Atraumatic versus cutting needles in fluoroscopic lumbar puncture

Abstract

Objectives:

Atraumatic needles are known to reduce complication rates of blind lumbar punctures (LP), however, their use in fluoroscopically guided LP is less studied. This study assessed the comparative difficulty of performing fluoroscopic lumbar puncture with atraumatic needles.

Methods:

Single-centre retrospective case-control study comparing atraumatic and conventional or “cutting” needles using fluoroscopic time and radiation dose (Dose Area Product or DAP) as surrogate markers. Patients were assessed from two comparable eight-month periods before and after a policy change to primary use of atraumatic needles.

Results:

105 procedures with a cutting needle were performed in the group prior to the policy change. Median fluoroscopy time was 48 sec and median DAP was 3.14. Of 102 procedures performed in the group after the policy change, 99 were performed with an atraumatic needle and three with a cutting needle after initial attempt with an atraumatic needle. Median fluoroscopy time was 41 sec and median DAP was 3.28. The mean number of attempts was 1.02 in the cutting needle group and 1.05 in the atraumatic needle group. There was no significant difference in median fluoroscopy time, median DAP, or mean number of attempts.

Conclusion:

Fluoroscopic screening time, DAP and mean number of attempts were not significantly increased with primary use of atraumatic needles for LP. Use of atraumatic needles should be considered in fluoroscopic LP given the lower complication rates.

Advances in knowledge:

This study provides new data showing that the use of atraumatic needles does not increase the difficulty of fluoroscopically guided LP.

Introduction

Lumbar punctures (LP) have long been used in medical practice for both diagnostic and therapeutic purposes. ¹ By providing access to the cerebrospinal fluid (CSF), LP may assist with the diagnosis of conditions including subarachnoid haemorrhage, infective meningitis, neurological disorders and intracranial hypertension. ^1–3 Intrathecal access may also be used to inject therapeutic agents such as chemotherapy, or to inject contrast agent to perform myelography or cisternography. ^1–3

Lumbar punctures may be performed using a clinical bedside technique, sometimes referred to as a “blind” approach given the absence of direct imaging guidance. ^2,3 For patients in whom a blind approach has failed, or patients with anatomical features that make a blind approach unfavourable (e.g., difficult spinal anatomy or obesity), image-guided LP may be appropriate and is commonly performed with fluoroscopic or computed tomography guidance in adults. ^2,3 The number of fluoroscopically guided LP procedures performed by radiologists has increased in recent years, ^2–4 and a recent study from Kroll et al suggested radiology departments are now the most common provider of diagnostic LP and the second most common provider of therapeutic LP in North America. ⁴

Both blind and fluoroscopically guided LP have a number of known complications, most notably post-dural puncture headache secondary to CSF leak. ^3,5,6 While the use of a conventional or “cutting” needle has been the existing standard of care for many years, recent studies have shown that the use of an atraumatic (pencil-tip) needle reduces the incidence of post-dural puncture headaches. ⁷ As a result, a clinical practice guideline released from the BMJ Rapid Recommendations panel strongly recommended the use of atraumatic needles in all patients. ⁸ Further, a recent systematic review demonstrated that use of an atraumatic needle in lumbar puncture results in a 62% decrease in the incidence of post-dural puncture headache, as well as decreased need to return to hospital for additional therapy. ⁷ Based on these and similar data, many institutions, including our own, have instituted a policy of primary use atraumatic needles in preference to cutting needles for both blind and fluoroscopic LP.

Some authors have advocated for the use of a cutting needle in fluoroscopically guided LP stating that cutting needles are easier to manoeuvre ⁵ and atraumatic needles may increase the number of puncture attempts, thereby increasing screening time. ^3,9 Minimising procedure duration and fluoroscopic screening time is desirable as it reduces ionising radiation exposure to both patient and proceduralist, and is a well-established principle of radiology practice. ¹⁰ Other authors have argued that as fluoroscopically guided LP has a lower rate of dural leak headache than blind LP, there is no need to use an atraumatic needle. ⁵

At our institution, a policy of primary use of atraumatic needles in fluoroscopically guided LP came into effect in October 2019. This study aimed to determine whether there was any significant difference in procedural difficulty, as indicated by the length of fluoroscopy screening time and radiation dose area product (DAP), for fluoroscopically guided LP with primary use of an atraumatic needle compared to a cutting needle.

Methods

Study design

The Radiology Information System (RIS) was used to find all fluoroscopically guided LP procedures performed at our institution. We identified two comparable eight-month periods before and after the official policy change to using atraumatic needles, with a one-year interval between the periods intentionally selected to avoid any potential learning curve effect of using new equipment. The learning curve effect describes the phenomenon in which performance of a procedure improves with increasing experience. ¹¹ The group before the policy change is referred to as the “cutting needle” group while the group after the policy change is referred to as the “atraumatic needle” group. This study was designed in accordance with the STROBE reporting guidelines for case-control studies (Supplementary Table 1).

Data collection

The records from each of these encounters were examined, and the following data were obtained: age, gender, procedure date, screening time, DAP, and type of proceduralist. The report and images (Figures 1–4) for each procedure were also reviewed to determine the needle type (cutting or atraumatic), level at which the procedure was performed, number of attempts required, referring speciality, and the reason for referral. The patients’ weights were obtained from our institution’s electronic medical records (EMR) system if available.

Figure 1.

Lumbar puncture performed using atraumatic pencil tip needle with introducer at L3-4

Figure 2.

Lumbar puncture performed using cutting needle at L2-3

Figure 3.

Lumbar puncture performed using atraumatic pencil tip needle without introducer at L2-3

Figure 4.

Lumbar puncture performed using cutting needle at L3-4

Reasons for referral were categorised into: diagnostic; diagnostic with opening pressure; follow up of known meningitis; follow up of known idiopathic intracranial hypertension; intrathecal chemotherapy; intrathecal hydrocortisone; and clinical trial.

Inclusion and exclusion criteria

All adult patients above 18 years old were included when fluoroscopic LP was attempted whether or not it was successful. Patients were excluded from the cutting needle group if a cutting needle (22 g Spinal Tap Cannula, Hakko Co., Ltd., Nagano, Japan with “Quincke” style bevel (USD $2.15 each)) was not used in the initial attempt, and excluded from the atraumatic group if an atraumatic needle (22 g RapID^TM Pencil Point Spinal Needle, Smiths Medical, Minneapolis, USA (USD $2.90 each)) was not used in the initial attempt. Patients were also excluded if the procedure was not actually attempted or if procedural information was unavailable.

Procedure details

At our centre, fluoroscopically guided LP is performed in the fluoroscopy procedure room (Artis VD11C, Siemens, Munich, Germany) as a sterile procedure with patients positioned in lateral decubitus on the fluoroscopy table. The point of entry for the needle is determined by palpating the lumbar spine for anatomical landmarks and using a metallic radiopaque marker under lateral and postero-anterior fluoroscopic screening to confirm position. Possible points of entry include the L1/2, L2/3, L3/4, and L4/5 interspinous spaces. The patient’s lumbar region is prepared with chlorhexidine wash (1% chlorhexidine in 70% alcohol) and draped as standard of care. Local anaesthetic (5 ml 1% lidocaine) is administered to the skin and subcutaneous tissues. An introducer needle is used with the atraumatic needles according to proceduralist preference. The spinal needle is then inserted until it appears to be appropriately positioned on fluoroscopic screening and/or puncture through the ligamentum flavum is felt. CSF is then collected and opening pressure may be measured. This technique is similar to the practice parameter published by the American College of Radiologists for myelography and cisternography. ¹² Use of an introducer needle with the atraumatic needle was according to proceduralist preference.

Statistical analysis

Descriptive statistics were used to compare demographic and procedural data. A Mann-Whitney u-test was performed on fluoroscopy time and DAP, a two-tailed unequal variance t-test was performed on age, and a chi-squared test was performed on gender to compare whether there was any statistically significant difference between the cutting and atraumatic needle groups. A p-value of <0.05 was used to determine statistical significance.

Results

A total of 219 records of lumbar puncture were identified during the study dates. Of these, 209 procedures in 159 patients were included in the analysis (Tables 1 and 2). Two records from the cutting needle group were excluded from the study: one because information about the procedure was unavailable, and one as the procedure was not actually attempted (cancelled following a discussion with the referring doctor). Nine records from the atraumatic needle group were also excluded from the study, as an atraumatic needle was not used in the initial attempt. The reasons for the primary use of a cutting needle after the policy change included perceived anatomical difficulties such as severe scoliosis.

Table 1.

Patient demographics, referring unit and reason for referral

	Primary use of cutting needles	Primary use of atraumatic needles	P-value
Number of patients	78	81	n/a
Age
Mean (SD)	53.4 (17.86)	51.0 (16.35)	0.39
Gender
Male/Female Female gender (percentage)	52.6%	44.4%	0.31
Referrer
Haematology	32	34	n/a
Neurology	30	24	n/a
Infectious diseases	17	17	n/a
Neurosurgery	6	1	n/a
Ophthalmology	4	7	n/a
General medicine	9	2	n/a
Respiratory	3	3	n/a
Medical oncology	2	3	n/a
Psychiatry	2	1	n/a
Gastroenterology	1	0	n/a
Emergency	1	1	n/a
Rheumatology	0	3	n/a
Renal	0	1	n/a
Cardiothoracics	0	1	n/a
Endocrinology	0	1	n/a
Reason for referral
Diagnostic	41	35	n/a
Intrathecal chemotherapy	28	34	n/a
Diagnostic with opening pressure	19	21	n/a
Follow-up of known meningitis	12	3	n/a
Follow-up of known malignancy	1	6	n/a
Follow-up of known idiopathic intracranial hypertension	4	3	n/a
Intrathecal hydrocortisone	1	0	n/a
Clinical trial	1	0	n/a
Weight
Mean (SD) in kilograms	90.7 (26.11)	92.1 (19.22)	0.89
Procedure completion
Procedure completed	105	102	n/a
Procedure attempted but not completed	2	0	1.00

Table 2.

Procedure details

	Primary use of cutting needles	Primary use of atraumatic needles	P-value
Number of procedures	105	102	n/a
Needle type (final)
Cutting	105	3	n/a
Atraumatic	0	99	n/a
Fluoroscopy time
Median (IQR) in seconds	48 (63)	41 (51.5)	0.271
Dose area product
Median (IQR)	3.14 (7.23)	3.28 (4.71)	0.542
Spinal level access on first attempt
L1/2	0	2	n/a
L2/3	37	46	n/a
L3/4	59	49	n/a
L4/5	9	5	n/a
Number of attempts
one attempt	103	98	n/a
two attempts	2	3	n/a
three attempts	0	1	n/a
Mean	1.02	1.05	0.30

107 procedures were performed on 78 patients in the cutting needle group with a mean age of 53.4 years (17.86). 52.6% of the patients were female. In the atraumatic needle group, 102 procedures were performed on 81 patients with a mean age of 51 years (16.35). 44.4% of the patients were female. All procedures were performed primarily by radiology trainees (registrars).The most common referring specialties for fluoroscopically guided LP in both groups were haematology (including bone marrow transplant and haemostasis thrombosis subunits), neurology and infectious diseases. The most common reasons for referral in both groups were diagnostic, intrathecal chemotherapy, and diagnostic with opening pressure. The mean weight of participants in the cutting needle group was 90.7 kg (data available for 54 out of 78 individuals), while the mean weight of participants in the atraumatic needle group was 92.1 kg (data available for 61 out of 81 individuals). The most common spinal level the procedure was performed at was L3/4 in both groups, and most procedures required a single attempt.

Of the 107 procedures included in the cutting needle group, 105 procedures were completed using a cutting needle and two were attempted but not completed – one because the patient was unable to tolerate the procedure and the other because the procedure was ultimately technically unsuccessful. The median fluoroscopy time was 48 sec (interquartile range IQR: 63 sec) and the median DAP was 3.14 (IQR: 7.23). Two attempts were required for two cases in the cutting needle group (mean number of attempts 1.02).

Of the 102 procedures in the atraumatic needle group, 99 procedures were performed with an atraumatic needle and three were performed with a cutting needle after an initial attempt with an atraumatic needle. The median fluoroscopy time was 41 sec (IQR: 51.5 sec) and the median DAP was 3.28 (IQR: 4.71). Two attempts were required for three cases in the atraumatic needle group, and three attempts in one case (mean number of attempts 1.05).

There was no significant difference in the median fluoroscopy time (p = 0.271), median DAP (p = 0.542), or mean number of attempts (p = 0.30) between the cutting and atraumatic needle groups.

Discussion

An increasing number of LP are being performed in radiology practices, with the main benefits of using imaging guidance including the lower likelihood of a traumatic tap ¹³ and higher success rate in patients with unfavourable anatomy for a blind bedside approach. ¹⁴ While fluoroscopy, computed tomography (CT) and ultrasound are used as modalities for imaging guidance, fluoroscopy is the main modality in adult LP. ¹⁵

All radiological procedures involve radiation exposure to the patient and proceduralist, and therefore the ALARA (As Low As Reasonably Achievable) concept underpins practices and every effort is made to minimise screening time in fluoroscopic procedures. The level of radiation exposure for a procedure varies both due to procedural factors such as the proceduralist’s experience level and equipment used, ^4,16,17 and patient factors including obesity ¹⁴ and anatomical complexity (including facet joint arthropathy, degenerative disc disease, and scoliosis). Studies have suggested methods of reducing radiation exposure in fluoroscopic LP including using a smaller field of view, reducing the use of continuous fluoroscopy and minimising the number of exposure X-ray images taken as these require higher radiation doses. ²

The use of atraumatic needles is well established to have a lower rate of post-dural puncture leak headache than cutting needles. ⁷ This study showed no statistically significant difference in the fluoroscopy time, DAP or mean number of attempts required for fluoroscopically guided LP using a cutting needle compared with fluoroscopically guided LP primarily using an atraumatic needle. There was also no statistically significant difference in the mean age or weight of the patients. There was a higher percentage of female patients in the cutting needle group whilst a higher percentage of male patients in the atraumatic needle group. There is no existing published data comparing fluoroscopy times for cutting and atraumatic needles in fluoroscopically guided LP; however, fluoroscopy time has been previously benchmarked at 0.48 min (28.8 sec) for a patient of normal body mass index (BMI). ¹⁴ Other studies have found that for a radiology trainee, the mean fluoroscopy time was 1.09 ± 0.65 min, although no information was provided on the type of needle used. ¹⁶ Given that fluoroscopically guided LP procedures at our institution are almost exclusively performed by radiology trainees, the median fluoroscopy times (48 and 41 sec) for both needle types in our study are comparable.

Our finding that there was no statistically significant difference in the mean number of attempts required is supported by recent literature on blind LP, which found no significant difference between the two needle groups. ⁷ This is contrary to concerns raised by some authors that the use of an atraumatic needle in fluoroscopically guided LP may increase the difficulty and number of attempts required. ^3,5,9 A retrospective review of blood patch rates for atraumatic and cutting needles in fluoroscopically guided LP also found that use of an atraumatic needle was less likely to require multiple attempts, and that if multiple attempts are required, use of an atraumatic needle results in a lower incidence of dural puncture headache and requirement for blood patch compared to multiple attempts with a cutting needle. ¹⁸

This paper provides new data comparing screening time and DAP in fluoroscopically guided lumbar puncture with cutting and atraumatic needles. However, limitations include the retrospective design, non-randomisation, and potential confounding factors. Firstly, the needle type was not documented in a small number of procedures and was determined on image confirmation alone. When using the atraumatic needle, proceduralists variably used an introducer needle withthe atraumatic needle according to personal preference, however the rate at which this occurred was not captured in the data. A cutting needle was used primarily in a small number of procedures after the official policy change to primary use of an atraumatic needle, which could reflect selection bias as the patient’s habitus may have influenced these decisions. Furthermore, although there was no significant difference in mean weight between the two patient groups, data were incomplete, which further confounds the utility of DAP as an imperfect estimator of radiation dose. Lastly, it should be noted that screening time does not capture total procedure time. A prospective and randomized study would obviate many of these biases.

Conclusion

This study showed that procedural difficulty, as assessed by screening time, DAP, and number of attempts, was not significantly different with use of atraumatic needles for LP under fluoroscopic guidance compared to conventional cutting needles. This supports the primary use of atraumatic needles, which are known to reduce the likelihood of post-dural puncture leak headaches.