+
    (95% CI)
    smoker 5 years 1.64
    (1.42 to 1.90)     		6
    (4 to 9)
    diabetes 5 years 1.60
    (1.35 to 1.90)     		6
    (4 to 11)
    acute severe heart failure on presentation 5 years 1.63
    (1.35 to 2.06)     		6
    (4 to 11)
    history of congestive heart failure 5 years 1.59
    (1.36 to 1.84)     		7
    (5 to 11)
    history of MI 5 years 1.28
    (1.11 to 1.48)     		14
    (8 to 35)
    unspecific symptoms on presentation 5 years 1.23
    (1.03 to 1.47)     		17
    (8 to 130)
    hypertension 5 years 1.20
    (1.05 to 1.37)     		19
    (11 to 77)
    age (increase per year added) 5 years 1.07
    (1.06 to 1.08)     		55
    (48 to 64)
    female 5 years 0.73
    (0.64 to 0.84)     		-14
    (-24 to -11)
    ECG on admission: signs of acute ischaemia 5 years 2.24
    (1.83 to 2.75)     		3
    (2 to 5)
    pathological but no sign of acute ischaemia 5 years 1.90
    (1.57 to 2.29)     		4
    (3 to 7)
    presence of q-waves 5 years 1.47
    (1.12 to 1.92)     		8
    (4 to 32)
    normal 5 years 0.49
    (0.41 to 0.58)     		-8
    (-9 to -7)

    Comments

    1. Few patients received thrombolysis or aspirin at the time of the study.

    Citation

    1. Herlitz J, Karlson BW, Lindqvist J, et al: predictors and mode of death over 5 years amongst patients admitted to the emergency department with acute chest pain or other symptoms raising suspicion of acute myocardial infarction. Journal of Internal Medicine 1998; 243: 41-48
    Search Terms: hand search
    Contributor: Chris Ball and Clare Wotton, Unknown Month 2000
    Reviewer:

    Clinical Question.
    Patient chest pain
    Intervention or Exposure risk factors
    Outcome death