Multivariate logistic regression demonstrating effect of covariates on potential diagnostic errors
| Covariate | Adjusted OR (95% CI) | P value |
| Age* | ||
| <2 years | Reference group | |
| 2–5 years | 1.21 (0.60 to 2.45) | 0.51 |
| 6–<12 years | 1.40 (0.69 to 2.85) | 0.99 |
| 12 years and older | 2.22 (1.15 to 4.30) | 0.02 |
| ESI acuity level (L3, L4, L5 compared with L1 and L2) | 2.15 (1.01 to 4.60) | 0.047 |
| Number of inpatient visits in the past 6 months (continuous) | 0.67 (0.45 to 0.98) | 0.04 |
*A linear relationship did not exist between the log of the odds of the outcome (diagnostic error) and the continuous independent variable (age) so could not be included as a continuous variable in logistic regression. Therefore, the age variable was split in four categories: <2 years, 2–5 years, 6–<12 years and ≥12 years. These age categories account for changes in the direction of the slope of the plot of the log odds of likely diagnostic error vs age and have clinical relevance from a developmental standpoint.
ESI, Emergency Severity Index.