Background: Lumbar punctures are an important diagnostic and therapeutic tool for internists. While rates of major complications are low, success rates in the literature are widely variable and dependent on patient body habitus, patient positioning, operator experience, and use of ultrasound guidance.1 Many trainees report discomfort with their skill in performing and supervising bedside procedures, in particular lumbar punctures. In order to address these concerns, our institution established a medicine procedure service (MPS) in 2014. While review and analysis of more than 1400 lumbar punctures performed by the MPS with trainees since 2014 is underway, we report on success rates and complications of 562 procedures performed between January 1, 2019 – December 31, 2020.
Methods: We conducted a retrospective analysis of all patients referred to the MPS for a lumbar puncture during the study period above. We performed manual chart review for all encounters to identify patient and operator dependent variables and conducted univariable and multivariable analyses for associations with 1) the MPS successfully obtaining cerebrospinal fluid (CSF) and 2) CSF red blood cell number >1000/mm3 (i.e., a traumatic tap). All variables in the univariable analyses with p < 0.20 were included in the multivariable analyses.
Results: Of the 562 lumbar punctures performed, CSF was obtained in 79.4% and 83% were ultrasound-assisted. Interestingly, of the 18.6% of patients referred to neuroradiology for fluoroscopic guidance, only 88% of those procedures were successful. In multivariable analyses, obesity (OR=0.50, p=0.039), lying position (OR=0.41, p=0.020) and pass number 0.42 (OR=0.42, p=< 0.001) were associated with inability to obtain fluid. Use of ultrasound was not significantly associated with increased likelihood of obtaining CSF. However, ultrasound guidance during the procedure was the only variable associated with decreased odds (OR=0.37, p=0.019) of a traumatic tap (CSF RBC < 1,000). There were no incidences of epidural hematoma or spinal cord injury. The most common complication was post dural puncture headache, which occurred in only 4% of procedures. Less common complications were transient radicular pain, tenderness at the puncture site, and severe back pain.
Conclusions: Lumbar punctures performed by an academic MPS with trainee participation had a success rate of almost 80% with low complications. This is the largest series performed primarily by trainees reported in the literature. Variables associated with failure to obtain CSF were obesity, lying position, and number of passes. Ultrasound guidance significantly decreased the odds of traumatic tap. Further review is underway and will include data on operator experience and ultrasound use and association with number of passes performed.
