| Abstract |
Background: Coma is associated with complete loss of consciousness and is considered an emergency and life-threatening situation. Investigating children with coma is a challenge since the causes may be numerous and heterogeneous. Detailed history, full clinical examination, laboratory testing, and appropriate neuroimaging are important for the correct diagnosis. Foremost and regardless of the etiology of coma, it is essential to perform primary assessment according to the ABCDE (Advanced Pediatric Life Support) approach and apply specific therapeutic protocols as soon as possible, in order to support the patient’s vital functions, stabilize the patient, and prevent secondary brain damage.
Imaging is a valuable diagnostic tool. Scoring systems based on imaging findings have been associated with severity and prognosis: in adults and children with traumatic coma the Marshall’s score, a 6-point Computed Tomography (CT) scale, is used; in children with hypoxic coma, an 8-point scale Magnetic Resonance (MR) score has been proposed. To the best of our knowledge, there is no imaging severity score system evaluating the severity of neuroimaging findings for pediatric coma regardless of its etiology.
Objectives: To construct a modified imaging-based severity score for assessment of children with coma (Coma Imaging Severity Score or Modified Marshall Score) separately for CT and MRI.
To investigate the contribution of neuroimaging, and in particular of Modified Marshall CT and MRI scores in the diagnosis and monitoring of children with coma, regardless of etiology (traumatic and non-traumatic).
To correlate imaging findings and modified imaging severity scores with clinical and laboratory markers with regard to the severity and etiology of pediatric coma.
Methods: This retrospective study included all patients with coma admitted to the Pediatric Intensive Care Unit (PICU) at the University Hospital of Heraklion from January 2014 to March 2018. Overall, 85 children aged 3 month to18 year-old (mean 7.56±5.31 years) were recruited and classified in three different clinical groups according to the etiology of coma: a) Traumatic coma (n=21), b) coma due to medical conditions like status epilepticus, encephalitis, meningitis, sepsis or acute diffuse encephalomyelitis (n=62) and c) non-traumatic surgical coma, due to arteriovenous dysplasia-subarachnoid hemorrhage or unexplained subdural hemorrhage (n=2). For each patient, demographic data (gender, age, ethnicity), final diagnosis, co-morbidities, outcome, length of stay, clinical severity scores (PRISM, PELOD), complications, treatment (antibiotics, inotropes, barbiturates, surgery analgesia, decompression craniotomy), mechanical ventilation, Glascow Coma Scale (GCS), intracranial measurement parameters (Intra Cranial Pressure (ICP), Cerebral Perfusion Pressure (CPP)), lumbar puncture, imaging methods and total number of the neuroimaging tests were recorded. A modified Marshall CT scoring system and a modified Marshall MR scoring system took into account possible lesions found in neuroimaging, by adding parameters to the Marshall CT score. Modified Marshall CT Scores and modified Marshall MR Scores based on imaging findings of test up on admission (1st CT and/ or MRI) were compared with respective scores on imaging tests at the day of worse PICU stay (2nd CT and/or MRI). The following criteria defined worse PICU day: reduction of the level of consciousness and GCS, new neurological signs and symptoms, raised ICP, reduced CPP. We examined whether imaging investigations provided important additional or new information that led to clinical diagnosis change and/or modify medical therapy or management modification. The statistical package SPSS25 was used for data analysis. P<0.05 was considered statistically significant.
Results: Percentages of children subjected to one and two MRIs were 62.1% and 5,2%, respectively while percentages of children subjected to 1 and 2 CTs were 67,2% and 20,7%, respectively. Modified Marshall CT scores were associated with increased predicted death probability (p<0.001, ANOVA), with an increased length of stay, increased mechanical ventilation duration (p<0.01), and a higher ICP and lower CPP of the 1st or of the worse PICU day (p<0.03). Although GCS showed a similar declining trend with lower values as the CT score increased, the relationship did not reach statistical significance.
Modified Marshall CT Scores tended to be higher on the 2nd CT - worst PICU day, compared to the 1st CT, in all clinical groups. In all cases, the 2nd CT added new information and/or led to treatment modification. Similar results were found for MRI, where 2nd MRI either added new information or led to treatment modification.
The initial ICP values were significantly related to those of the worse PICU day (r=0.87, p<0.001), the clinical disease severity (PRISM) and the predicted mortality (r=0.75, p<0.001), and were inversely related with lower CPP values (r=-0.78, p<0.001).
Higher ICP values (>15 mmHg) were associated with increased Modified Marshall CT and MR Scores and with a variety of imaging abnormalities in the same study (p<0.001). Lower CPP values (<50mmHg) were associated with a higher number of CTs, increased Modified Marshall CT and MR Scores and with a variety of imaging abnormalities in the same test (p<0.001). Patients with CPP <50mmHg had higher mortality rates compared to patients with CPP>50mmHg or patients without CPP monitoring (p<0.03)
Conclusions: Neuroimaging in comatose children with a deteriorating neurological condition and increased ICP during PICU stay might contribute by addition of new information and consequent treatment modification. Assessment of pediatric traumatic and non-traumatic coma could be performed by neuroimaging-based severity CT and MRI scores as they seem to correlate with outcome prediction probability, clinical severity indices, length of PICU stay, length of mechanical ventilation as well as increased ICP and decreased CPP values. These modified imaging severity scores should be validated in large multicenter studies in order to confirm their reliability as diagnostic and prognostic tools.
|
Search
