Skip to main content
. 2005 Aug 11;5:26. doi: 10.1186/1472-6947-5-26

Table 2.

Comparison of pretest probability estimates for acute coronary syndrome using two techniques in a validation population.

Category of Pretest Probability
Very Low (<2%) Low (2 to <10%) Moderate (10 to 30%) High (>30%) Indeterminate*

Attribute Matching
Number with ACS† 28 189 250 9 125
Total in category‡ 1670 2953 1453 28 955
Prevalence of ACS 1.7% 6.4% 17.2% 32.1% 13.1%
% of all patients assessed 23.7% 41.8% 20.6% 0.4% 13.5%
Logistic Regression
Number with ACS† 4 59 272 266 NA
Total in category‡ 244 1356 3813 1646 NA
Prevalence of ACS 1.6% 4.4% 7.1% 16.2% NA
% of all patients assessed 3.5% 19.2% 54.0% 23.3% NA

*Estimates from attribute matching with match size <50. Does not apply to logistic regression.

†Total number of patients with ACS = 601

‡Total number patients in validation study = 7059

Abbreviations: ACS acute coronary syndrome.