11 non-academic hospitals reviewed 760 consecutive patients who had a brain CT for an acute headache that was followed by an LP 12 hrs from the onset of headache from January 2007 to January 2013. These 11 centers diagnosis roughly 250 subarachnoid bleeds annually. In this study, the patients presented within 6 hours from onset of the headache, and all had a negative CT read by staff radiologists, and were independently reviewed by two neuroradiologists and one stroke neurologist. At these 11 centers, of the 760 patients with a negative CT read by staff radiologists, 52 patients had CSF positive for bilirubin (7%). Of these 52 patients, there was one patient identified to have a non-aneurysmal perimesencephalic SAH on repeat review of the images. This one patient had a benign outcome. There were 8 others who had an aneurysm on CTA, DSA, or MRA (3 of which had been previously coiled). All of them were deemed as having rupture unlikely for various reasons (RBC <100, no bilirubin on spectrophotometry, etc).
So, with a negative CT read at a non-academic center by non-neuro radiologists, at the high end, we have a 1 in 760 miss rate if we *only* miss perimesencephalic bleeds on CT. These types of bleeds account for about 5% of SAH, so, potentially, at the low end, we are looking at a miss rate of 1 in 15,200. Essentially, the lumbar puncture is not a very useful test to diagnose SAH – with a posttest probability of 1.9% in cases with a positive CSF spectrophotometric result (a previous study reported about 8% PPV for xanthochromia)
Unfortunately, it is not mentioned how SAH was diagnosed throughout the study period. It would be nice to know if they were made via CT in the ED, as that would help solidify the author’s suggestion that CT/LP is a dinosaur in the age of shared decision making. Speaking of which…
I’ve had a few colleagues who have said, “show them the needle” as a somewhat subversive way to have patients either sign AMA or a refusal to consent for an LP for a subarachnoid hemorrhage. Few, if any, have actually said, “tell them the evidence.”
What sounds better to you for well informed shared decision making?
A) “You could die from this. You need a spinal tap. If you don’t do it, you could die.”
B) “Ultimately, I think your risk of a bleed is low, but I want you to understand that there is significant consequences to a subarachnoid and over half of those diagnosed may die. With that said, studies show that with a normal CT scan, your risk of having this condition is well below 2%. “Normal” spinal tap results performed 12 hours from the onset of your headache helps to further reduce your risk, but also comes with a significant number of false positives. While an abnormal spinal tap is concerning for a subarachnoid, it is not specific, and you will likely require admission for further testing if your spinal tap is abnormal.”
more reading on this: