Assessment of the prognostic value of Helsinki computer tomography score in severe traumatic brain injury patients at Kenyatta National Hospital.

Abstract

Background: Head injury is usually considered as a silent epidemic and continues to exist as an everlasting consternation conferring high mortality and disability worldwide. In the early assessment of patients with injury to the head, the computer tomography(CT) is of utmost importance and millions of CT scans are conducted yearly. These CT scans contain information which can be used to determine the patient's prognosis and to form a baseline in identifying risk in clinical trials. The Helsinki computer tomography scoring classification was created in 2014 as a tool for determining outcomes in those patients with TBI. In Kenya, severe TBI accounts for 10.3% of all brain injuries seen at the Kenyatta National Hospital and 14.3% of all adults admitted at the Critical Care Unit. The Helsinki computer tomography score being the latest in the armamentarium of outcome predictors and having outperformed previous CT scoring system in the European and Asian subcontinent, it will be necessary to assess its capacity to predict outcome in the African subcontinent on severe traumatic brain injury patients and hence guide clinical decision making. Objectives: This study assessed the prognostic value of the Helsinki Computer Tomography score among patients with severe traumatic brain injury. Methodology: Following ethical approval, a prospective observational study involving forty two patients above 18 years of age with severe TBI were recruited by convenience sampling at the Kenyatta National Hospital Accident and Emergency Department and Critical care units over a period of four months(October 2019- January 2020). Clinical parameters of blood pressure, pupillary reactivity, random blood glucose, age, GCS and Helsinki computer tomography(CT) score were evaluated at admission and subsequent follow up done at 6 weeks for Glasgow outcome scoring. Data was collected using a structured questionnaire and recorded in excel sheets and analysis done using Statistical Package for Social Sciences version 23.0. Results: A total of 42 patients were recruited with 90% males and mean age of 33 years old. Overall mortality was 64.3%. RTA was the commonest mode of injury at 64% followed by assault at 26% and falls at 10%. Patients with non reactive pupil had mortality of 67% while slow reacting pupil had 63%. Patients with systolic BP > 90 mmHg comprises 95% of the study population with a resultant mortality at 67.5%. The most common random blood glucose level was < 10 mmol/l at 80% with a mortality of 58.8%. Patients with GCS of 3-4 had the highest mortality of 100% while GCS of 7-8 lowest mortality of 60.9%. GCS of 3-4 had no favourable outcome at 6 weeks while GCS 7-8 had favourable outcome in 30.4%. The Helsinki CT score of 4 had mortality of 33.3% while Helsinki CT score of 11 had mortality of 100%. Patients with contusions and intracerebral hematomas had mortality of 80% while in ASDH and EDH the mortality were 53.8% and 44.4% respectively. In correlation analysis the Helsinki score was significantly associated with GOS at 6 weeks(p=0.004), and death(p=0.009). Age was significantly correlated with 6 weeks GOS(p=0.03) and mortality(p=0.02). Systolic BP was only associated with mortality at p value of 0.043. The other clinical parameters did not show any statistical significance with both 6 weeks GOS and mortality. The specificity, sensitivity and accuracy for Helsinki CT score for mortality were 88.9%, 53.3% and 71% respectively; and for an unfavourable outcome, these values were 81.8%, 55.6% and 69% respectively. After performing logistic regression analysis to determine the predictors of outcome, we found that the odds ratio(OR) for the Helsinki CT score to predict mortality to be 9.1(95% CI 1.9-44) and unfavourable outcome at 5.6(95% CI 1.2-27.4). Conclusion: Severe TBI carries a high mortality and disability in Kenya. The age of patient, the systolic blood pressure on admission and the initial Helsinki CT score were significant predictors of outcome(p < 0.05). The Helsinki CT score correlated well with the clinical parameters at predicting outcome. Recommendations: A change to new computer tomography scoring system may be warranted and the Helsinki CT score can be used as a predictor of outcome in Kenyan hospitals and in our African population. Future studies comparing the different computer tomography scores available in correlation with clinical parameters on predicting outcome should be done as a multi-center study.