• 
• 

Three Part Question

In [children who present to the emergency department] with [recurrent headaches, neurological abnormalities and recurrent vomiting] is [CT scan better than MRI].

Clinical Scenario

A 9 year old child comes to the emergency department accompanied by his mother. The child complains of recurrent headaches which started a month ago. The headaches awakens him from sleep and are worse in the morning. He has also had recurrent vomiting associated with his headaches. On questioning the childs mother she reports that her son tends to be confused during the headache episodes, and that there has been a recent change in her sons personality. You are worried that this child could have a brain tumor. You wonder whether a CT scan is better than an MRI scan to rule out the diagnosis of a brain tumor.

Search Strategy

Medline 1950-June 2007 using Ovid Interface
Embase 1980 to 2007 Week 25 using Ovid Interface
[(exp brain tumor OR exp brain cancer OR exp headache) AND (neurological abnormalities.mp. OR neurological symptoms.mp. OR exp.hemiplegia OR exp diplopia) AND (exp vomiting OR emesis.mp. OR vomit.mp.) AND (exp MRI scan OR exp CT scan OR exp imaging OR exp diagnostic imaging) AND (LIMIT to children and english)].

Search Outcome

4 papers were identified on Medline of which 1 was relevant and 10 papers were identified on Embase of which 1 was relevant.

Relevant Paper(s)

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
Anne M.Comi et al, 1997 New York	40 children less than 3 years.	Diagnotic study. Children with a diagnosis of ependymomas were given imaging.	Features seen on CT scan.	Isodense ependymomas seen in 57% of children. Hyperdense ependymomas seen in 26% of children. Hypodense ependymomas seen in 9% of children.	The size of the patient group was small.Only some of the patients were given contrast for enhancement and the reason for this was not specified.
			Features seen in 15 T1 weighted MRI scans	67% demonstrated low signal intensity lesions as compared with grey matter. 20% were isosignal with grey matter.
			Features seen in 16 T2 weighted MRI scans	15/16 revealed high signal tumors compared with surrounding grey matter.
			% of enhanced tumors using contrast CT scans.	29 patients were given contrast CT scans, 97% of the tumors were enhanced.
			Tumor enhancement using MRI scans + Gadolinium.	3/5 patients who received Gadolinium showed enhancement.
L.Santiago Medina et al, 1997 Boston	315 children, mean age of 11 years.	Retrospective study	Sensitivity	MRI 92%, CT scan 81%	The patient group were children who had been referred to a pediatric neurologist, therefore they were from a higher risk population. If the presence of a symptom was not mentioned by the patient then it was assumed to be absent.
			Specificity	MRI 99%, CT scan 92%
			Positive predictive value	MRI 96%, CT scan 93%
			Negative predictive value	MRI 97%, CT scan 86%
			Number of false positives	3 with MRI, 2 with CT scans
			Number of false negatives	6 with MRI, 6 with CT scans

Clinical Bottom Line

The results from the first study show that the key results with CT and MRI scans were totally different. Sensitivity and specificity of MRI scans was higher than that of CT scans, therefore MRI scans should be used for identifying brain tumors.

References

1. Anne M.Comi, James W.Backstrom, Peter C.Burger, Patricia K.Duffner and pediatric oncology group. Clinical and neuroradiologic findings in infants with intracranial ependymomas. Pediatric neurology 23-29, 1997.
2. L.Santiago Medina, Joseph D.Pinter, David Zurakowski, Ronald G.Davis, Karl Kuban, Patrick D.Barnes. Children with headache: clinical predictors of surgical space-occupying lesions and the role of neuroimaging. Pediatric neurology 819-824, 1997