Predictors of poor knowledge of AMI symptoms by multiple logistic regression (n=1804)
| Variables | Odds ratio (95% CI) | P values |
| Age group | ||
| 65 – 74 | 1 | |
| ≥ 75 | 1.47 (1.16 to 1.86) | 0.001 |
| Sex | ||
| Male | 1.47 (1.16 to 1.87) | 0.001 |
| Female | 1 | |
| Education level | ||
| Primary or below | 3.69 (2.55 to 5.34) | < 0.001 |
| Secondary | 1.67 (1.16 to 2.40) | 0.005 |
| Bachelor degree or above | 1 | |
| History of heart disease | ||
| Yes | 1 | |
| No | 1.87 (1.24 to 2.83) | 0.003 |
*‘Poor knowledge about AMI symptoms’ was defined as failure to identify at least five of the common symptoms experienced by the local patients with AMI (ie, sweating, shortness of breath, dizziness, chest pain, stomach ache, neck pain, shoulder pain, arm pain and nausea).
†Hosmer and Lemeshow test showed acceptable goodness-of-fit (P=0.990).
‡Age group, sex, marital status, education level, occupation, living status, monthly household income, history of diabetes, history of heart diseases, history of hypertension, history of stroke, history of high cholesterol and previous history of AMI were added to the full model. Insignificant factors were removed according to backward elimination procedure one by one. Only significant factors remained in the final model.