Safety of dry needling to the upper lumbar spine: a pilot cadaver study

ABSTRACT

Objective: When inserting a dry needle laterally into the upper lumbar spine (L1-L3) there is an increased risk of piercing the kidney; therefore, the objective of this study was to determine a zone of safety for practitioners to needle in the upper lumbar spine.

Methods: Ten cadavers were screened for inclusion. L1 spinous process was identified and confirmed with ultrasound imaging. A digital caliper was used to measure laterally at 1.5 cm, 2.0 cm, and 2.5 cm. Dry needles were inserted maximally at each point and a binary decision, yes or no, was made to determine if bony contact was made. Needle depth and abdominal width measurements were also recorded. Safety of the dry needling procedure was interpreted as such if bony contact was made by the needle. If bony contact was made, then it was assumed that the needle cannot advance further into pleura or kidney.

Results: Forty-four percent of needles did not make bony contact at 2.5 cm lateral of the L1 spinous process, whereas 22% did not make bony contact at 1.5 cm and 2.0 cm. There was a weak to moderate negative correlation between abdominal width measurements and needle depth at 1.5 cm (−0.48) and 2.0 cm (−0.45), and at 2.5 cm (−0.39).

Conclusion: A safety zone of needling less than 2.5 cm is likely safe, but needs to be confirmed with future study. Dry needling 2.5 cm lateral appears more risky due to the higher frequency of not contacting a bony backdrop.

Introduction

Needling therapies such as dry needling and acupuncture involve the insertion of a small gauge monofilament needle into muscle, tendon, and ligaments [1]. Needling is used by physical therapists, athletic trainers, physicians, chiropractors and acupuncturists, and has been shown to be effective at reducing low back pain in the short term [2–6]. It is suggested that the needle can safely be inserted in the paraspinal musculature of the lumbar spine perpendicularly approximately one to two fingerbreadths lateral of the lumbar spinous process [7]. Even though this is a common technique amongst practitioners who perform needling therapies, no study to date has validated the safety of this needling procedure. Of particular importance, the further lateral a needle is inserted into the upper lumbar spine (L1-L3), there is an increased risk of piercing the kidney or missing the targeted muscles and entering the pleura [8–10]. Needling therapies have been accepted as a safe intervention with few adverse events [8,10]. A previous acupuncture study found that 2.2% (4,963 of 229,230 patients) had an adverse event that required medical attention [11]. The most common severe adverse event occurred when needling in the thoracic region near the lung field, where a practitioner may inadvertently cause a pneumothorax [11]. Previous research has reported the incidence of a pneumothorax from needling as less than one in 100,000 patients [11]. Additionally, researchers have identified techniques to maximize safety and reduce the risk of pneumothorax when needling in the thoracic region by needling in a safety zone of 1.0 cm lateral to the thoracic spinous process; however, a safety zone has not been established for the upper lumbar spine [12]. A prospective study of physiotherapists in Ireland reported that mild adverse events from needling occurred in 19.18% of treatments (1463 of 7629), such as bleeding, pain, and bruising [13]. Although no significant adverse events were reported in this study, it was estimated that the risk of a severe adverse event was 0.04% or less (1 of 2543 treatments) [13].

Although researchers, practitioners, and instructors of needling therapies teach caution with needling in the upper lumbar spine, no study has reported adverse events occurring to the kidneys from needling. Adverse events have been reported with renal biopsies, where a portion of the kidney is removed, such as hematuria, pain, arteriovenous fistula or infection due to hematoma [14–18]. Although renal biopsies are more invasive than dry needling, it is imperative that the provider understand all possible risks should the intended muscle be missed, including the risk of kidney penetration. The incidence of adverse events from renal biopsies is also low [14–18].

No study has been published validating the safety of needling in the upper lumbar spine. Some researchers advocate only needling in this area if using ultrasound imaging to confirm the depth of the needle [10]. In the thoracolumbar region, key structures are the diaphragm which separates the abdomen and pelvis from the thorax, the pleura space and kidneys [8]. The goal of this pilot study was to develop a safety zone of needling in the upper lumbar spine to avoid adverse events related to piercing the kidney. Caution should be used when needling by the kidney due to variations in the morphology, as well as differences in the positioning from left to right, with the right kidney positioned more inferior than the left kidney due to the liver in the right quadrant of the abdomen [9,10]. The primary objective of our study was to determine if dry needling in the upper lumbar spine could be safely performed without piercing the pleura or kidney. We have implemented a pilot cadaver study to help inform current practice of dry needling to the upper lumbar spine and guide future cadaver studies assessing dry needling and safety. For the purposes of this study, if the needle contacted bone, it was indirectly implied that the technique was safe since it could not have pierced the pleura or kidney. We hypothesize that bony contact will be made most frequently lateral of the L1 spinous process at 1.5 cm and 2.0 cm, supporting the use of this dry needling technique, and that needling 2.5 cm lateral would result in the least percentage of bony contact being made relative to the other two needle placements. A secondary aim of this study was to determine if cadavers with larger abdominal width measurements had any relationship to the percentage of bony contact or needle depth, and if the depth of penetration is significantly different lateral of the spinous process at 1.5 cm, 2.0 cm and 2.5 cm. Ultimately, this pilot study will be used to guide future dry needling safety studies with cadavers.

Methods

Study design and setting

This was a pilot cadaver study where data were collected cross-sectionally. The study was conducted in an anatomy laboratory of The Ohio State University. Project approval was received in accordance with departmental policy by The Ohio State University Body Donor Program Committee prior to data collection. Ten formalin-embalmed cadavers were randomly selected for use from 16 available cadavers in the anatomy laboratory. Unembalmed cadavers were not available.

Data collection

One researcher (CM) with experience with palpation clinically assessed the cadavers in prone to locate the first lumbar spinous process (L1). He first identified the L4-L5 interspinous space by palpating the tops of the iliac crests and moving midline with his thumbs. Then moving superiorly, he identified the L4 spinous process and counted each spinous process until he reached the L1 spinous process [19,20]. Next, a purple surgical marker was used to mark the L1 spinous process. Another researcher (LB) confirmed the location of the spinous process of L1 with ultrasound imaging. A Philips Lumify portable ultrasound unit (Philips Corp, Andover MA), using a 12–4 MHz linear array transducer was utilized (Figure 1). In long-axis, the spinous process of L5 was first identified with ultrasound imaging and then the transducer was moved superiorly until each lumbar spinous process was identified, and the L1 spinous process was confirmed. The transducer was then rotated at the L1 spinous process into a short axis view and saved for each subject. The spinous process was confirmed by locating the hyperechoic (white) part of the image with the hypoechoic (dark) portion deep (Figure 2).

Figure 1.

A Philips Lumify portable ultrasound with a 12–4 MHz linear transducer (Philips, N.V., 2004–2019) was utilized to confirm the L1 spinous process.

Figure 2.

The L4 spinous process is easily confirmed by identifying the hyperechoic (white) bone with the hypoechoic shadow deep to it in the image.

Ultrasound visualization of the kidney was attempted with the cadaver in a prone position and the transducer placed on the flank of the cadaver [21–23]. The goal was to assess if the needle made contact with the kidney with a lateral view, since we would be unable to dissect the cadaver for visual confirmation. Unfortunately, the kidney could not be visualized with ultrasound due to the embalming process. Additionally, dissection of the kidney was not possible due to the educational requirements of a simultaneously occurring anatomy course.

A digital caliper was then used to measure laterally to the right from the L1 spinous process and a surgical marker was used to mark 1.5 cm, 2.0 cm, and 2.5 cm laterally (Figure 3). HWATO stainless steel Chinese style acupuncture needles (0.25 x 75 mm) were inserted perpendicularly and inserted maximally until the needle could not advance further [24]. Typically, 50 mm needles are used in this area for added safety. However, for the purposes of this study, 75 mm needles were used to gain a better idea of the depth of penetration possible since the primary aim was to determine the depth of the needle in areas where the needle may not contact the bone. Due to the increased risk of right kidney puncture, since it is positioned more inferiorly than the left, the needles were inserted to the right of the L1 spinous process. The needling technique was performed by a physical therapist (CM) with 5 years’ experience, was board certified clinical specialist in orthopedic physical therapy, a Fellow of the American Academy of Orthopedic Manual Physical Therapists, and has completed three dry needling continuing education courses. Bony contact was defined as a perceived reverberation from the needle by the practitioner. A binary decision, yes or no, if the practitioner perceived bony contact. Once the needle was maximally inserted at each point on the cadaver, the digital calipers were used again to measure the vertical distance from the skin to the bottom of the handle of the needle. To determine the depth of needle insertion, this number was subtracted from 75 mm. If the needle was inserted maximally, it was assumed the depth of the needle was 75 mm. Lastly, we took width measurements of the abdomen at the level of L1.

Figure 3.

A digital caliper was then used to accurately measure laterally from the L1 spinous process and surgical marker was used to mark 1.5 cm, 2 cm, and 2.5 cm laterally.

The dry needling procedure was interpreted as being safe if bony contact was made by the needle. It was then assumed that the needle cannot advance further into pleura or kidney and thereby was considered safe. If bone was not contacted by the needle and there was a ‘miss’ then we are less certain of the consequences but know that there is an increased risk of the needle piercing the pleura or kidney. Our methodology does not quantify where the tip of the needle ended up in the scenario where bony contact is not made by the needle.

Data analysis

Statistical analysis for all data was competed using STATA (1996–2019 StataCorp LLC). Descriptive statistics of cadaver characteristics and frequency of dry needling not making bony contact was reported. Correlation analysis was performed between the depth of needle insertion and abdominal width measurements. Interpretation of strength of correlation was based on the following: negligible (0.00–0.10), weak (0.10–0.39), moderate (0.40–0.69), strong (0.70–0.89), and very strong (0.90–1.00) [25]. In order to assess statistical differences between depth of needle penetration at 1.5 cm, 2.0 cm, and 2.5 cm, analysis of variance (ANOVA) was performed between each needle placement with a-priori p-value set at 0.05. Needle depth was assessed to determine the length of penetration until bone was contacted and to assess for differences at each needle placement. If statistical significance was found, then post-hoc t-tests were performed.

Results

Characteristics of sample

Ten cadavers were utilized based on resources available. One cadaver was excluded due to the difficulty of needle penetration. The pilot study was designed after cadavers were embalmed. Future studies, informed by this pilot, will examine the technique using unembalmed cadavers. Another cadaver was excluded from needle depth analysis due to a needle being removed prior to digital caliper measurement, leaving eight cadavers with a full data set. Of the nine cadavers, five were female and four were male. The average age at death was 81 years old (range: 73–89). The mean abdominal width at L1 was 33.67 cm [95% confidence interval 30.36–36.99].

Primary analysis

Four of nine (44.44%) of 75 mm needles did not make bony contact at 2.5 cm lateral compared to 22.2% at 1.5 cm and 2.0 cm lateral, as perceived by the practitioner (Table 1). Interestingly, for two cadavers, the practitioner was able to insert a 75 mm needle to its full depth at 1.5 cm without making bony contact. One of these cadavers had a large torso (35.5 cm) and the other had scoliosis with a left thoracic curve and increased concavity on the right aspect of the spine. The cadaver with scoliosis had an abdominal caliper reading of 25.8 cm. Bony contact was not made at any needle position on this subject (1.5 cm, 2.0 cm, 2.5 cm). Another cadaver yielded bony contact at 1.5 cm and 2.5 cm, but not 2.0 cm lateral from L1 spinous process. Abdominal caliper reading for this cadaver was 35.2 cm.

Table 1.

Primary and secondary analysis.

Measurements	DN at 1.5 cm	DN at 2.0 cm	DN at 2.5 cm
Frequency of DN where no bony contact made (%)	2 (22.2)	2 (22.2)	4 (44.44)
ANOVA: Mean depth of dry needle mm (95% CI)	46.7 (31.0,62.5)	43.9 (28.4,59.33)	49.6 (26.07,73.24)
Correlation between needle depth and abdominal width	−0.48	−0.45	−0.39

Secondary analysis

There was a weak to moderate negative correlation between abdominal caliper measurements and needle depth at 1.5 cm (−0.48), 2.0 cm (−0.45), and 2.5 cm (−0.39) (Table 1). Mean depth of dry needling at 1.5 cm was 46.72 mm (95% CI: 31.0,62.5), 43.86 mm at 2.0 cm (95% CI: 28.4,59.33), and 49.65 mm at 2.5 cm (95% CI: 26.07,73.24). There was no statistically significant difference between any needle depth and location (p > 0.05) (Table 1).

Discussion

Main findings

Based on our operational definition of a needling procedure being safe, inserting a 75 mm needle closer (1.5 cm or 2.0 cm) to the L1 spinous process is safer than inserting the needle more lateral (2.5 cm). This interpretation is based on the frequency of the needle not contacting the bone at each location (1.5 cm, 2.0 cm, 2.5 cm) laterally to the L1 spinous process. For two cadavers, the 75 mm needle did not make contact with the bony lamina of L1 at 1.5 cm lateral to the spinous process. This may have been due to one cadaver having a larger abdominal width (35.5 cm), and the other cadaver demonstrating a left thoracic scoliotic curve. This demonstrates that needling perpendicularly 1.5 cm (approximately one to two fingerbreadths) lateral to the L1 spinous process is relatively safe due to the needle making contact with the lamina of L1. However, this needling technique should be used with caution when patients have scoliosis.

Based on our pilot data, a safety zone of needling less than 2.5 cm lateral to the spinous process may be safe but needs to be confirmed with future study with ultrasound confirmation and/or dissection. The researcher (CM) that performed the dry needling on the cadavers noted that 2.5 cm lateral appeared to be on the border of the L1 transverse process. Since the L1 transverse process is more cranial to the spinous process, there is a chance that if the provider inserting the needle palpated the mid-portion or inferior aspect of the spinous process and inserted the needle 2.5 cm directly lateral of the spinous process then they would likely ‘miss’ the transverse [26]. However, if the provider palpated the superior aspect of the L1 spinous process, and inserted the needle 2.5 cm laterally, this placement was more likely to make contact with the transverse process due to an overlap in plane between the inferior aspect of the transverse process and the superior aspect of the spinous process [26]. Depending on where you palpate L1 spinous process and insert the needle laterally, you may or may not be in line with the transverse process. If you are in line with the transverse process, then dry needling less than 2.5 cm lateral may be an optimal safety zone. However, if you are not in line with transverse process and your needle insertion is in between two transverse processes it may lead to a ‘miss’ of the vertebra. Therefore, a more narrow safety zone should be utilized. There was no significant difference between the needling positions (1.5 cm, 2.0 cm, and 2.5 cm) and needle depth. However, the mean depth and upper tier of the 95% confidence interval were higher [49.6 (26.07, 73.24)] at the 2.5 cm insertion compared to the other needle positions [46.7 (31.0, 62.5); 43.9 (28.4, 59.33)]. It is possible that with a larger sample of data, the trend may lead to a significant difference at the 2.5 cm position. There was also a fair negative correlation (−0.48, −0.45, −0.39, respectively) between needle depth and abdominal width measurements for each needling position. This suggests that needle depth is negatively correlated with abdominal caliper measurements.

Comparison of findings of those reported in the literature

Only clinical commentaries [8,10] and expert opinions [9] have been published regarding the safety of dry needling in the upper lumbar spine. Prevalence and incidence data on adverse events from acupuncture or dry needling have not mentioned adverse events related to the kidney [13]. This study supports clinical commentaries that call for more caution when needling in this region and supports conservative safety recommendation to needle within 1.5 cm from the lateral spinous process. One study highlights the risk of puncturing the kidney when attempting to needle the psoas muscle and emphasizes the importance of screening for urogenital diseases such as upper urinary tract pathology, or kidney infection/inflammation [9]. Another clinical commentary highlights the importance of needling the upper lumbar spine with the use of ultrasound imaging to assess the depth of the needle to ensure the kidney is not punctured [10]. Although the focus of our project was to utilize ultrasound imaging to identify the L1 spinous process, we did try to visualize the dry needle. However, this was challenging with the needles we used, and we could only get a sense of where the needle was in the muscle by moving the needle up and down and visualizing the movement of the lumbar multifidus muscle. Other studies utilizing dry needling and cadavers have been published regarding accuracy of inserting a dry needle without piercing the sciatic nerve [27], and needling in the suboccipital region [28]. More research is needed with cadavers and dry needling to fully assess the safety of dry needling in high risk areas such as the upper lumbar spine in order to substantiate if the techniques practitioners use are safe.

Limitations

This pilot study has several limitations. We had a low number of cadavers due to the time restriction in the lab since the cadavers were being used for an anatomy course at the university. The embalming process likely affected the depth of the needle penetration, which may limit the applicability of the results to a living patient. Also, this cadaver study only assessed the safety of dry needling perpendicularly at the L1 spinous process level. Other techniques are utilized where the needle is placed approximately two fingerbreadths lateral to the spinous process and angled medially toward the lamina of the lumbar vertebrae. We did not assess the safety of this technique, because it appears to be safer than the technique we investigate due to the direction of needle insertion. Finally, bony contact, puncture of kidney, or needle insertion into pleura were not confirmed with ultrasound imaging or dissection. Visualization of the kidney and needle with ultrasound was compromised by the embalming of the cadavers. Future studies should utilize unembalmed cadavers. Furthermore, dissection of the area could not be performed due to the cadavers being used for an anatomy course.

Implications for clinical practice

Dry needling courses advocate the use of 50 mm needles or smaller, inserted one to two fingerbreadths lateral of the L1 spinous process, and needle inclination perpendicular or angled medially to ensure optimal safety [29,30]. However, depending on the fingertip size, two fingerbreadths lateral from the spinous process and needle inserted perpendicularly could be approximately 2.5 cm lateral of L1 spinous process. Our pilot data suggest this may increase the risk of an adverse event due to the decreased frequency of making bony contact at 2.5 cm lateral to the spinous process. Since 2.5 cm, at times, appeared to be on the border of where the L1 transverse process is located, needling less than 2.5 cm laterally is likely safe since the provider is making contact with the lamina of the L1 and less likely to ‘miss.’ Due to the study design, we are limited in our conclusions regarding the appropriateness and safety of the 50 mm depth in this region. However, we did find a weak to moderate negative correlation between needle depth and abdominal caliper measurements. These results suggest that caution should be used for individuals of smaller body types with less muscle and fat. Based on our findings and the current literature, we have highlighted safety recommendations for needling in the upper lumbar spine (Table 2) [8–10].

Table 2.

Safety recommendations for dry needling in upper lumbar spine (L1-L3) [8–10].

Screen for urogenital disease	Upper urinary tract pathology Kidney infection/inflammation Color of urine Incontinence/retention of urine Previous history of urogenital disease Murphy Percussion Test
Dry needle selection	50 mm needle maximum depth If patient has a small body type, consider using a smaller needle
Safety zone of needle insertion	Needle less than 2.5 cm lateral of lumbar spinous process to ensure needle has bony backdrop
Needle direction	Instead of inserting needle perpendicularly, consider needling medially toward the lamina of the lumbar vertebra for optimal safety

Implications for research

This study informs practitioners who needle in the upper lumbar area and can inform future research on the safety of dry needling. Utilizing unembalmed cadavers would enhance a future cadaver study. Furthermore, a different research design may have led to more meaningful conclusions regarding the safety and use of dry needling in the upper lumbar spine, such as implementing the needling procedure and then dissecting to visually confirm if the kidney was pierced at different points laterally. Future research should focus on substantiating a safety zone for practitioners dry needling in the upper lumbar spine. From the experience of the researchers to this project, needling at 1.5 cm appears to be safe, and in most cases, 2.5 cm lateral appears to hit or be just lateral of the transverse process of L1, Whereas needling more medially appeared to make bony contact with the lamina of L1. This could suggest a potential safety zone of needling the lumbar spine of less than 2.5 cm lateral. However, a more rigorous research design should be implemented to substantiate this claim.

Conclusion

Dry needling 1.5 cm to 2.0 cm lateral to the L1 spinous process is a safe intervention in individuals without morphological abnormalities of the spine. Needling 2.5 cm laterally appears more risky due to the higher frequency of not contacting a bony backdrop. Inserting a needle 2.5 cm lateral to spinous process appeared to be on the border of hitting the transverse process of L1 versus not hitting the transverse process. Based on our pilot data, a safety zone of needling less than 2.5 cm is likely safe but needs to be confirmed with a future study.

Biographies

CJ Mansfield is a Fellow of the American Academy of Orthopedic Manual Physical Therapists and is working towards becoming a clinician scientist with a focus on understanding the effects of dry needling on acute and chronic musculoskeletal pathologies.  He is pursuing his PhD training through the School of Health and Rehabilitation Sciences at OSU and works as a physical therapist in the Sports Medicine department at OSU.

M Harr is is an undergraduate student at OSU with an interest in research regarding biomechanics, sports medicine, and  low back injuries.

M Briggs is a clinical assistant professor in the Department of Orthopaedics, researcher at the Sports Medicine Research Institute, lecturer in the Division of Physical Therapy and also serves as the director of the Sports Physical Therapy Residency for Sports Medicine.

J Onate main research focus is developing functional motion assessment research that bridges the gap across research to clinical systems to allow for evidence-based outcomes for aiding individuals to sustain optimal health and performance throughout their careers and lives. He is an associate professor and co-director of the Sports Medicine Movement Analysis and Performance program.

LC Boucher is an assistant professor and her research focuses on pediatric injury biomechanics and more recently has started to explore the role of using musculoskeletal ultrasound in teaching physical exam skills, exploring how tissue responds to stress, as well as applications to aid in clinical decision making.

Acknowledgments

We would like to acknowledge the contributions of individuals and their families who donate their bodies to the University which aid in the education and research of health-care providers.

Disclosure statement

No potential conflict of interest was reported by the authors.