lumbar puncture halp from a pleading senior

This forum made possible through the generous support of SDN members, donors, and sponsors. Thank you.

Fungi121

Full Member
10+ Year Member
Joined
Jun 3, 2012
Messages
65
Reaction score
3
Kind of a last ditch effort here. I'm sort of out of ideas and youtube videos to watch.

I suck at lumbar punctures. I've tried doing them laying left lateral decub, sitting up, etc. My success rate is 50% which I'm kind of embarrassed about.

My issues:
1) I feel like I have trouble staying midline/finding midline, especially in left lateral. In large people, I can't even feel the spinous processes

2) "Once you hit bone you can find your way" But there's midline bone and then random process bone.

3) Where am I supposed to be aiming in left lateral vs upright? I'm really struggling.

Any advice? I'm really concerned.

Members don't see this ad.
 
Have you ever tried Ultrasound?
It's saved my butt on more than a few and has gotten me some champagne taps on obese people.
Vasc probe w Vertical orientation w the marker to the head, find the SP shadows and center on the space between. I make a hash to the left, right, center marks above and below and then mark the center w a sterile pen. Enter there, profit
 
  • Like
Reactions: 1 user
This is a procedure that can psych folks out, but everybody misses sometimes.

95% of success is based on patient position. Focus heavily on that.

I tend to do mine upright with a bedside table with blankets or pillows for the patient to rest their head on. If the patient can't put their feet on the floor put something under their feet for support. Take the time to make sure the patient is comfortable so they won't fatigue and move during the procedure.

Some general tips to getting midline: as you have the patient bend their head down, focus on aligning the c7 spinous process with the top of the patients crack, that'll help you find the midline. Then focus on ensuring the patients neck/shoulders/hips/knees are all in line with each other to avoid spinal rotation. Then have the patient arch their back. Explain to the pt how everything should be aligned and that they gotta hold the position. Have a nurse or tech with hands on the patient's shoulders to help maintain the position you need once you get everything aligned properly. Palpate and mark L3 and L4 interspaces and don't be afraid to use L3. If the patient has a really elevated BMI and you can't feel anything, just use the iliac crest level to estimate L4 and just focus on staying midline--even if you hit bone the first poke it can help you understand where you are.

Don't use the crappy drape as it can cause you to loose your landmarks (ie the crack) but instead betadine more real estate than you otherwise would and make sure your patient has a cap on and you've got all your gear on as well.

Insert the needle on a trajectory towards the patient's belly button (ie somewhat cephalad).

Good luck.
 
  • Like
Reactions: 4 users
Members don't see this ad :)
Kind of a last ditch effort here. I'm sort of out of ideas and youtube videos to watch.

I suck at lumbar punctures. I've tried doing them laying left lateral decub, sitting up, etc. My success rate is 50% which I'm kind of embarrassed about.

My issues:
1) I feel like I have trouble staying midline/finding midline, especially in left lateral. In large people, I can't even feel the spinous processes

2) "Once you hit bone you can find your way" But there's midline bone and then random process bone.

3) Where am I supposed to be aiming in left lateral vs upright? I'm really struggling.

Any advice? I'm really concerned.

In terms of identifying landmarks I see a lot of people using the intergluteal cleft as the landmark. I think it can be very misleading in terms of where the midline actually is, particularly in the lateral decubitus position. Instead, sit people up at first to find the midline (even if you plan to do the LP in the lateral decubitus). Feel for the midline several spaces up. Feel for it in the thoracic spine if you have to. Once you've found the line, use the marker to draw a huge line following the midline all the way down the patient's back, past the point at which you would expect to insert the needle. Then identify the vertical dimensions of the intervertebral space of interest. This is also easier to do sitting up. I would recommend drawing and upper and lower margin to the space you are feeling. Be generous with this marking too. Don't be afraid to go way lateral until it looks like there is a giant cross on the patient's back. It's not like the skin marker is painful or harmful. People may find this extensive marking funny, but that doesn't matter nearly as much as developing your confidence with LPs. Extensive marking will help orient you way better than the usually useless circle or skin indentation people place. Eventually you won't need it any more.

Another thing that completely changed my experience with LPs is premedicating the patients with a benzodiazepine. I recommend midazolam or diazepam. Relaxes the patient, which makes the procedure easier both because the muscles will be more relaxed, but also because YOU will be more relaxed. Also, patients love it.
 
  • Like
Reactions: 1 user
Feel the spinous processes all the way down the spine to get midline. Position well. And just don’t quit.
 
This might be a dumb question, and honestly I feel embarrassed to even ask it...

But why do we make such a big deal about obtaining CSF?

We empirically treat septic patients with antibiotics without a source routinely. If a patient looks like crap and you think they have meningitis, why not just treat them empirically and admit them to the hospital? Who cares if you get the confirmatory test or not?

We do this ALL the time. How often have you attempted an LP multiple times in the ED and failed, and you just start empiric treatment. At my hospital, we will attempt an LP in the ED, really more than anything just to placate the inpatient team. If we can't get it, we start meningitic dosing of antibiotics in the ED and admit the patient to the floor overnight after which point they will get a fluoro-guided LP in the AM.

I suspect at least one part of the equation is that we want to identify an organism so we can narrow antibiotics and not have to keep patients on broad spectrum coverage for weeks. I understand resistance is a real issue.

By all means, attempt an LP to get extra data and support your diagnosis. But if you can't get it after one or two sticks, most of the time it probably has more to do with patient body habitus than it does with your skills. If the clinical suspicion is there (toxic appearing patient, neuro deficits, altered, seizing, febrile), I really don't care about what the CSF shows.
 
  • Like
Reactions: 1 users
95% of success is based on patient position. Focus heavily on that.

100% agree patient positioning means everything! I bring extra pillows...blankets and stuff them between the patients knees (for proper hip posture) and put one under the RIGHT arm in left lateral decub position for reasons I write below...

I do all of my in left lateral decub position. My success rate is prob 80-90% (16-18/20 seems about right). I tell the nurse who is helping me exactly what I need.

When lying on your left side, you have to make sure patients don't coil up in a way that contorts their spine. What I want is their shoulders and hips to be perfectly vertical. That is....if I were make a line connecting the humeral heads that line needs to be 100% vertical. Think about that for a second....lie on your side and think about the position of the shoulders. Same thing goes for the hips....a line connecting the femoral heads needs to be vertical. It's usually not a problem for the hips.

Another thing that completely changed my experience with LPs is premedicating the patients with a benzodiazepine. I recommend midazolam or diazepam. Relaxes the patient, which makes the procedure easier both because the muscles will be more relaxed, but also because YOU will be more relaxed. Also, patients love it.

I use versed 1-2 mg IVP (usually 2 for adults) prior to every single LP (except for infants.) It's great! And it wears off in 1 hour. I even gave it once to a meningitic lady whose BP was 90/50. 1 mg versed calmed her down and got the LP.
 
  • Like
Reactions: 1 users
This might be a dumb question, and honestly I feel embarrassed to even ask it...

But why do we make such a big deal about obtaining CSF?

We empirically treat septic patients with antibiotics without a source routinely. If a patient looks like crap and you think they have meningitis, why not just treat them empirically and admit them to the hospital? Who cares if you get the confirmatory test or not?

We do this ALL the time. How often have you attempted an LP multiple times in the ED and failed, and you just start empiric treatment. At my hospital, we will attempt an LP in the ED, really more than anything just to placate the inpatient team. If we can't get it, we start meningitic dosing of antibiotics in the ED and admit the patient to the floor overnight after which point they will get a fluoro-guided LP in the AM.

I suspect at least one part of the equation is that we want to identify an organism so we can narrow antibiotics and not have to keep patients on broad spectrum coverage for weeks. I understand resistance is a real issue.

By all means, attempt an LP to get extra data and support your diagnosis. But if you can't get it after one or two sticks, most of the time it probably has more to do with patient body habitus than it does with your skills. If the clinical suspicion is there (toxic appearing patient, neuro deficits, altered, seizing, febrile), I really don't care about what the CSF shows.

I think most of this is reasonable. If we can confirm the suspected diagnosis w/ a relatively simple procedure though, why not? Personally I don't find LPs that hard (I probably just doomed by next couple attempts by saying that). I definitely used to dread them as an intern, so I'm empathetic to the OP. As said above ad nauseam, positioning is everything. Also, just like 80% of medicine, a lot of the classic teaching is bull****. Personally I feel I could almost do an LP blindfolded, and the reason is I've got the feel down. It isn't a pop, at all. It's a sudden increase in resistance that will still give way to steady pressure on the needle, and once you're past that ligament, you're in.

The three reasonable pieces of advice in my mind are 1. Spend time getting great positioning, 2. Leave the stylet out once you're past the subcutaneous tissues, and 3. Have an extra needle. Once you employ these 3 things, it's really just about time in the field until you get good at them.
 
This might be a dumb question, and honestly I feel embarrassed to even ask it...

But why do we make such a big deal about obtaining CSF?

We empirically treat septic patients with antibiotics without a source routinely. If a patient looks like crap and you think they have meningitis, why not just treat them empirically and admit them to the hospital? Who cares if you get the confirmatory test or not?

We do this ALL the time. How often have you attempted an LP multiple times in the ED and failed, and you just start empiric treatment. At my hospital, we will attempt an LP in the ED, really more than anything just to placate the inpatient team. If we can't get it, we start meningitic dosing of antibiotics in the ED and admit the patient to the floor overnight after which point they will get a fluoro-guided LP in the AM.

I suspect at least one part of the equation is that we want to identify an organism so we can narrow antibiotics and not have to keep patients on broad spectrum coverage for weeks. I understand resistance is a real issue.

By all means, attempt an LP to get extra data and support your diagnosis. But if you can't get it after one or two sticks, most of the time it probably has more to do with patient body habitus than it does with your skills. If the clinical suspicion is there (toxic appearing patient, neuro deficits, altered, seizing, febrile), I really don't care about what the CSF shows.

Yea I do agree....in-so-much that getting "cultures" prior to antibiotic treatment is recommended.

I tell hospitalists that LP's are not really an emergency procedure. We are trained to do it, but it's not life saving. An LP does not save someone's life. It can be delayed. They get pissed off about this and don't understand. The only time you need to do one is if you want to exclude SAH and want to send the patient home. Then you have to do it.

We get blood cultures prior to antibiotics
We get urine cultures prior to antibiotics
It's reasonable to try to get CSF cultures prior to antibiotics, but due to the nature of procedure it's like...infinitely more difficult to get CSF cultures prior to antibiotics.
 
  • Like
Reactions: 1 users
This might be a dumb question, and honestly I feel embarrassed to even ask it...

But why do we make such a big deal about obtaining CSF?
Yea I do agree....in-so-much that getting "cultures" prior to antibiotic treatment is recommended.

I tell hospitalists that LP's are not really an emergency procedure. We are trained to do it, but it's not life saving...

Agreed. CSF anaylsis is nice to have for the inpatient side, and an opening pressure is nice to have, but in the vast majority of emergent pathologies, it's not key. Think it's meningitis? Treat it. Immunosuppressed? Add the weird stuff. Think it's an SAH? CT then CTA if inside the 4-6 hr window. If not it's a bit murkier, but if they have no aneurysm on a CTA, what are you going to do emergently about a non-aneurysmal SAH? Think it's IIH and they have no vision changes? Still not an emergency. GBS? Clinical diagnosis that is aided by an LP.

I'm happy to do LP's, and will continue to do them as much as possible in residency, but have yet to have one change management in the ED.
 
It’s important to get for two reasons:
1) culture data guides antibiotic therapy - drug and duration
2) if we assume meningitis but are wrong, we need to find the real diagnosis. If we just say “seems like meningitis, do the LP at your convenience” it may be a day later before we realize that we’re actually dealing with status or TTP or something else weird that we may not have considered and may not be treating.
 
  • Like
Reactions: 3 users
Kind of a last ditch effort here. I'm sort of out of ideas and youtube videos to watch.

I suck at lumbar punctures. I've tried doing them laying left lateral decub, sitting up, etc. My success rate is 50% which I'm kind of embarrassed about.

My issues:
1) I feel like I have trouble staying midline/finding midline, especially in left lateral. In large people, I can't even feel the spinous processes

2) "Once you hit bone you can find your way" But there's midline bone and then random process bone.

3) Where am I supposed to be aiming in left lateral vs upright? I'm really struggling.

Any advice? I'm really concerned.

You're doing the procedure how 95% of EPs teach it, and at the same time you're doing everything you can to make LP's as hard as possible by using a midline approach. Do your LPs with a paramedian approach. It’s far superior technique. There is zero increased risk and extreme improvements in success rates, ease & efficiency.

Think of a figure of 8, sideways (infinity symbol). You're shooting an arrow. You're insisting you strike the point where the lines cross, every time. That's the space between the spinous processes. Using the paramedian approach allows you to aim for the entire circle on either side of the sideways, figure of eight.

Start 1 level below and a couple of cm lateral to your target level. Head a little bit midline and upwards and your space opens up by 10-fold. There will be no interspinous ligament in the way or to cause pain. If you hit bone, redirect until you pop in. Use this approach and your success rate will approach 95+% soon. Watch an LP or epidural steroid injection under fluoro, 1 time, and see how much more space there is paramedian vs midline and you'll immediately slap yourself across the face and chant, "SHAME...SHAME...SHAME..." to yourself, for ever attempting a midline LP, through thick painful ligament, and a needle-eye target.

This is not a made up, or voodoo technique. It's straight from anesthesia/pain/IR practice. It's well supported in EM lit also, if that makes you feels better.

Techniques for Performing Paramedian Approach to Lumbar Puncture - ACEP Now

Currently I do spine injections under fluoro and when doing an ESI, though not going all the way intrathecal, I would never, every torture myself by going midline. I always go paramedian, even when I want my needle tip to end up midline. It's simply the best approach.

(Disclaimer: I struggled with LPs early on, too, in my ED training days. Use this approach and they'll get much easier with zero increased procedural risk).
 
Last edited:
  • Like
Reactions: 1 user
Yes I think both viewpoints are valid. CSF is important to have, and in general it's not an emergency procedure.

I can think of several occasions were I will try the LP two or three times (inserting the needle 2-3 times)...can't get it and I stop. I'm not going to sit in the room for 30 minutes or more struggling to try to get the LP, or giving more morphine or versed to keep the patient calm, or this and that to try to get the CSF.

Unfortunately one example is for neonates for the full sepsis workup for fever...you would think it's easy because they are so small and landmarks are easy to come by. Most of the time despite my insistence that parents leave the room, if they insist on being in there then I let them stay. If I can't get it after a handful of attempts I stop and tell the pediatrician or NICU doc who will be taking care of the patient.

For the fat people...sometimes I'll pull up a recent Abd CT and measure the distance from skin to spinal canal at the lumbar level. If it's long then I tell the inpatient team "this needs to be done under fluoro with a long needle. It's hard to guide a bendy 20 g spinal needle 7-10 cm deep into a small opening to get CSF.


Overall I just don't like LPs....it's a time consuming procedure with variable and unpredictable difficulty that rarely changes ED management. I can be in there for 10-15 minutes positioning the patient (because nurses never do that properly), setting up, poking around......then you get the CSF and it comes out like 1 drop every 3 seconds....and you sit there feeling like a dope for another 10 minutes just to get 6 cc's of CSF....doh!
 
  • Like
Reactions: 1 user
Members don't see this ad :)
C'mon now, we all know the LP needs to be done on these pt's. It's important both diagnostically and therapeutically for their inpatient management. That's like asking why we get blood cultures on septic patients. What you don't want to do is delay therapy due to prolonged set up for the LP. I think the ID literature dictates initiating therapy within 30 mins if you really think they have it. I knew of a guy involved in a lawsuit over a pediatric pt who reportedly presented very atypically with GI sx. Long story short, the history and work up finally pointed him in the direction of meningitis but the LP and dx was delayed by 2-4 hours. He ended up getting the CSF and making the dx but there was a sig delay in therapy. The pt got flown to local pediatric ICU and I'm not sure what happened to the pt but he was sued over the delay in therapy. I'm not sure what happened as it was told to me by one of his colleagues. Just goes to show, don't delay therapy while your nurses set up for an LP that might take > 1h if you think you really have meningitis.

Unfortunately, the situation I often find myself in is initiating therapy with delayed set up for the LP though I will try to get the CFS within the first hour at the very least. You kill meningococcus pretty rapidly, I think within 15 mins to 1 hour, however many of the other pathogens you can culture 2, 4 even 8 hours out.

Interpretation of Lumbar Puncture Cultures After Parenteral Antibiotic Pretreatment
 
  • Like
Reactions: 1 user
C'mon now, we all know the LP needs to be done on these pt's. It's important both diagnostically and therapeutically for their inpatient management. That's like asking why we get blood cultures on septic patients. What you don't want to do is delay therapy due to prolonged set up for the LP. I think the ID literature dictates initiating therapy within 30 mins if you really think they have it. I knew of a guy involved in a lawsuit over a pediatric pt who reportedly presented very atypically with GI sx. Long story short, the history and work up finally pointed him in the direction of meningitis but the LP and dx was delayed by 2-4 hours. He ended up getting the CSF and making the dx but there was a sig delay in therapy. The pt got flown to local pediatric ICU and I'm not sure what happened to the pt but he was sued over the delay in therapy. I'm not sure what happened as it was told to me by one of his colleagues. Just goes to show, don't delay therapy while your nurses set up for an LP that might take > 1h if you think you really have meningitis.

Unfortunately, the situation I often find myself in is initiating therapy with delayed set up for the LP though I will try to get the CFS within the first hour at the very least. You kill meningococcus pretty rapidly, I think within 15 mins to 1 hour, however many of the other pathogens you can culture 2, 4 even 8 hours out.

The patients that lose out are the "not sick" meningitis patients, or the low risk ones whom probably don't even need it done anyway.....

Patient A: Fever, AMS, Meningismus, BP 80/40.....they get everything immediately, resuscitation, Abx, LP, because they are supersick.
Patient B: AMS, Na 108, +EtOH, who has come in like 3 times like this previously. Not meningitis. The hospitalist very nicely told me she was not going to take the patient unless I did an LP. :-(

I just won't sit in the room forever trying to get an LP on someone if they are being treated appropriately. I don't have any shame saying "I tried and couldn't get it."

BTW I agree with you as I wrote above that we get blood cultures and urine cultures in septic patients prior to antibiotics. They are easy to get. Thankfully I can't remember the last time I had a really sick meningitic patient. Been a few years I think. Never seen one in kids.
 
Why EM attendings teach midline LP technique boggles my mind. It's a far inferior technique. I think they're worried about hitting a nerve root or something laterally, but if you redirect towards the midline there's no way your needle will end up in the foramen. Plus, there are nerve roots in the thecal sac you're aiming at, when you do midline LPs also. There's zero reason to do them midline. There's much, much, much more room with the paramedian approach. Much smaller passageway, with painful, thick ligament and much more bone to hit, midline. Learn this technique and you be super confident with LPs and rarely if ever miss one. Sit the patients up, too, if able.

See image:
 

Attachments

  • Screen Shot 2018-09-25 at 10.03.08 PM.png
    Screen Shot 2018-09-25 at 10.03.08 PM.png
    202.7 KB · Views: 176
  • Like
Reactions: 1 user
C'mon now, we all know the LP needs to be done on these pt's. It's important both diagnostically and therapeutically for their inpatient management. That's like asking why we get blood cultures on septic patients.
Blood culture data is far and away unreliable. Contamination rates are high. When you look at multiple studies regarding blood cultures drawn in the ED, the positive yield ranges somewhere from 2-6% with only ~2% of positive cultures actually influencing patient management. I agree that we need guided therapies and narrowing antibiotic therapy is essential, but overall the yield for this test is quite low IMO.

Just because it's part of the CMS sepsis bundle doesn't mean its clinically relevant in the ED. That being said, I will echo what thegenius said, namely, blood cultures are low hanging fruit and relatively easy to obtain. CSF is a obviously going to be harder. I'm not saying because it's a time sink we should just abandon doing LPs, but after one or two tries, in my opinion, you are getting diminishing returns.

Side note: apparently people with meningococcus will have positive BLOOD cultures ~60% of the time, so I would argue for these patients blood cultures are pretty much mandatory. Growth in CSF cultures is even higher, so make no mistake, the diagnostic value of the LP is awesome (it's the gold standard for a reason). It's just a cost/benefit analysis of actually doing the procedure in my mind.

I find the LP to provide useful but non-essential information from the ED standpoint. If the thought of meningitis even enters my mind and the patient looks sick, they are getting meningitic coverage irrespective of whether I obtain CSF or not. I completely understand that this approach is plagued with issues, but it's one of the only options we have. I believe that for some patients, just given the constraints of their body habitus, it's unreasonable to expect to easily obtain a CSF sample without fluoro-guidance (which I obviously am not trained to do).

Maybe this is why I'm not a neurologist or an ID doc. Or maybe I just need to get better at doing LPs. To those of you that have perfected the art of the LP, I applaud you. You should continue to clinch the diagnosis and perform them. Me on the other hand, I still need some work and will be unfortunately contributing the wave of antibiotic resistance we have until I figure it out :)
 
Last edited:
Why EM attendings teach midline LP technique boggles my mind. It's a far inferior technique. I think they're worried about hitting a nerve root or something laterally, but if you redirect towards the midline there's no way your needle will end up in the foramen. Plus, there are nerve roots in the thecal sac you're aiming at, when you do midline LPs also. There's zero reason to do them midline. There's much, much, much more room with the paramedian approach. Much smaller passageway, with painful, thick ligament and much more bone to hit, midline. Learn this technique and you be super confident with LPs and rarely if ever miss one. Sit the patients up, too, if able.

I believe this is a promising technique, but I'm not just gonna try doing it on the next patient requiring an LP. It doesn't "boggle" my mind that EM attendings don't teach it because most don't do it. I looked at a YouTube.com video on Paramedian approach (for epidurals in obstetrics) and 96% of anesthesiologists in England do the midline approach while 4% do paramedian.

I've never even heard of this approach until I read this thread.
 
  • Like
Reactions: 1 user
The patients that lose out are the "not sick" meningitis patients, or the low risk ones whom probably don't even need it done anyway.....

Patient A: Fever, AMS, Meningismus, BP 80/40.....they get everything immediately, resuscitation, Abx, LP, because they are supersick.
Patient B: AMS, Na 108, +EtOH, who has come in like 3 times like this previously. Not meningitis. The hospitalist very nicely told me she was not going to take the patient unless I did an LP. :-(

I just won't sit in the room forever trying to get an LP on someone if they are being treated appropriately. I don't have any shame saying "I tried and couldn't get it."

BTW I agree with you as I wrote above that we get blood cultures and urine cultures in septic patients prior to antibiotics. They are easy to get. Thankfully I can't remember the last time I had a really sick meningitic patient. Been a few years I think. Never seen one in kids.

Regarding pt B: I'd escalate issues like that with your medical director to address. Was the hospitalist demanding an LP without even seeing the pt? Regardless, if you say they don't need an LP, then they don't get one. Period. If the hospitalist is refusing an admission due to a procedure that you clearly think the pt does not require and clearly has risks associated with it then I'd escalate that fairly quickly. You have a variety of options in the heat of the moment.... a) Consult neuro to obtain buy in on your management decision, then re-consult hospitalist for admission, b) admit to someone else, c) escalate with hospital administration, d) formally consult the hospitalist to bedside and for placement of note as bylaws usually dictate any consultant respond within a reasonable timeframe (usually 30 mins or an hour), e) transfer the pt to a facility that will admit them and tell your administrator that you're worried about potential EMTALA violation due to refusal of admission 2/2 jacka** hospitalist (there's usually an area on your COBRA form where you are asked to put the name of any consultant that refused to provide treatment for the pt resulting in your need to transfer). LOL, hopefully by option E you've stirred the hornets nest enough to result in your administrator calling the hospitalist director who has in turned ripped your hospitalist a new one over the phone and told them to admit the pt immediately.

Of course, it goes without saying that if you decide to play nasty at any point, be absolutely certain that they didn't need that LP! ;)
 
  • Like
Reactions: 1 user
Regarding pt B: I'd escalate issues like that with your medical director to address. Was the hospitalist demanding an LP without even seeing the pt? Regardless, if you say they don't need an LP, then they don't get one. Period. If the hospitalist is refusing an admission due to a procedure that you clearly think the pt does not require and clearly has risks associated with it then I'd escalate that fairly quickly. You have a variety of options in the heat of the moment.... a) Consult neuro to obtain buy in on your management decision, then re-consult hospitalist for admission, b) admit to someone else, c) escalate with hospital administration, d) formally consult the hospitalist to bedside and for placement of note as bylaws usually dictate any consultant respond within a reasonable timeframe (usually 30 mins or an hour), e) transfer the pt to a facility that will admit them and tell your administrator that you're worried about potential EMTALA violation due to refusal of admission 2/2 jacka** hospitalist (there's usually an area on your COBRA form where you are asked to put the name of any consultant that refused to provide treatment for the pt resulting in your need to transfer). LOL, hopefully by option E you've stirred the hornets nest enough to result in your administrator calling the hospitalist director who has in turned ripped your hospitalist a new one over the phone and told them to admit the pt immediately.

Of course, it goes without saying that if you decide to play nasty at any point, be absolutely certain that they didn't need that LP! ;)

LMAO. I vaguely remember the case. The Hospitalist saw the patient and first asked me "you think he might need an LP? he isn't really responsive to much at all." I replied....."if he hadn't come in like this 3 times previously, then maybe I would agree with you. I think he is minimally responsive due to being drunk..."

I was irritated that she wouldn't take him, but it was like 1:00 AM and I went back into the room with the hospitalist and we both looked at him. He was grunting to noxious stimuli. It wasn't the worst thought ever...his ETOH was like 200-250 which wasn't that high for him. So I did the LP. It was one cell away from being a champagne.


However....I 100% agree in general. We cannot be forced to do a procedure. I think a hospitalist once asked me to do a thoracentesis on a guy whose left lung field was totally whited out on CXR. He wasn't sick but he needed admission. I told him that there is no indication for me to do this right now, its not a life saving procedure, and if I do this I would only pull out 50 cc, which will anger you and the patient when you send him to IR the next day for a large volume thoracentesis...he's going to ask why do I have to get his done twice.....so just have IR the next day to a large volume thoracentesis. "I will not do a procedure where the risks outweigh the benefits" I remembered saying. I even admitted to them that I've done like 10-20 thoracenteses in my life. Despite the fact that they are easy, **** happens every now and then, no? I even said "if someone has abdominal pain and you want the surgeon to do surgery, and the surgeon consults and says no.....you can't demand the surgeon do surgery."
 
I will admit to you, when I graduated from residency I was similarly frustrated with LPs and felt they were my worse procedure. My technique drastically appeared in my first year of attending practice. Here are things that are key:

1.)always sitting up right. Too hard to find the midline in lateral decubitus. Patients are always rotated about their axis in lateral decub. I even do neonates held upright.

2.)really nail the positioning, measure twice, cut once. Make sure you have squared up the patients hips, shoulders, neck, and back. Also, do not tell the patient to "bend forward" because all they do is flex their hips, what you really need them to do is "push your low back towards me." That puts their L spine in a little bit of kyphosis and opens the space up.

3.)I mark the landmarks with a marking pen including illiac crests, midline, spinous processes, so I see where the intraspinous spaces are. This also helps once you inject lidocaine and the skin wheal/induration creates a firm mark that is difficult to differentiate from bones under skin.

4.)All patients get ativan preprocedure. Even the people who seem cool. Something happens, they feel that big needle going deeper...and deeper...and deeper in their back, and they start freaking out and moving, and now the procedure is way harder.

5.)Lido with epi, not plain lido. more champagne taps, less confusion with traumatic taps in SAH rule outs.

6.)start with your needle as close to the inferior spinous process as you can and then angle cephalad.

7.)its deeper than you think, keep advancing the needle.
 
  • Like
Reactions: 2 users
I know of people who will do the LP sitting up, then once the needle is in the subdural space then will lay them lateral to get an opening pressure. Is that really a good idea when you have a spinal needle in someones intrathecal space?

Seems, prima facie, crazy to do that!
 
For the folks that are saying they don't see the value of the LP: I would encourage you to look at what's best for the patient overall, not what is technically necessary from the ED perspective. Is it best for them to have the CSF tapped as early as possible, ideally before antibiotics are started? I think the answer is clearly yes, for all the reasons mentioned above (being able to narrow antibiotics or decide on the duration of therapy). Just because I can bully the hospitalists into taking the patient without doing an LP or pretend I don't understand why it's necessary that it be done as early as possible (and therefore in the ED), doesn't mean that's the right thing to do.
 
  • Like
Reactions: 3 users
You're doing the procedure how 95% of EPs teach it, and at the same time you're doing everything you can to make LP's as hard as possible by using a midline approach. Do your LPs with a paramedian approach. It’s far superior technique. There is zero increased risk and extreme improvements in success rates, ease & efficiency.

Think of a figure of 8, sideways (infinity symbol). You're shooting an arrow. You're insisting you strike the point where the lines cross, every time. That's the space between the spinous processes. Using the paramedian approach allows you to aim for the entire circle on either side of the sideways, figure of eight.

Start 1 level below and a couple of cm lateral to your target level. Head a little bit midline and upwards and your space opens up by 10-fold. There will be no interspinous ligament in the way or to cause pain. If you hit bone, redirect until you pop in. Use this approach and your success rate will approach 95+% soon. Watch an LP or epidural steroid injection under fluoro, 1 time, and see how much more space there is paramedian vs midline and you'll immediately slap yourself across the face and chant, "SHAME...SHAME...SHAME..." to yourself, for ever attempting a midline LP, through thick painful ligament, and a needle-eye target.

This is not a made up, or voodoo technique. It's straight from anesthesia/pain/IR practice. It's well supported in EM lit also, if that makes you feels better.

Techniques for Performing Paramedian Approach to Lumbar Puncture - ACEP Now

Currently I do spine injections under fluoro and when doing an ESI, though not going all the way intrathecal, I would never, every torture myself by going midline. I always go paramedian, even when I want my needle tip to end up midline. It's simply the best approach.

(Disclaimer: I struggled with LPs early on, too, in my ED training days. Use this approach and they'll get much easier with zero increased procedural risk).

You've been saying this for years, and I've yet to actually try it. If I do, and I find that you've been right all along, a "SHAME...SHAME...SHAME..." will indeed be in order.
 
I believe this is a promising technique, but I'm not just gonna try doing it on the next patient requiring an LP. It doesn't "boggle" my mind that EM attendings don't teach it because most don't do it. I looked at a YouTube.com video on Paramedian approach (for epidurals in obstetrics) and 96% of anesthesiologists in England do the midline approach while 4% do paramedian.

I've never even heard of this approach until I read this thread.
1. Don't "just try it" or do it because you heard it on SDN from me. Ask someone who knows how to do it, show you how to do it. It's not rocket science. It's just moving a needle a couple cm laterally and a couple cm down with zero increase in procedural risk, and increased ease of access.

2. Most "don't do it" precisely because most EM attendings don't teach it, because they are teaching what they were taught. It's "turtles all the way down." That's how medical dogma works. But that's not the same as proof of superiority.

3. With all do respect, quoting stats you heard on you tube is not credible. Posting a relevant link from a reputable medical source, like the link above I posted from ACEP, is credible. Also, threading a catheter like in a labor epidural (or spinal cord stimulator lead, for example) is not the same as simply getting access with a needle. When threading a catheter or lead, staying midline is more important because if your introducer needle penetrates the epidural space laterally, your catheter or lead will drift into the lateral recess of the canal, get caught up on exiting nerve roots laterally or even go anterior, which is not what you want. When doing an LP, you just need to penetrate the thecal sac and get fluid. You don't need to be 100% midline. You're comparing apples to oranges.

Regardless, when you do a paramedian LP, if you're doing it right, your needle heads towards the midline, anyways. Similarly, if you're not doing your midline LP correctly, your needle can drift laterally anyways. For some reason, there's an anxiety amongst some EP's in training in doing LPs (I think primarily due to patient anxiety) but it's an extremely safe procedure. The complications you can get (post dural puncture headache, epidural hematoma or abscess) have nothing to do with your needle ending up too far laterally. If you go to far laterally, you just hit the bone of the facet joint.

Do your LPs however you want. But I'm telling you, as someone who struggled with midline non-image guided LPs during EM training, who now does spine injections under fluoro, without a shadow of a doubt, using a paramedian approach, with a steeper angle, starting a level below your target, opens up your entry space dramatically, avoids the painful interspinous ligaments, and avoids trying to thread the needle between a narrow interspinous space. Seeing all this bony anatomy under fluoro, I cringe when I think of all the times I went midline and struggled. I never choose that approach now (for my fluoro-guided epidural injections) and that's even with the luxury of having imaging which makes getting through a tighter space so much easier. Even when I want my needle tip to penetrate at the midline (sometimes I want that, sometimes I don't, based on laterality of symptoms) I still use a paramedian approach. It's simply a less painful technique, that's easier for the patient and doctor and faster, with no increase in procedural risk. I think more EPs should try it for their LPs.
 
  • Like
Reactions: 1 user
1. Don't "just try it" or do it because you heard it on SDN from me. Ask someone who knows how to do it, show you how to do it. It's not rocket science. It's just moving a needle a couple cm laterally and a couple cm down with zero increase in procedural risk, and increased ease of access.

2. Most "don't do it" precisely because most EM attendings don't teach it, because they are teaching what they were taught. It's "turtles all the way down." That's how medical dogma works. But that's not the same as proof of superiority.

3. With all do respect, quoting stats you heard on you tube is not credible. Posting a relevant link from a reputable medical source, like the link above I posted from ACEP, is credible. Also, threading a catheter like in a labor epidural (or spinal cord stimulator lead, for example) is not the same as simply getting access with a needle. When threading a catheter or lead, staying midline is more important because if your introducer needle penetrates the epidural space laterally, your catheter or lead will drift into the lateral recess of the canal, get caught up on exiting nerve roots laterally or even go anterior, which is not what you want. When doing an LP, you just need to penetrate the thecal sac and get fluid. You don't need to be 100% midline. You're comparing apples to oranges.

Regardless, when you do a paramedian LP, if you're doing it right, your needle heads towards the midline, anyways. Similarly, if you're not doing your midline LP correctly, your needle can drift laterally anyways. For some reason, there's an anxiety amongst some EP's in training in doing LPs (I think primarily due to patient anxiety) but it's an extremely safe procedure. The complications you can get (post dural puncture headache, epidural hematoma or abscess) have nothing to do with your needle ending up too far laterally. If you go to far laterally, you just hit the bone of the facet joint.

Do your LPs however you want. But I'm telling you, as someone who struggled with midline non-image guided LPs during EM training, who now does spine injections under fluoro, without a shadow of a doubt, using a paramedian approach, with a steeper angle, starting a level below your target, opens up your entry space dramatically, avoids the painful interspinous ligaments, and avoids trying to thread the needle between a narrow interspinous space. Seeing all this bony anatomy under fluoro, I cringe when I think of all the times I went midline and struggled. I never choose that approach now (for my fluoro-guided epidural injections) and that's even with the luxury of having imaging which makes getting through a tighter space so much easier. Even when I want my needle tip to penetrate at the midline (sometimes I want that, sometimes I don't, based on laterality of symptoms) I still use a paramedian approach. It's simply a less painful technique, that's easier for the patient and doctor and faster, with no increase in procedural risk. I think more EPs should try it for their LPs.

It makes anatomic sense and I also like the fact that the patient does not need to be flexed.
 
I have a fairly high success rate. Another vote for:

1. Measure twice. Poke once. Positioning. Premedication if needed.

2. Measure twice.

3. MEASURE TWICE. As in, position, take your time, and be sure of where you're poking and your landmarks.

4. I frequently adjust my angle a bit based on the image in my head of how the patient's spine is -- but often find myself starting out with a paramedian approach. Not 90 degree insertion. Just like Birdstrike said.
 
  • Like
Reactions: 1 user
1. Don't "just try it" or do it because you heard it on SDN from me. Ask someone who knows how to do it, show you how to do it. It's not rocket science. It's just moving a needle a couple cm laterally and a couple cm down with zero increase in procedural risk, and increased ease of access.

2. Most "don't do it" precisely because most EM attendings don't teach it, because they are teaching what they were taught. It's "turtles all the way down." That's how medical dogma works. But that's not the same as proof of superiority.

3. With all do respect, quoting stats you heard on you tube is not credible. Posting a relevant link from a reputable medical source, like the link above I posted from ACEP, is credible. Also, threading a catheter like in a labor epidural (or spinal cord stimulator lead, for example) is not the same as simply getting access with a needle. When threading a catheter or lead, staying midline is more important because if your introducer needle penetrates the epidural space laterally, your catheter or lead will drift into the lateral recess of the canal, get caught up on exiting nerve roots laterally or even go anterior, which is not what you want. When doing an LP, you just need to penetrate the thecal sac and get fluid. You don't need to be 100% midline. You're comparing apples to oranges.

Regardless, when you do a paramedian LP, if you're doing it right, your needle heads towards the midline, anyways. Similarly, if you're not doing your midline LP correctly, your needle can drift laterally anyways. For some reason, there's an anxiety amongst some EP's in training in doing LPs (I think primarily due to patient anxiety) but it's an extremely safe procedure. The complications you can get (post dural puncture headache, epidural hematoma or abscess) have nothing to do with your needle ending up too far laterally. If you go to far laterally, you just hit the bone of the facet joint.

Do your LPs however you want. But I'm telling you, as someone who struggled with midline non-image guided LPs during EM training, who now does spine injections under fluoro, without a shadow of a doubt, using a paramedian approach, with a steeper angle, starting a level below your target, opens up your entry space dramatically, avoids the painful interspinous ligaments, and avoids trying to thread the needle between a narrow interspinous space. Seeing all this bony anatomy under fluoro, I cringe when I think of all the times I went midline and struggled. I never choose that approach now (for my fluoro-guided epidural injections) and that's even with the luxury of having imaging which makes getting through a tighter space so much easier. Even when I want my needle tip to penetrate at the midline (sometimes I want that, sometimes I don't, based on laterality of symptoms) I still use a paramedian approach. It's simply a less painful technique, that's easier for the patient and doctor and faster, with no increase in procedural risk. I think more EPs should try it for their LPs.

Slow down tiger....I'm not quoting random s&^t from youtube. It's a talk given at an Anesthesiology Conference last year.
At time index 2:25 is where I took the 96% vs 4%.

I sense a little bit of frustration or animosity from your post...I have written nothing to impugn the paramedian approach. It's almost like your angry at me for mentioning that I haven't heard of it prior to this thread. I would just like to witness it in person before trying it, but I think it's going to be difficult to witness because it is used or discussed so rarely.

BTW, the anesthesiologist in the video said that it ought to be easier to thread a catheter using the paramedian approach (time index 20:30 shows a picture)
 
  • Like
Reactions: 1 users
Slow down tiger....I'm not quoting random s&^t from youtube. It's a talk given at an Anesthesiology Conference last year.
At time index 2:25 is where I took the 96% vs 4%.

I sense a little bit of frustration or animosity from your post...I have written nothing to impugn the paramedian approach. It's almost like your angry at me for mentioning that I haven't heard of it prior to this thread. I would just like to witness it in person before trying it, but I think it's going to be difficult to witness because it is used or discussed so rarely.

BTW, the anesthesiologist in the video said that it ought to be easier to thread a catheter using the paramedian approach (time index 20:30 shows a picture)

I posted in favor of paramedian approach. You disagreed, saying midline is better and your supporting link is a lecture called, "Paramedian technique: Why you should adopt it."
 
Last edited:
I posted in favor of paramedian approach. You disagreed, saying midline is better and your supporting link is a lecture called, "Paramedian technique: Why you should adopt it."

Alright...I don't know where I wrote midline is better. All I said was it looks promising and I haven't heard of it before. And I wasn't going to do it for my next LP. Dunno how that's interpreted as midline is better. I believe you that paramedian is easier. My lack of adopting it immediately just means I'm not going to practice it on live patients before I have time to study it more and see it done by someone else.
 
Alright...I don't know where I wrote midline is better. All I said was it looks promising and I haven't heard of it before. And I wasn't going to do it for my next LP. Dunno how that's interpreted as midline is better. I believe you that paramedian is easier. My lack of adopting it immediately just means I'm not going to practice it on live patients before I have time to study it more and see it done by someone else.
I thought you were a resident or student at first. I now see you're an attending. My intended audience for the posts in this thread was more for students and residents still in the learning phase. If you have good results with midline LPs, keep doing them. I can't say there's any need to change if you're having good results. It's just a technique I (and some others) think is easier. Not everyone agrees.
 
In terms of tactile feel, do you tend to encounter the same type of resistance while going paramedian, that you get while going midline?
 
My average patient is a turbo-arthritic senior with a curly-cue spine. Never again will I go midline with an LP.
 
  • Like
Reactions: 1 user
In terms of tactile feel, do you tend to encounter the same type of resistance while going paramedian, that you get while going midline?

It doesn't sound like it...I have trained myself to feel the resistance from the spinous ligament and have a general idea of how long it is....looking at the model for the paramedian approach: it appears you traverse muscle until you puncture the ligamentum flavum. Then you go like another 1-2 mm and you puncture the dura and BOOM you are in. I am interested in this approach because if it alleviates (or mitigates) the fear and uncertainty of doing an LP I might do them more.
 
In terms of tactile feel, do you tend to encounter the same type of resistance while going paramedian, that you get while going midline?
It's a much softer feel for the operator and the patient's react in pain, less, because you're not crunching through the interspinous ligaments between the spinous processes.
 
I'm a resident, and after reading this thread, I youtubed how to do it. Did it paramedian for the first time today. Got it on the first try. Will be doing LPs paramedian until I miss. Thanks for the tip!
 
  • Like
Reactions: 4 users
Kind of a last ditch effort here. I'm sort of out of ideas and youtube videos to watch.

I suck at lumbar punctures. I've tried doing them laying left lateral decub, sitting up, etc. My success rate is 50% which I'm kind of embarrassed about.

My issues:
1) I feel like I have trouble staying midline/finding midline, especially in left lateral. In large people, I can't even feel the spinous processes

2) "Once you hit bone you can find your way" But there's midline bone and then random process bone.

3) Where am I supposed to be aiming in left lateral vs upright? I'm really struggling.

Any advice? I'm really concerned.

Ultrasound is your friend. Even if you just use the vascular probe at max depth with a skin marker to scout the area and mark the spinous processes and other things you see, you'll end up with a patient's back that looks a little like the screen from Luke's targeting computer from Star Wars. Then just turn the ultrasound off, grab your needle, and draw the fluid. Piece of cake. You don't even have to use sterile procedures with the ultrasound, clean the back after you map it.
 
Ultrasound is your friend. Even if you just use the vascular probe at max depth with a skin marker to scout the area and mark the spinous processes and other things you see, you'll end up with a patient's back that looks a little like the screen from Luke's targeting computer from Star Wars. Then just turn the ultrasound off, grab your needle, and draw the fluid. Piece of cake. You don't even have to use sterile procedures with the ultrasound, clean the back after you map it.


Yessssssss.

 
  • Like
Reactions: 1 users
I never found US helpful for LP to be honest.

I like to find the midline with my anesthetic needle. When you meet resistance with the longer 22 g needle when injecting lidocaine (you won't be able to inject), you are in the ligamentum flavum. That is the direct midline. If you can find this when injecting the lidocaine, then you know exactly where your midline is and it will make the LP very very easy.

I always have the needle perpendicular or slightly angled with the needle tip pointing more towards the head. I go just below the top spinous process I'm going between if I can easily feel them.

In an adult, if you hit bone fairly shallow, you are hitting the spinous process, your issue is one of needing to go more superior or inferior to get between the two.

If you hit bone deeper, you are hitting lateral elements of the vertebra, you need to go more R or L depending on which way you are missing.

I never tried the paramedian approach, but I'm willing to try it on my next difficult LP. To be honest though, with doing my LPs how I described above, I haven't missed one in years.
 
  • Like
Reactions: 1 users
You're doing the procedure how 95% of EPs teach it, and at the same time you're doing everything you can to make LP's as hard as possible by using a midline approach. Do your LPs with a paramedian approach. It’s far superior technique. There is zero increased risk and extreme improvements in success rates, ease & efficiency.

Think of a figure of 8, sideways (infinity symbol). You're shooting an arrow. You're insisting you strike the point where the lines cross, every time. That's the space between the spinous processes. Using the paramedian approach allows you to aim for the entire circle on either side of the sideways, figure of eight.

Start 1 level below and a couple of cm lateral to your target level. Head a little bit midline and upwards and your space opens up by 10-fold. There will be no interspinous ligament in the way or to cause pain. If you hit bone, redirect until you pop in. Use this approach and your success rate will approach 95+% soon. Watch an LP or epidural steroid injection under fluoro, 1 time, and see how much more space there is paramedian vs midline and you'll immediately slap yourself across the face and chant, "SHAME...SHAME...SHAME..." to yourself, for ever attempting a midline LP, through thick painful ligament, and a needle-eye target.

This is not a made up, or voodoo technique. It's straight from anesthesia/pain/IR practice. It's well supported in EM lit also, if that makes you feels better.

Techniques for Performing Paramedian Approach to Lumbar Puncture - ACEP Now

Currently I do spine injections under fluoro and when doing an ESI, though not going all the way intrathecal, I would never, every torture myself by going midline. I always go paramedian, even when I want my needle tip to end up midline. It's simply the best approach.

(Disclaimer: I struggled with LPs early on, too, in my ED training days. Use this approach and they'll get much easier with zero increased procedural risk).

Great read. Will definitely try this next time. Last LP I did was on a 300 lb woman, regular midline approach, getting 1 mL of lidocaine at a time (don't ask). Lots of good tips in this thread.
 
  • Like
Reactions: 1 user
Anesthesiologist here, sorry for the intrusion. Thought perhaps I could lend some experience. In my limited experience most of us do midline epidurals (17g tuohy needle in mostly young laboring women) as it’s easier to stay midline while threading a catheter (which needs bilateral setup for labor pain). That being said no one would blink an eye if one were to do paramedian approach for epidurals as we often do that for thoracic epidurals, which tend to be more difficult.

Access to CSF is different. We use much smaller needles (25-27g) to minimize risk of PDPH. Such small needles meet more resistance and the needle itself (cutting or pencil point) won’t pass through bony elements that larger tuohy needles are able to direct around. In short, for access to CSF paramedian makes a lot of sense, and it’s what I see most anesthesiologists do. Midline is fine, but paramedian will give access in cases were midline is difficult.
 
  • Like
Reactions: 3 users
Anes here as well, thought i'd give my 2 cents. We usually access pts that are usually in the reproductive age for epidurals with bigger touhy needles, but lumbar puncture should be a very similar if not the same skill.

The way i got really good was by looking at and manipulating a spine model. You'll notice that the space is somewhat mobile and variable depends on the positioning, you will also notice the space you're getting into is about 10 or 15 times the size of the touhy needle. That should give you confidence. I don't understand the paramedian vs midline debate in lumbar punctures. The difference matters some in thoracic approaches, but it should really matter very little in lumbar approaches if you get your fundamentals correct. The beginner should start with the sitting position (unless you need an opening pressure).

The key is to be able to deduce where you are in relation to the target. E.g. If you felt a divot in the back, then you can deduce that you're between the spinous processes. If your first attempt resulted in hitting in bone and your 2nd approach is aimed more caudad but results in hitting the bone at a shorter needle mark, you're probably not along the lines of of the gradual slope of the midline, but perhaps hit something else less smooth either to the right or the left.

I also highly encourage copious amounts of local anesthetic. When you're using the much smaller 25G local needle to inject, don't be afraid to use the needle it self to map out the anatomy of the patient. Furthermore, inject when you feel bone, that really lessens the pain from repeated approaches.

Unfortunately there is no substitute for experience, so perhaps ask your OB anes collegues to see if they will let you attempt a spinal (or even watching them do a few is good mental reps). Elective c-sections will be a very calm and controlled area for you to get some good beginner reps.

I also agree that ultrasound can be very helpful, but that is somewhat advanced topic and a long talk, in short: curvelinear probe for BMI > 30 and good marking to denote your trajectory.
 
  • Like
Reactions: 1 user
The key is to be able to deduce where you are in relation to the target. E.g. If you felt a divot in the back, then you can deduce that you're between the spinous processes. If your first attempt resulted in hitting in bone and your 2nd approach is aimed more caudad but results in hitting the bone at a shorter needle mark, you're probably not along the lines of of the gradual slope of the midline, but perhaps hit something else less smooth either to the right or the left.

I think this is a key that has not been discussed too much at this point, using the data from your missed attempts and then using that to visualize the anatomy in three dimensional space and recalibrate properly on your next attempt.

Examples: If you hit bone very shallow you likely hit the dorsal surface of a spinous process and you know that your next attempt needs to be adjusted inferiorly/superiorly.

If you hit bone at what feels very deep you may be hitting facet or lamina and you need to move more midline on your next attempt.

If you hit bone at medium depth, you are probably in the right space; but your angle may be too sharp or shallow and thus you are hitting the spinous process along its length and rather than the top of it as in case #1; thus you need to adjust your angle more caudal/rostral.

I am enjoying the anesthesia input by the way dchz.
 
Top