LPs: Stop Managing by Myth


If you want to read a medical fairly tale, pick up an older textbook and read the chapter on lumbar punctures. Although you won’t read about unicorns or dragons, some of the recommendations are no less mythical. Let’s take a shot at disproving these myths with a silver bullet of truth.

Educational Objectives – After evaluating this article, participants will be able to:1. Develop strategies to perform lumbar puncture more efficiently without sacrificing patient safety

2. Incorporate rational, evidence-based strategies into clinical practice for the performance of lumbar puncture


3. Improve efficiency of diagnostic evaluations using lumbar puncture

In the past year or so, I began questioning common practices that frequently get in the way of efficient and sensible patient care. “That’s the way we’ve always done it,” just isn’t good enough anymore. Why do many, if not most of us, feel obligated to perform a CT scan before performing an LP? What about glucose and protein? Why do we use the needle in the LP kit? What is the real story of post-LP positioning?

Myth #1: The Exploding Head Theory


We were probably all taught, at one point or another, that a CT should be performed before ever putting a needle in someone’s back. Hypothetically, if they have increased intracranial pressure, they could herniate. Whether the literature supports this theory or not, fear of the exploding, or imploding, head has been reason enough to subscribe to this myth. In addition, many articles report that “pathophysiologic arguments” are the primary reason for this concern (Joffe AR. Lumbar puncture and brain herniation in acute bacterial meningitis: a review. J Intensive Care Med. 2007 Jul-Aug;22(4):194-207.).

So, what is the truth? Herniation can occur from LP. However, it is extremely rare. When it does occur, it is due to “brain shift” or “raised CSF pressure” whether or not papilledema is present or if intracranial pressure is actually elevated (van Crevel H, et al. Lumbar puncture and the risk of herniation: when should we first perform CT? J Neurol. 2002 Feb;249(2):129-37.). Those who advocate for the “CT-first approach” should recognize that in those at risk for herniation, the CT is frequently normal and a normal CT does not insure the safety of LP (Joffe AR).

Well, if you plan on doing a CT anyway, why not just get it first to be safe? The reason is efficiency. This is particularly important when performing a CT prior to LP for meningitis, as the CT is performed purely for safety reasons. But, even back in 1993, Archer reported that there is no evidence supporting the routine performance of cranial CT prior to lumbar puncture in suspected meningitis. If the patient shows atypical features (i.e. abnormal neurological examination), CT may be indicated (Archer BD. Computed tomography before lumbar puncture in acute meningitis: a review of the risks and benefits. CMAJ. 1993 Mar 15;148(6):961-5.). As neither CT nor LP are the treatment, make certain administration of antibiotics is timely, regardless whether the CT will be performed first or not.

What happens if you tap someone at risk for herniation? In a study to determine low-risk characteristics for LP, 3 patients that were misclassified underwent LP and did not herniate, and of the 56 patients with abnormal CTs, 52 had LPs performed and none suffered herniation (Hasbun R, et al. Computed Tomography Of The Head Before Lumbar Puncture In Adults With Suspected Meningitis. N Engl J Med 2001;345:1727-33.).


I completely agree that in the evaluation for an unusual headache (i.e. suspicion for subarachnoid hemorrhage) a CT and LP are both indicated. However, we should challenge what order to perform them in. If the CT must be performed first, the length of stay will likely double for those patients (Hasbun R). Performing tests in series is less efficient than in a parallel fashion. If we perform the CT, wait for the results and then perform the LP, the patient is in for a 6-hour length of stay in many cases (Hasbun R). However, if the LP is performed and the CT is performed directly after, we can wait for both results simultaneously, reducing the total length of stay considerably.  Tempting, isn’t it?

Do your patients frequently refuse their LP after hours have gone by, their pain is resolved or you just told them their CT was normal? Many do. Despite the fact that response to analgesics does not reduce the likelihood of a serious cause of headache, many EPs and their patients are unduly comforted by such reductions in pain (Pope J, Edlow J, Favorable Response to Analgesics Does Not Predict a Benign Etiology of Headache. Headache: The Journal of Head and Face Pain, PG: 944-950, 2008.). Your best opportunity to perform the LP is when you and your patient share the same concern, directly after the word “aneurysm” comes out of your mouth.

In the mid 1990s, safe LP, in the absence of papilledema or focal neurological findings, was reported in the literature and accepted by many, including specialty societies. Despite this practice-changing endorsement, many still are concerned about ICP and LPs. Let’s dig into the roots of the concern to dismiss this myth altogether. Although the concern for herniation in “increased ICP” sounds perfectly reasonable, this practice has little to no basis in the literature. It appears that this concern is based on a study published in 1967 by Dr. Duffy (G. P. Duffy. Lumbar Puncture in the Presence of Raised Intracranial Pressure, Br Med J. 1969;1:407-409, 15 February). Duffy observed 30 patients with end-stage brain tumors. All had LPs performed. 50% herniated immediately and the other half herniated 12 hours later. So, 100% herniated. This is where the mass hysteria originated. Interestingly, all of Duffy’s patients had progressive headache, an altered mental status and localizing neurological findings. Would any of us be crazy enough to tap these patients anyway?

They all were well on their way to herniation before the needle ever pierced the epidermis. It’s wise to proceed with caution and perform a thorough neurological examination prior to LP. However, blind adherence to the performance of CT first should be abandoned. Many forward thinking physicians have raised the argument that LP alone is sufficient for the evaluation of subarachnoid hemorrhage. I would have to agree, if we are willing to put all of our diagnostic eggs into the SAH basket. However, the CT may provide alternative, plausible explanations for the headache such as mass lesion, stroke or lytic lesions. One caveat is the radiologist report of sinusitis. Although acute sinusitis may cause headache, many experts feel that chronic sinusitis does not. Thus, reported findings of chronic disease, such as mucous retention cysts, mucosal thickening and chronic sinusitis should not result in aborting the search for bad disease (Jones NS. Sinus headaches: avoiding over- and mis-diagnosis. Expert Rev Neurother. 2009 Apr;9 (4):439-44).

Myth #2: Watch That Glucose!

For years, I would run to the printer, anticipating the arrival of my CSF cell counts, only to be disappointed to find the glucose and protein results. It never fails, the glucose and protein always seem to be reported first. 4 years ago, I just quit ordering them. Guess what? No phone calls and no complaints. Now, the lab always runs my cell counts first. Can any of us claim we have made the diagnosis of meningitis or other entity, solely based on the glucose or protein? I doubt it. “Wow! Look at that low glucose. This must be a particularly hungry strain of bacteria.” You can always save a tube for others to order worthless tests later. However, in the ED, the cell counts are where the money is.

So, what is the source of this tradition with CSF glucose? It appears to date back to 1965, where Swartz, et al. reported that 50% of cases of meningitis had a glucose below the reference range. Th
at’s it? Yes. Furthermore, we should question where the reference ranges came from to begin with. CSF glucose ranges were set with the presumption that the CSF glucose should be two-thirds that of the serum glucose – not very scientific. Omitting these tests will only reduce cost and improve the return rate on the tests that really make a difference.

Myth #3: Use the LP Needle in the Kit

Size matters and so does the sharpness of your tool. Post-dural puncture headache rates of around 30% have been consistently reported in the obstetric and anesthesia literature when the standard Quincke (cutting, needle) is used. The blunt, or pencil tipped, Whitacre or Sprotte needles are preferred. Further, the larger the needle you use, the greater the risk of spinal headache. Hatfield reported that the blood patch rate for the 20 gauge Quincke was 29% and 15.2% for the 22 gauge. However, when a 22 gauge Whitacre was used, the rate decreased to 4.2% (Hatfield, M.K., et al. Blood patch rates after lumbar puncture with Whitacre versus Quincke 22-and 20-gauge spinal needles, Am J Roent 190(6):1686, June 2008.). This is just one of many studies that proves this theory. Small and dull is better than sharp and big. So, should we use the needle in the kit? My opinion is no.  As many providers are using alternative needles, the manufacturers benefit from selling the standard LP kit and the atraumatic needles separately. However, if they didn’t put a needle in the kit, they would be hard pressed to call it an LP kit. So, most kits come equipped with an inexpensive, expendable, 20-gauge Quincke needle, knowing many will discard it anyway.

Myth #4: Post LP Positioning Is Important

What about positioning following an LP? Is it important or necessary? No and no. The classic instruction is, “Lay still on your back for one hour.” This is uncomfortable for the patient and adds an additional hour to their length of stay. Does patient positioning impact the likelihood of post-dural puncture headache? The short answer is no. Just like other dogma (medical dog food), this recommendation is full of filler with little substance. In the Canadian Medical Association Journal in 2001, a meta-analysis included 16 studies and 2,213 patients, testing bed rest up to eight hours, brief bed rest and immediate mobilization. Bed rest does not reduce the likelihood of spinal headaches. Furthermore, prolonged bed rest was associated with increased likelihood of spinal headache in some patients. (Thoennissen, Jana, Herkner, Harald, Lang, Wilfried, Domanovits, Hans, Laggner, Anton N., Mullner, Marcus. Does bed rest after cervical or lumbar puncture prevent headache? A systematic review and meta-analysis. CMAJ 2001 165: 1311-1316.). Rest assured, after LP, patients can go to CT, sit up and eat a meal, walk to the bathroom or take a nap. It really doesn’t matter. The likelihood of post dural puncture headache is based on the trauma caused by the procedure not what happens afterwards. If you use a large needle, use Quincke needles or have a challenging tap, your patients are more likely to experience a spinal headache.

I realize that there may be varying opinions on these topics and wide-range of comfort in accepting what I am proposing. However, it is our responsibility to keep an open mind and be ever vigilant about challenging tradition to the benefit of our patients. Outdated practices and poor literature have probably done more to negatively impact the care we provide than good literature has resulted in improvements.  It’s time to stop managing by myth.



  1. Penny Karson on

    Why won’t you sedate patients if they are extremely anxious? I would probably have a panic attack and jerk all over the place. I had a horrible experience with needles placed in my back and just thinking about it wants to make me puke. I would refuse a lumbar puncture because of this. Why don’t you offer sedation to hesitant patients instead of torturing them? You are not looking out for the patient’s best interests and by not offering sedation you may traumatize them. I didn’t visit a doctor for many years because of several horrible experiences. I find that doctors just don’t care about the emotional and psychological damage they inflict on some patients. We are all different so don’t treat us all the same. A patient’s emotional well-being should be considered when performing medical tests as it effects whether the patient will seek further medical care. Some people just can’t tolerate certain procedures. Everything should be done to make it as tolerable as possible. USE SEDATION if the patient wants it.

    • A very reasonable question. It is unfortunate that it was not answered by the physicians. Would sedation affect the CSF? If not…?

Leave A Reply