Table 3

Multiple logistic regression model based on ECG and other clinical characteristics of patients (n = 627a) who come to the emergency department with acute chest pain and acute coronary syndrome (ACS). I-STamp, ST amplitude in lead I; aVF-STamp, ST amplitude in lead aVF; V2-STamp, ST amplitude in lead V2; STamp38, ST amplitude at the end of the third out of eight equal intervals between the ST-J point and the end of the T wave.


Estimate
95% CI

Baseline odds for ACSb
0.0163
0.0073 – 0.0362



Odds ratios


     Age (no. of years above 40)
1.031
1.014 – 1.047
     Hypertension
1.7
1.1 – 2.8
     Angina pectoris ≤ 1 month
4.1
0.97 – 17
     Congestive heart failure
0.48
0.24 – 0.94
     Previous myocardial infarction


          Yes, ≤ 6 months
2.7
1.2 – 6.4
          Yes, > 6 months
2.1
1.2 – 3.8
          No
1.0
-
     Previous CABG
0.23
0.09 – 0.60
     Chest discomfort at presentation
1.9
1.2 – 3.1
     Symptom duration


          0 – 6 h
3.8
2.0 – 7.1
          7 – 12 h
2.8
1.2 – 6.5
          > 12 h
1.0
-
     I-Stamp


          I-STamp > 50 and I-STamp38 > I-Stamp
2.4
0.74 – 7.7
     aVF-Stamp


          aVF-STamp>100 and aVF-STamp38>aVF-Stamp
9.4
2.7 – 33
          aVF-STamp < -100 and aVF-STamp38 < aVF-STamp
4.1
0.72 – 23
          None of above
1.0

     V2-Stamp


          V2-STamp > 200 and V2-STamp38 > V2-Stamp
3.4
1.5 – 7.4
          100 < V2-STamp ≤ 200 and V2-STamp38 > V2-STamp
1.6
0.90 – 2.8
          V2-STamp < -100 and V2-STamp38 < V2-Stamp
2.6
0.54 – 13
          None of above
1.0
-

a Data on at least one of characteristics were missing for 7 (4 with ACS and 3 without ACS) of the original 634 patients.

b Baseline odds for ACS for a 40-year old patient who belongs to the reference category with respect to all other characteristics. The corresponding risk (probability) for ACS can be calculated as Odds/(1+Odds).

Björk et al. BMC Medical Informatics and Decision Making 2006 6:28   doi:10.1186/1472-6947-6-28