Article
Can standardised comprehensive assessment of patients with traumatic brain injury lead to better outcome prediction?
Kann eine standardisierte, umfassende Evaluation von SHT Patienten zu einer besseren Berechnung des Outcomes führen?
Search Medline for
Authors
-
Fares Komboz
- Universitätsmedizin Göttingen, Neurochirurgische Klinik, Göttingen, Deutschland
-
Hiba Douja Chehade
- Universitätsmedizin Göttingen, Neurochirurgische Klinik, Göttingen, Deutschland
-
Dorothee Mielke
- Universitätsmedizin Göttingen, Neurochirurgische Klinik, Göttingen, Deutschland
-
Veit Rohde
- Universitätsmedizin Göttingen, Neurochirurgische Klinik, Göttingen, Deutschland
-
Tammam Abboud
- Universitätsmedizin Göttingen, Neurochirurgische Klinik, Göttingen, Deutschland
| Published: | May 25, 2022 |
|---|
Outline
Text
Objective: In patients with TBI, accurate outcome prediction is crucial in the process of decision-making and managing ICU resources. GCS is the gold standard in predicting clinical outcome in patients with TBI. In the last decade, the CRASH score was presented as a new prognostic model. It reflects more neurological findings than those in the GCS. The IMPACT prognostic model emerged as a comprehensive score for clinical outcome prediction in TBI patients and additionally takes lab findings into account. The aim of this retrospective study was to validate evaluating these scores as prognostic tools in a large cohort of patients.
Methods: All patients who were treated for TBI at a major trauma center between 2008 and 2019 and had an initial GCS of 14 or less were included. We evaluated GCS at admission and calculated the CRASH outcome prediction at 6 months. We then used the IMPACT score to estimate prediction with CT data and lab findings. Glasgow outcome score (GOS) and mortality at hospital discharge were the outcome measures. We then calculated the Kendall tau-b correlation coefficient of each predictor with each of the outcome measures.
Results: 544 patients could be included in the study. The patients’ mean age was 62.2 ±21.5 years, the median GCS in the emergency room was 14, IQR=12. 497 were alive at the time of discharge (91.4%). The median GOS at discharge was 4 ±1.
The CRASH score significantly correlated with each of the recorded outcome measures, i.e., with death (p = 1.06e-11) and GOS (p = 1.27e-20), as GCS did with mortality (p = 0.004) and GOS (p = 8.33e-20). CRASH score correlated significantly higher with mortality than GCS (τb = 0.29 vs τb = -0.11, respectively) and with GOS (τb = -0.36 vs τb = 0.33, respectively) at discharge.
The IMPACT score significantly correlated with each of the recorded outcome measures, i.e., with death (p = 2.37e-10) and GOS (p = 2.25e-14), as GCS did with mortality (p = 0.004) and GOS (p = 8.33e-20). IMPACT score correlated significantly higher with mortality than GCS (τb = 0.36 vs τb = -0.11, respectively) and with GOS (τb = -0.41 vs τb = 0.33, respectively) at discharge.
Conclusion: These results highlight the need for implementing more comprehensive assessment methods in routine clinical practice in the acute evaluation of patients with TBI. Specifically, approaches applying clinical and radiological features like the CRASH and IMPACT scores can lead to superior outcome prediction than that of the gold-standard GCS.
