Univariate and multivariate regression analyses of factors associated with major hemorrhage diagnostic errors in 947 injured patients
| Variable | Factors associated with missed MH diagnosis | Factors associated with MH overdiagnosis | ||||||
| Univariate | Multivariate* | Univariate | Multivariate* | |||||
| OR (95% CI) | P value | OR (95% CI) | P value | OR (95% CI) | P value | OR (95% CI) | P value | |
| Patient factors | ||||||||
| Age | 0.989 (0.967 to 1.01) | 0.27 | 0.984 (0.963 to 1.00) | 0.12 | ||||
| Sex (female) | 1.62 (0.682 to 3.42) | 0.24 | 2.21 (1.05 to 4.36) | 0.03* | ||||
| MOI (penetrating) | 2.45 (1.30 to 4.74) | 0.006 | 2.37 (1.24 to 4.65) | 0.010 | 1.77 (0.970 to 3.25) | 0.06* | ||
| GCS | 0.999 (0.929 to 1.08) | 0.98 | 1.03 (0.953 to 1.12) | 0.52 | ||||
| SBP | 0.993 (0.985 to 1.00) | 0.09 | 0.997 (0.988 to 1.01) | 0.56 | 0.983 (0.976 to 0.990) | <0.001 | 0.986 (0.978 to 0.994) | <0.001 |
| HR | 1.00 (0.989 to 1.01) | 0.98 | 1.01 (0.996 to 1.02) | 0.22 | ||||
| Polytrauma | 1.11 (0.890 to 1.36) | 0.33 | 1.41 (1.16 to 1.71) | <0.001 | 1.33 (1.08 to 1.63) | 0.006 | ||
| Bleeding chest injury | 3.58 (1.55 to 7.53) | 0.001 | 2.36 (0.963 to 5.33) | 0.05 | 2.69 (1.13 to 5.75) | 0.02 | 1.28 (0.510 to 2.91) | 0.57 |
| Bleeding abdominal injury | 4.90 (2.18 to 10.2) | <0.001 | 4.04 (1.74 to 8.70) | <0.001 | 1.72 (0.579 to 4.15) | 0.27 | ||
| Unstable pelvis | 1.40 (0.222 to 4.86) | 0.65 | 2.01 (0.469 to 5.92) | 0.26 | ||||
| Peripheral bleeding injury | 2.16 (0.338 to 7.73) | 0.31 | 0.909 (0.0502 to 4.49) | 0.93 | ||||
| Major vascular injury | 5.25 (2.74 to 9.98) | <0.001 | † | 1.82 (0.863 to 3.57) | 0.09 | † | ||
| Clinician factors | ||||||||
| <1 year experience | 1.00 | 1.00 | ||||||
| 1–5 years’ experience | 0.644 (0.190 to 1.66) | 0.41 | 0.538 (0.159 to 1.37) | 0.25 | ||||
| 5–10 years’ experience | 1.95 (0.452 to 5.80) | 0.29 | 1.05 (0.167 to 3.64) | 0.95 | ||||
| >10 years’ experience | 1.63 (0.452 to 5.80) | 0.52 | 0.658 (0.0364 to 3.22) | 0.68 | ||||
| Base specialty: EM | 1.00 | 1.00 | ||||||
| Base specialty: anesthetics | 1.14 (0.575 to 2.20) | 0.70 | Could not fit regression | |||||
| Base specialty: ICM | 0.904 (0.211 to 2.68) | 0.87 | Could not fit regression | |||||
| Clinician diagnostic uncertainty | 1.41 (0.524 to 3.21) | 0.45 | 4.57 (2.31 to 8.69) | <0.001 | 3.67 (1.77 to 7.30) | <0.001 | ||
| Environment factors | ||||||||
| Nightshift | 1.09 (0.584 to 2.08) | 0.78 | 1.18 (0.645 to 2.18) | 0.6 | ||||
Referents in the model (not listed in the table): for female sex was male; for penetrating MOI was blunt; for bleeding chest injuries, was no chest bleeding; for bleeding abdominal injuries, was no abdominal bleeding; for unstable pelvis, was stable pelvis; for peripheral bleeding injuries, was no peripheral bleeding; for major vascular injury, was no major vascular injury; for clinician diagnostic uncertainty, was certainty; for nightshift was dayshift.
*Multivariate variables were limited to four due to event rate of 41 patients for the missed MH diagnosis analysis, and 45 patients for the MH overdiagnosis analysis. For this reason (on the overdiagnosis analysis), although female sex and penetrating MOI had p<0.1 on univariate, they were not included in the multivariate model because included variables were more significant on univariate.
†Major vascular bleeding injury excluded from multivariate due to collinearity with other bleeding injuries (chest, abdomen, unstable pelvis, peripheral), as it combined these injuries in one category.
EM, emergency medicine; GCS, Glasgow Coma Scale; ICM, intensive care medicine; MH, major hemorrhage; MOI, mechanism of injury; Polytrauma, number of Abbreviated Injury Score (AIS) body regions injured; SBP, systolic blood pressure.