- Prognostic risk
- Many prognostic models for predicting risk of
- Mortality
- Unfavourable outcome
- Although useful for more reliably informing relatives of the anticipated outcome, it should be recognized that a prediction model is a population- based estimate and yields no more than a risk estimate with a substantial element of uncertainty, which can be expressed in the confidence interval.
- For this reason, the greatest application of prognostic analysis is not at the level of the individual patient, but rather at the ‘group’ level for quantifying and stratifying the severity of brain injury, as a reference for evaluating quality of care and for stratification and covariate adjustment in clinical trials
- It is worthy to add that recent studies are showing the superiority in early MR imaging as a prognostic tool in TBI, but there is still uncertainty about its discriminative predictive value and which acute phase lesion patterns correlate with long- term outcome
- The two largest prognostic models
- For any TBI patient regardless of severity
- MRC CRASH trial prognosis for a patient with isolated head injury, bilateral reactive pupils, CT evidence of petechial hemorrhage, SAH, cistern effacement/obliteration of third ventricle and midline shift (but without non-evacuated hematoma) in a high income country
- For patient with moderate and severe TBI;
- Hypoxia hypotension and hyperglycaemia has poor outcomes
CRASH (CRASH calculator)
GCS | Age | Risk of 14-day mortality | Risk of unfavorable outcome at 6 months |
Mild (13-14) | ≤40 years | 14.3–16.4% | 38–41.9% |
ㅤ | 75-year-old | 66.2–69.8% | 87.7–89.4% |
Moderate (9-12) | ≤40 years | 18.7–27.2% | 45.9–58.1% |
ㅤ | 75-year-old | 73.1–81.5% | 90.8–94.2% |
Severe (3-8) | ≤40 years | 30.5–49.7% | 62–78.8% |
ㅤ | 75-year-old | 83.8–92.1% | 95–97.7% |
Made with Bullet