Most doctors no longer do spinal taps… and that is a problem. As a resident, I performed dozens of lumbar punctures. We even had our own lab in the ICU where the internal medicine residents could spin down cerebrospinal fluid specimens and do diff-quick stains for cell counts and India ink preps for Cryptococcus in the middle of the night. In truth, we had a lot more meningitis back then, with few adults having received Pneumococcal vaccines and AIDS being rampant. But today, there are less lumbar punctures being performed and fewer physicians trained in how to do them.
Back in the 1980’s, all internal medicine residents learned how to do LPs. If you were unsuccessful, you found another resident who was more experienced to do it. If that resident was unsuccessful, you called a neurosurgeon. And all LPs were performed at the bedside, using only palpation of lower back anatomy to guide needle insertion.
Many doctors no longer perform spinal taps
Several factors have led to fewer patients getting lumbar punctures. Most older adults have had a pneumococcal vaccine, reducing the incidence of pneumococcal meningitis. Most children have received a Haemophilus vaccine, reducing the incidence of Haemophilus meningitis. Most people with HIV are receiving anti-retroviral therapy, reducing the incidence of cryptococcal meningitis. Use of colony stimulating factors with chemotherapy and improved immunosuppressant regimens have reduced the incidence of opportunistic infections causing meningitis. Advances in CT and MRI imaging have reduced the need for lumbar puncture for the diagnosis of subarachnoid hemorrhage.
A second factor is that bedside lumbar punctures have become more difficult. The two most important reasons for a lumbar puncture attempt to fail are patient obesity and spinal arthritis. Obesity makes it more difficult to palpate the lumbar landmarks. Arthritis causes deformities of the spinous processes and reductions in the diameter and shape of the channel that the spinal needle must pass through the interspinous ligaments. As the baby boomers have aged, the percentage of older Americans has increased and with that age has come more spinal arthritis. And as Americans have become more obese in the past three decades, the average patient needing a spinal tap has become more obese.
A third factor has been an evolution of the hospital credentialing process. In the past, lumbar puncture was generally a “core privilege” for internists, family physicians, pediatricians, emergency medicine physicians, anesthesiologists, and surgeons. As a core privilege, any physician in those specialties was allowed to perform a lumbar puncture. As the U.S. physician credentialing process has become more strict, hospitals started requiring new physicians to provide procedure logs documenting successful performance of some minimum number of each bedside procedure – typically 10 for lumbar punctures. Many internal medicine, surgery, and family practice residents were unable to perform that many LPs during residency and thus were unable to apply for hospital privileges to perform lumbar punctures as an attending physician. Because a resident is technically acting under the supervision of an attending physician, if the attending physician was not credentialed to do lumbar punctures, then the resident working under that attending physician was not permitted to do an LP either. This has led to a snowballing of fewer and fewer residents doing enough LPs to be able to get credentialed to do them as an attending physician. Lumbar punctures have dropped off of the core privilege lists in many hospitals. Once a procedure drops off of the core privilege list, physicians have to provide documentation of maintenance of skill by doing some minimum number of each procedure every 1 or 2 years. This has resulted in a further drop in the overall number of physicians credentialed to perform lumbar punctures.
As a result of all of these factors, interventional radiologists now perform a large portion of lumbar punctures on inpatients. In our hospital’s ICU, many physicians would not attempt an LP, even if they had hospital credentials to perform them, simply because it was easier to just put in an order for an interventional radiologist to do it. Uniquely, I actually liked performing spinal taps and did a fairly large number of them to the point that I’d get called to do them in the ICU, even if I was not assigned to inpatient duties that week.
The thing is that doing a spinal tap is an art. You have to be able to palpate the iliac crest and the different components of the spine. You have to know which spinal interspace to choose for needle insertion. You have to know how to angle the needle while inserting it. You have know how deep to insert the needle based on how thick the muscles (or fat) on any given patient’s back. You have to know how it feels when the needle enters the spinous ligament and how it feels when the needle pops through the dura mater. And if your needle hits bone, you have to know what part of the patient’s spine you hit so that you can redirect for successful insertion. In other words, it takes skill and like any skill, it takes practice and more practice to be good at it.
There are three problems with relying on interventional radiologists to perform your hospital’s spinal taps. First, they don’t like doing them – they don’t pay very well and take up a lot of time that the radiologist could be doing more lucrative procedures. Second, they aren’t always available – few radiologists are going to come into the hospital at midnight on a Saturday to do a fluoroscopically-guided LP but if you think a patient has bacterial meningitis, that LP really can’t wait until sometime the following Monday when the radiologist does elective procedures. And third, they often don’t send the spinal fluid for the correct tests – you only want to put your patient through an LP once and you don’t want to have to do a second procedure because the radiologist didn’t know to order a bacterial or fungal culture on the spinal fluid.
A possible solution…
There is a way to bring lumbar punctures back to the bedside where they belong, rather than routinely sending all patients to the radiology department for their spinal taps… ultrasound.
30 years ago, all ultrasound was relegated to radiologists and cardiologists. Then, in the 1990’s we bought a small portable ultrasound machine for our medical intensive care unit and I started using it to help identify jugular veins when placing central venous catheters. Bedside ultrasound soon became the imaging test of choice to evaluate patients with blunt abdominal trauma in the emergency room. Today, few residents even know how to place a central venous line, place an arterial catheter, perform a thoracentesis, or perform a paracentesis without ultrasound guidance.
Bedside ultrasound can also be used to assist performance of lumbar punctures. A 2018 study of emergency department patients found a 90% success rate using ultrasound guidance vs 81% with conventional anatomic landmark guidance; this study also showed shorter procedure times, lower patient pain scores, and fewer needle passes using ultrasound. A 2020 study found that neurologists had a success rate of 93% using ultrasound guidance vs. 68% without ultrasound. Moreover, this study found that the procedure time decreased from 38 minutes to 17 minutes with the use of ultrasound. A 2024 meta-analysis of LPs in children found that pediatricians had an improved success rate of 72% with ultrasound vs. 60% without ultrasound and it also resulted in fewer traumatic taps. In 2024, a meta-analysis of patients undergoing epidural and spinal anesthesia found a significant increase in first-attempt success rates using ultrasound guidance. In 2019, the Society for Hospital Medicine recommended that ultrasound be used to guide lumbar puncture whenever it is available.
So, what should hospitals do now?
- Residency programs should incorporate simulation laboratory training in lumbar puncture using ultrasound.
- Residency programs should encourage use of ultrasound guidance when performing lumbar punctures on patients whenever feasible.
- Hospitals should purchase portable bedside ultrasound machines that can be used in emergency departments, intensive care units, operating rooms, and general inpatient nursing units.
- Hospitals should ensure that existing bedside ultrasound machines have an appropriate transducer for lumbar spine imaging.
- Hospitals should develop simulation laboratory training in ultrasound guided lumbar puncture for those hospitalists, emergency medicine physicians, intensivists, and anesthesiologists who have not been trained in ultrasound guidance.
- Hospitals should develop a training and credentialing process for physicians who do not currently have lumbar puncture privileges. This could include simulation laboratory training coupled with some number of supervised lumbar puncture procedures in patients as a pathway to obtaining hospital privileges.
- Hospitals currently utilizing “procedures teams” of physicians, nurse practitioners, and/or physician assistants should incorporate ultrasound guidance for performance of lumbar punctures.
- Because most hospitals do not perform a sufficient number of lumbar punctures for all physicians to perform a minimum number of procedures annually to maintain credentials, an alternative pathway to re-credentialing lumbar puncture privileges using a simulation lab should be adopted.
It’s time to embrace ultrasound
I was one of the first physicians to use bedside ultrasound to guide central venous catheter placement in our intensive care unit. But I have to admit that after a few months, I stopped using it except in rare occasions because it slowed me down. I had placed so many internal jugular vein, subclavian vein, and femoral vein catheters that I could do them quickly and successfully just by look and feel. But most physicians have not placed hundreds of central lines without ultrasound guidance and most physicians have not performed dozens of lumbar punctures using anatomic landmark identification alone.
When a patient is suspected of having meningitis or encephalitis, they need a lumbar puncture and they need that LP right now. Choosing the wrong antibiotic that doesn’t cover the bacteria in a patient’s cerebrospinal fluid or failing to give an antiviral drug for the patient with herpes encephalitis can mean the difference between complete recovery and permanent brain damage (or death). We need to give hospital-based physicians the tools and training that they need to take optimal care of these patients. And that means embracing ultrasound guidance for lumbar punctures.
April 24, 2025
