Safety and anatomical accuracy of dry needling procedures in musculoskeletal system: a systematic review of cadaveric studies

ABSTRACT

Introduction

In recent years, invasive physiotherapy techniques have seen a significant rise in use for the treatment of musculoskeletal pain. However, there is currently no standardization in the anatomical landmarks, needle length, and insertion angles applied across different muscles, which is critical for ensuring both safety and reliability, especially in non-ultrasound-guided interventions.

Objective

To analyze the results of studies evaluating invasive puncture techniques performed on cadaveric specimens and to assess their methodological quality.

Methods

A systematic review was conducted according to PRISMA guidelines. A comprehensive search was performed in the MEDLINE, Scopus, Web of Science, CINAHL, and SPORTDiscus databases, including all records up to 23 December 2024. Studies were included if they assessed the anatomical safety and procedural accuracy of invasive physiotherapy techniques on cadavers. The quality of reporting was assessed using the QUACS scale, and risk of bias was evaluated with the JBI Critical Appraisal Checklist for Systematic Reviews.

Results

Twenty-one cadaveric observational studies were included. All studies demonstrated anatomically safe approaches, with no reported injury to adjacent neural or vascular structures. However, there was variability in methodological quality and consistency in reporting across studies. The mean QUACS score across the included studies was 8.95 out of 13, corresponding to an average methodological quality of 68.84% (30.76% to 92.30%).

Conclusion

The current evidence supports the anatomical safety of invasive physiotherapy techniques applied to cadavers. Nevertheless, the development and implementation of standardized, reproducible protocols are necessary to improve the consistency, safety, and clinical applicability of these interventions, particularly when performed without ultrasound guidance.

1. Introduction

In recent years, the use of invasive procedures in physiotherapy has increased significantly [1,2]. Commonly used techniques like dry needling, percutaneous electrolysis, and electrical nerve stimulation involve inserting fine, non-beveled, surgical-grade stainless steel needles into soft tissues to minimize tissue damage [3].

Dry needling is one of the most frequently used invasive techniques in physiotherapy [2], especially for treating myofascial pain syndromes [4]. It involves inserting a needle into a painful spot within the muscle, known as a myofascial trigger point [1], described as a hyperirritable nodule within a taut band of skeletal muscle associated with motor dysfunction and characteristic referred pain [5]. Invasive techniques can target musculoskeletal impairments, or even areas adjacent to peripheral nerves or neurovascular bundles [3].

Depending on the technique applied, these interventions may be guided by ultrasound or performed without imaging. Non-ultrasound-guided dry needling relies on the clinician’s palpation skills to identify taut bands, local tenderness, and the elicitation of a local twitch response [4]. Ultrasound-guided approaches enable visualization of soft tissue layers within the patient’s pain region, which may enhance anatomical accuracy and safety [5]. While both techniques are widely used and valued in clinical practice, ongoing efforts to improve standardized protocols are essential to ensure reproducibility, precision, and safety [5].

Cadaveric studies represent a valuable method for developing safe and precise clinical protocols, enabling the identification of anatomical safety zones for needle placement by trained professionals [5]. The design of methodologically reproducible invasive approach strategies increases the safety of invasive approaches. These strategies are based on the clear design of issues such as the definition of needle entry points, anatomical landmarks, needle size, and entry angle, among others. In this context, this review implements the QUACS tool, a validated checklist for evaluating methodological quality in cadaveric observational studies [6] and integrates the Joanna Briggs Institute (JBI) Reviewer’s Manual for risk of bias assessment [7].

Therefore, the main objective of this systematic review is to summarize the available literature on cadaveric studies evaluating the anatomical accuracy and safety of deep dry needling techniques in the musculoskeletal system, both with and without ultrasound guidance.

2. Methods

2.1. Study design

The review adhered to the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines [8]. The study protocol was registered in the PROSPERO system under registration number CRD42025631084.

2.2. Search strategies

A comprehensive literature search was performed using the electronic databases MEDLINE, Scopus, Web of Science, CINAHL, and SPORTDiscus, including all records up to 23 December 2024. The search aimed to identify studies investigating the safety, precision, and anatomical validation of invasive procedures targeting the musculoskeletal system. The search strategy was designed in collaboration with an expert health sciences librarian. Reference lists from the selected studies and relevant reviews were manually examined to identify additional relevant articles. Duplicate entries were detected and removed using a combination of automated tool (Rayyan QCRI, Doha, Qatar) [9] and manual verification.

The inclusion criteria were defined using the PICOS as follows [8]:

Participants: Studies involving adult cadaver models or evaluations of musculoskeletal structures.

Intervention: Procedures such as dry needling, percutaneous electrolysis, or other needle-based invasive techniques.

Comparator: No specific comparator was required; studies may include comparisons between techniques, evaluators, or populations.

Outcomes: The primary outcomes assessed were measures of safety, accuracy, or anatomical validation of the procedures.

Study Design: Observational, cadaveric, or experimental studies were included to capture the relevant data.

A detailed description of the search strategies for each database is provided in Supplementary material S1.

2.3. Eligibility criteria

Eligibility criteria included cross-sectional or observational cadaveric studies published in English between 1 January 2000, and 31 December 2024. Studies eligible for inclusion focused on the application of invasive needle techniques, such as dry needling, percutaneous electrolysis, or other similar approaches, to muscles of the musculoskeletal system in cadaveric specimens. The review emphasized studies that evaluated anatomical accuracy, safety regarding neurovascular structures, and procedural feasibility within anatomy lab settings.

Studies were excluded if they were systematic reviews, narrative reviews, scoping reviews, meta-analyses, book chapters, editorials, or letters to the editor. Additionally, studies involving interventions not relevant to the musculoskeletal system (e.g. non-muscle-focused procedures) or pharmaceutical approaches, such as steroid injections, were not included. Studies without sufficient procedural descriptions or those not focused on invasive techniques were also excluded.

2.4. Data extraction

The process of screening, selecting studies, and extracting data for this systematic review was meticulously organized to ensure precision and consistency. Two independent researchers (MSM and AGF) reviewed the titles and abstracts of all identified studies. Duplicate records were eliminated using a combination of automated tools and manual cross-checking. An initial eligibility assessment was conducted based on titles and abstracts, followed by an in-depth examination of full-text articles to determine their suitability according to the inclusion criteria. Any disagreements between reviewers during the selection phase were resolved through discussion or, when necessary, by consulting a third reviewer (JAP).

Data extraction was carried out independently by two researchers using a standardized extraction form designed to align with the objectives of the review. The extracted information included details such as study characteristics (e.g. author, year, country, sample size, and cadaver preservation techniques), methodological aspects (e.g. needle type, insertion methods, depth of penetration, and anatomical regions examined), and procedural data (e.g. targeted muscles, approach utilized, and anatomical landmarks).

Outcomes related to procedural safety (e.g. avoidance of neurovascular structures and identification of potential adverse events) and accuracy (e.g. precision in needle placement) were also recorded. Validation techniques, such as cadaveric dissection or imaging confirmation, were documented to ensure a thorough evaluation of methodological consistency. Any discrepancies in the extracted data were resolved through consensus, maintaining the integrity and reliability of the dataset for the subsequent narrative synthesis.

2.5. Assessment of methodological quality and risk of bias

The Quality Appraisal for Cadaveric Studies scale (QUACS scale) was used to critically evaluate the methodological quality assessment of the studies that were part of this systematic review. This checklist-based scale comprises thirteen items, each scored 0 (no/not stated) or 1 (yes/present), with points awarded solely when criteria were unequivocally met. The highest possible score is 12 or 13, depending on the study’s substance, because one item – ‘Applied statistics are appropriate’ – was not always relevant. The quality rating was provided as a percentage value (reached score/maximum score (%)) [6].

The risk of bias in the included studies was assessed using the Joanna Briggs Institute (JBI) Critical Appraisal Tool, a structured framework specifically designed for observational and descriptive studies. This tool evaluates key domains, such as the clarity of inclusion criteria, appropriateness of study design, sampling strategies, reliability of data collection methods, adequacy of sample sizes, and validity of outcome measures. Each study was systematically appraised against these criteria using a binary scale (‘Yes,’ ‘No,’ or ‘Unclear’) to determine whether the criterion was fully met, unmet, or insufficiently described. Higher scores indicated lower risk of bias and greater methodological rigor. Studies were categorized as high, moderate, or low quality based on the number of criteria met. Detailed scoring results were documented in a summary table [7].

For the agreement between reviewers in the QUACS and JBI assessment, Cohen’s kappa coefficient was calculated to quantify inter-rater reliability.

3. Results

The initial search across multiple databases (MEDLINE, CINAHL, SPORTDiscus, SCOPUS, and Web of Science) identified a total of 567 records. After removing duplicates (n = 115), 452 records remained for screening. Of these, 405 were excluded after title and abstract review, leaving 47 reports for full-text screening. After detailed examination, 26 reports were excluded for not meeting the inclusion criteria, leaving 21 studies included in the systematic review (Figure 1).

Figure 1.

Prisma flow diagram.

3.1. Targeted structures in the included studies

Reviewed studies investigated muscle and tendon structures across different anatomical regions, including cranio-cervical, thoracolumbar, abdominal, upper and lower limb, and tendinous structures (Table 1).

Table 1.

Characteristics of included cadaveric studies assessing dry needling.

Author	Study type	Muscle/Structure	Objective	Intervention	Needle	Sample	Outcome measurement	Results
Fernández-de-las-Peñas et al. [10]	Descriptive cadaveric and human study	Obliquus capitis inferior muscle	Evaluate needle penetration accuracy using ultrasound and cadaveric dissection.	40 mm sterile needles inserted in 10 healthy subjects and 5 fresh cadavers. Placement verified via ultrasound in live subjects and dissection in cadavers.	Sterile stainless steel needles with a cylindrical plastic guide, 40 mm long and 0.32 mm gauge.	10 healthy subjects without cervical pain and n = 5 fresh cadavers. Formalin, alcohol and lysoformin preservation	Needle placement accuracy, distance to C2 lamina, relation to greater occipital nerve, pain response.	90% accuracy in needle placement. Mean distance to C2 lamina: 2 mm. 50% reduction in pain. No adverse effects reported.
Mesa-Jiménez et al. 2020 [11]	Descriptive cadaveric and human study	Medial pterygoid muscle	Determine the accuracy of dry needling placement in the medial pterygoid muscle.	Dry needling validation performed on 5 cadaveric heads and 5 patients with temporomandibular disorder (TMD). Needles (40 mm) inserted perpendicularly at the inferior angle of the mandible, advancing superiorly up to 30 mm. Placement verified by dissection in cadavers and self-reported pain in patients.	Sterile stainless steel needles with a cylindrical plastic guide, 40 mm long, 0.32 mm gauge	n = 5 fresh cadaver heads and 5 live patients diagnosed with DTM. Cryopreserved preservation. Fresh	Needle placement accuracy, pain response, and pain characteristics in patients.	Successful needle penetration into the medial pterygoid muscle in all cadaveric cases, without neurovascular damage. In patients, 80% (4 out of 5) reported referred pain in the mouth and temporomandibular region during needling, supporting its therapeutic potential.
Pérez-Bellmunt et al. 2022 [12]	Descriptive cadaveric study	Tibialis posterior muscle	Evaluate needle penetration accuracy and safety in the tibialis posterior muscle using cadaveric models.	Dry needling on 11 cryopreserved legs using posterior/medial (mid-leg) and anterior (upper-leg) approaches. Precision assessed via anatomical dissection and photometric analysis.	Sterile stainless steel needles with a cylindrical plastic guide, 70 mm long, 0.32 mm gauge	n = 11 cryopreserved legs. Conservation: Cryopreservation.	Distance from needle tip to tibial nerve (posterior) and deep fibular nerve (anterior), distance to nearest vascular branch.	100% accuracy in muscle targeting. Posterior approach had significantly lower nerve and vascular distances, suggesting it is safer.
Rodríguez-Sanz et al. 2021 [13]	Cadaveric study	Popliteus muscle	Assess accuracy and safety of dry needling in the popliteus muscle.	Manual palpation and identification of anatomical landmarks. Dry needling performed with a 50 mm needle at an anterior angle toward the posterior tibia. Latex injected to mark needle position for post-intervention analysis.	Sterile stainless steel needles with a cylindrical plastic guide, 50 mm long, 0.32 mm gauge	n = 11 legs. Preservation: Cryopreserved	Needle depth (avg. 25.7 mm ±6.7 mm), distance from needle tip to tibial nerve (avg. 17 mm ±6mm), distance to nearest vascular branch (avg. 15 mm ±0.7 mm).	91% accuracy in targeting the popliteus muscle. Safe insertion margins: tibial nerve (17 mm), vascular structures (15 mm). Dry needling in the popliteus muscle was safe and effective when performed by an experienced clinician.
Calderón-Díez et al. 2022 [14]	Cadaveric and human study	Distal biceps tendon (electrolysis)	Validate the safety and effectiveness of percutaneous electrolysis in the distal biceps tendon.	Needle insertion guided by ultrasound and anatomical dissection in 5 cadaveric elbows and 10 live subjects. Two approaches: anterior (extended elbow) and posterior (flexed elbow).	Solid filiform needle, 25 × 0.3 mm	10 healthy subjects/n = 5 samples of cadaver elbows. Preservation: embalmed in Thiel technique.	Distance between the distal biceps tendon and the brachial artery, demographic characteristics, ultrasound visualization.	Posterior approach had a greater distance to the brachial artery, suggesting lower vascular risk. Needle placement was successful in all cases. Ultrasound guidance recommended for clinical use.
Belón-Pérez et al. 2022 [15]	Cadaveric and ultrasound validation study	Supinator muscle – Arcade of Frohse	Validate percutaneous electrolysis in the supinator muscle at the Arcade of Frohse for radial tunnel syndrome treatment.	25 × 0.3 mm needle inserted under ultrasound guidance in supination and pronation positions to evaluate technique accuracy.	Solid filiform needle, 25 × 0.3 mm	5 healthy volunteers/n = 6 arms (3 cadavers). Preservation: embalmed in Thiel technique.	Needle insertion accuracy, distance from the needle tip to deep radial nerve in supination and pronation.	100% accuracy in reaching the supinator muscle. Mean needle-to-neurovascular bundle distance: 15.3 mm (supination), 11.2 mm (pronation). Ultrasound guidance ensured safe and precise needle placement.
Calderón-Díez et al. 2022 [16]	Cadaveric and ultrasound validation pilot study	Achilles tendon	Validate the safety and effectiveness of percutaneous electrolysis in the Achilles tendon interface.	Ultrasound-guided electrolysis in 10 healthy volunteers and 10 cadaveric legs. Needle inserted 5 cm from calcaneal insertion, avoiding the sural nerve. Performed by an experienced physiotherapist.	Filiform solid dry needle 25 × 0,3 mm (AguPunt)	10 healthy volunteers and n = 10 legs. Storage: Fresh	Distance between Achilles tendon and sural nerve, ensuring correct needle placement.	100% accuracy in electrolysis application. No sural nerve or minor saphenous vein damage. Mean tendon-to-nerve distance: 5.28 ± 0.07 mm. No significant differences between live subjects and cadavers (p = 0.268).
Williams et al. 2023 [17]	Cadaveric study	Multifidus – Thoracolumbar region	Assess the risk of neurological injury during dry needling in the thoracolumbar region (T12-L1) and evaluate ultrasound guidance for improved accuracy and safety.	Dry needling performed on a fresh, lightly embalmed 88-year-old female cadaver. Ultrasound guidance used to monitor needle insertion.	Filiform solid dry needle 50 × 0,3 mm (AguPunt)	n = 1 corpse. Preservation: light embalming	Needle penetration depth (40–100 mm), insertion angle (0°-45°), and entry location (1.5 cm lateral to the spinous process, within one finger width of midline).	Ultrasound-guided needling improved needle placement accuracy and reduced the risk of neurological injuries compared to traditional techniques. Needle penetrated the ligamentum flavum in all cases, indicating access to deep structures. A significant number of insertions entered the spinal canal, highlighting safety concerns. Depth and angle variation suggested that clinical experience affects precision and safety.
Fernández-de-las-Peñas et al. 2022 [18]	Descriptive cadaveric study	Pronator teres muscle	Determine the safety and accuracy of dry needling in the pronator teres muscle.	Dry needling performed on cadavers with a 30 mm × 0.32 mm needle. Depth and neurovascular proximity were recorded to evaluate safety.	Stainless steel filiform needles with a plastic guide cylinder, 30 mm long, 0.32 mm gauge	n = 20 arms (10 cadavers). Conservation: Cryopreserved	Needle penetration depth, distance to median and ulnar nerves, distance to brachial artery, insertion angle.	100% accuracy in targeting the pronator teres muscle. No neurovascular structures were compromised. Mean distances: 1.5 cm to median nerve, 2.0 cm to ulnar nerve. No significant sex-based differences in measurements.
Pérez-Bellmunt et al. 2023 [19]	Cadaveric study	Pronator quadratus muscle	Assess the accuracy and safety of dry needling in the pronator quadratus muscle.	Needle inserted into 10 cryopreserved forearms, 3 cm proximal to the ulnar styloid process, directed anteriorly. Distance to neurovascular structures measured to assess safety.	Solid filiform needle 30 × 0.32 mm	n = 10 forearms. Preservation: cryopreserved	Needle penetration depth, distance to ulnar nerve, ulnar artery, median nerve, and anterior interosseous neurovascular bundle.	90% accuracy in needle penetration. No neurovascular structures were punctured. Mean distances: ulnar nerve (15.1 ± 4.8 mm), ulnar artery (15.6 ± 7.6 mm), median nerve (11.2 ± 3.3 mm), anterior interosseous bundle (4.9 ± 1.4 mm). Further clinical studies needed to confirm safety and efficacy.
Kearns et al. 2018 [20]	Cadaveric study	Piriformis muscle	Assess the accuracy and safety of dry needling in the piriformis muscle using cadaveric specimens.	Dry needling performed on 19 hips of unembalmed cadavers. Needle placement confirmed via dissection to ensure correct targeting and avoid sciatic nerve injury.	Dry needle puncture needles with a needle length between 50–100 mm	n = 19 hips (10 cadavers) Preservation: unembalmed, cryopreserved	Fat thickness, gluteus maximus thickness, perpendicular distance from piriformis to sciatic nerve, BMI, age, height, and weight.	Successful needle placement in 17/19 cases. No sciatic nerve perforation. Fat and gluteus maximus thickness influenced insertion difficulty, highlighting the need for anatomical considerations in clinical practice.
Mansfield et al. 2020 [21]	Pilot cadaveric study	Upper lumbar spine (L1-L3)	Determine a safe zone for dry needling in the upper lumbar spine (L1-L3) and assess the relationship between abdominal width and needle penetration depth.	Needle insertions in 10 cadavers at different lateral distances (1.5 cm, 2.0 cm, 2.5 cm) from the spinous process. Depth and proximity to bone assessed to identify a safe insertion distance.	HWATO Stainless Steel Chinese Style Acupuncture Needles (0.25 ×75 mm)	n = 10 cadavers. Preservation: Embalmed with formalin	Needle depth at different lateral distances, percentage of bone contact, abdominal width.	44.44% of 75 mm needles did not contact bone at 2.5 cm laterally, compared to 22.2% at 1.5 cm and 2.0 cm. Mean depths: 46.7 mm (1.5 cm), 43.9 mm (2.0 cm), 49.6 mm (2.5 cm). Negative correlation be
Fernández-de-las-Peñas et al. 2017 [22]	Cadaveric and human study	Cervical multifidus muscle	Determine whether a needle can accurately reach the cervical multifidus muscle during dry needling or acupuncture.	Needle inserted perpendicular to the skin, 1 cm lateral to the spinous process at C3-C4, advancing in a posterior-to-anterior direction with a 10° inferior-medial angle. Ultrasound used to guide insertion.	Sterile stainless steel needles 40 mm long and 0.32 mm gauge with a cylindrical plastic guide.	5 patients with mechanical neck pain/n = 2 cadavers. Preservation:fresh	Needle positioning in the cervical multifidus via ultrasound, anatomical assessment of surrounding structures, patient response to treatment.	Needle successfully reached the cervical multifidus muscle and positioned correctly against the vertebral laminae, ensuring protection of spinal structures. Clinical effects in patients were not detailed.
Mesa-Jiménez et al. 2015 [23]	Cadaveric anatomical study	Lateral pterygoid muscle	Determine if a needle can reach the lateral pterygoid muscle during dry needling.	Needles inserted into the superior (sphenoid) and inferior (pterygoid) heads of the lateral pterygoid muscle in two fresh cadavers. The superior head was accessed via the zygomatic process, and the inferior head between the mandibular condyle and coronoid process. Needle placement confirmed by dissection.	Needles with cylindrical plastic guide, length 50 to 60 mm, gauge 0.32 mm.	n = 2 cadavers. Preserved: Fixed in a mixture of formalin, alcohol and lysoformin.	Needle reach to superior and inferior heads of the lateral pterygoid muscle, assessed via stepwise dissection and osteotomy of the mandibular ramus.	The needle successfully reached both heads of the lateral pterygoid muscle. The superior head was accessed through the sphenoid, and the inferior head through the mandibular region. Final dissection confirmed proper needle positioning in all cases.
Hannah et al. 2016 [24]	Descriptive comparison study	Lumbar multifidus muscle	Compare the posteroanterior needle insertion angle with the commonly used inferomedial angle in dry needling of the lumbar multifidus muscle.	A comparative analysis of two needle insertion angles (inferomedial and posteroanterior) was conducted in cadaveric specimens. A total of 8 needle placements were performed, and the trajectory and final resting position of the needle in the lumbar lamina were observed.	A Dritz® needle (12.7 cm in length, 1.09 mm in diameter)	n = 8 (4 cadavers). Preservation: embalmed with Maryland embalming solution, containing 5.6% formaldehyde, 27.8% phenol, 33.3% methanol and 33.3% glycerin, mixed with 2 parts water.	Needle trajectories, final placement in the lumbar lamina, and the ability of the needles to penetrate all layers of the lumbar multifidus muscle. Additionally, the effectiveness of the inferomedial and posteroanterior angles in reaching the target tissue was evaluated.	In all cases, the needle successfully passed through the thoracolumbar fascia and all layers of the lumbar multifidus, resting on the bony structure of the lumbar lamina. All four inferomedial insertions effectively reached the target tissue, confirming that both insertion techniques were successful in achieving precise tissue dissection.
Hannah et al. 2023 [25]	Observational study	Piriformis muscle	Compare the safety and accuracy of three different needle insertion sites in the piriformis muscle using human donors.	Dry needling was performed on preserved human donors at three insertion sites: lateral, middle, and medial. Needles were assessed for penetration of the piriformis muscle and avoidance of the sciatic nerve.	Dritz® 1.1 mm diameter, 12.7 cm length	n = 14 (10 cadavers) Preservation: Maryland embalming solution containing 5.6% formaldehyde, 27.8% phenol, 33.3% methanol and 33.3% glycerin, mixed with two parts water.	Needle penetration into the piriformis muscle, impact on the sciatic nerve, insertion depth, and surrounding structural integrity.	Lateral insertion: penetrated piriformis in 3/14 cases without affecting the sciatic nerve. Medial insertion: penetrated both piriformis and sciatic nerve in 11/14 cases. Middle insertion: penetrated piriformis in 11/14 cases with lower sciatic nerve involvement (21.4% vs. 78.6% for medial).
Fernández-de-las-Peñas et al. 2021 [26]	Cadaveric study	Supinator muscle	Determine the safety of dry needling in the supinator muscle by measuring the distance between the needle and surrounding neurovascular structures in a cadaveric model.	Dry needling performed on the supinator muscle using sterile filiform acupuncture needles (30 mm). Needle inserted dorsally to ventrally at the upper third of the radius, 4 cm distal to the lateral epicondyle, with an average penetration depth of 16.4 mm. Blue latex injected to verify needle placement.	Sterile stainless steel filiform needles, 30 mm long and 0.32 mm gauge with cylindrical plastic guide	n = 10 Conservation: Cryopreserved	Needle insertion depth, distance from needle tip to deep and superficial branches of the radial nerve, anthropometric characteristics of cadavers (height, weight, forearm circumference).	100% accuracy in needle penetration into the supinator muscle. No neurovascular structures were perforated. Mean distances from the needle tip to deep and superficial branches of the radial nerve were 7.8 mm and 8.6 mm, respectively. Differences observed between male and female cadaver characteristics.
McConnel et al. 2024 [27]	Cadaveric study	Rectus abdominis muscle	Evaluate the accuracy of dry needling in the rectus abdominis muscle without penetrating the peritoneum.	Two physiotherapists performed dry needling on a cadaver, each attempting 10 insertions into the rectus abdominis. Two common renal artery needling approaches were used. Ultrasound verified needle depth and safety.	Filiform solid dry needle 40 × 0,25 mm (AguPunt)	n = 1 Conservation: Light embalming	Needle placement accuracy, safety of insertions (classified as safe or unsafe), ultrasound verification of needle position, tissue response (e.g. local contraction).	70% of attempts were classified as safe, while 30% were unsafe. Accurate rectus abdominis needling without peritoneal penetration occurred in 55% of attempts.
Kearns et al. 2022 [28]	Cadaveric study	Inferior oblique muscle of the head	Compare the influence of clinical experience on the accuracy and safety of dry needling in the C2 laminar arch using cranial-medial and caudal-medial techniques in unembalmed cadavers.	Three physiotherapists inserted 3 dry needles (50 mm) toward the C2 laminar arch, targeting the inferior oblique muscle of the head. Needle placement was visualized using ultrasound to ensure accuracy and safety.	Dry needling needles 50 mm long	n = 14 (7 cadavers) Conservation: Cryopreserved	Examiner experience level (novice, experienced, expert), needle insertion technique (cranial-medial or caudal-medial), accuracy of needle placement in the posterior laminar arch of C2.	Examiner experience and insertion direction influenced accuracy. Novices had 73.8% accuracy in caudal-medial insertions but only 14.3% in cranial-medial. Experts had 71.4% accuracy in caudal-medial and 66.7% in cranial-medial. Significant probability ratios were found between expert and experienced examiners in the cranial-medial direction, with an 86.6% correct prediction rate for experts.
Kearns et al. 2025 [29]	Cadaveric study	Rectus capitis posterior major	Establish guidelines on needle length and insertion angle to mitigate risks during dry needling of the rectus capitis posterior major.	Cadavers were placed prone for cervical access. A midline incision was made to dissect the skin, adipose tissue, and muscles down to the C2 spinous process. Depth measurements were taken using a metal ruler, and anatomical structures were dissected. Three angles were measured: posterior (from the frontal plane), lateral (from the midline), and horizontal distance from the external occipital protuberance to the lateral rectus capitis posterior major.	Needle for dry needling < 40 mm in length	n = 15 Preservation: embalmed with a mixture of water, ethanol, formaldehyde, ethylene glycol, phenol and glutaraldehyde.	Depth from skin to C2 spinous process, posterior angle relative to frontal plane, lateral angle relative to midline, horizontal distance from external occipital protuberance to lateral rectus capitis posterior major.	Mean depth from skin to C2 spinous process: 37 mm (±7.3 mm). Mean posterior angle: 65.2° (±10°). Mean lateral angle: 34.7° (±12.9°). Mean horizontal distance from external occipital protuberance: 30.6 mm (±9.3 mm).
Arias-Buria et al. 2023 [30]	Cadaveric study	Patellar tendon	Compare the accuracy and safety of ultrasound-guided vs. palpation-guided needle placement in the patellar tendon	Five physical therapists performed 100 needling insertions (50 palpation-guided, 50 ultrasound-guided) in cryopreserved cadaveric knees to reach the interface between the patellar tendon and Hoffa’s fat pad	Solid stainless steel filiform needles 0.30 × 40 mm.	n = 5 Conservation: cryopreserved	Distance to target, surface contact, needle passes, insertion time, unintended punctures	Ultrasound guidance improved accuracy (100% vs. 80%), reduced unintended patellar tendon punctures (16% vs. 52%), and improved contact surface (15.5 mm vs. 4.7 mm). However, it required more time (54.8s vs. 23.75s) and more needle passes (2.55 vs. 1.5) compared to palpation-guided needling

Cranio-Cervical Region: Obliquus capitis inferior muscle [10]; Inferior oblique muscle of the head [28]; Rectus capitis posterior major [29]; Cervical multifidus muscle [22]; Medial pterygoid muscle [11]; Lateral pterygoid muscle [23].

Thoracolumbar and Lumbar Region: Multifidus muscle (thoracolumbar region) [17]; Lumbar multifidus muscle [24]; Upper lumbar spine (L1-L3) multifidus region [21].

Abdominal Region: Rectus abdominis muscle [27].

Upper Limb Muscles: Distal biceps tendon (electrolysis application) [14]; Pronator teres muscle [18]; Supinator muscle – Arcade of Frohse [15]; Pronator quadratus muscle [19].

Lower Limb Muscles: Piriformis muscle [20,25]; Popliteus muscle [13]; Tibialis posterior muscle [12].

Tendinous Structures: Achilles tendon [16] and patellar tendon [30].

3.2. Cadaver sample size and preservation techniques

In this systematic review, we distinguished between cadaveric anatomical specimens (individual anatomical pieces such as isolated limbs, trunk segments, or specific regions) and cadaveric specimens (whole-body cadavers).

3.2.1. Cadaveric anatomical specimens

The number of cadaveric anatomical specimens varied in the different studies, from a single cadaver to a maximum of twenty (Table 1). While two studies conducted their procedures on a single specimen [17,27], others included larger sample sizes, such as Fernández-de-las-Peñas et al. [22] and Mesa-Jiménez et al. [23], with two cadavers. The studies by Fernández-de-las-Peñas et al. [10], Mesa-Jiménez et al. 2020 [11], Calderón-Díez et al. [14] and Arias-Buria et al. [30] used five anatomical pieces, while Belón-Pérez et al. [15] used six anatomical pieces and Hannah et al. [24] used eight anatomical pieces. Larger sample sizes were observed in Calderón-Díez et al. [16], Pérez-Bellmunt et al. [19], Mansfield et al. [21] and Fernández de las Peñas et al. [26] where they used ten anatomical pieces; in Pérez-Bellmunt et al. [12] and Rodríguez-Sanz et al. [13] they used 11 anatomical pieces. Among the studies with the largest sample sizes are Hannah et al. 2023 [25] and Kearns et al. [28], with fourteen, and Kearns et al. [29] fifteen. The studies with the largest sample size were Kearns et al. [20] with nineteen and Fernández-de-las-Peñas et al. [18] with twenty anatomical specimens.

3.2.2. Cadaveric specimens

The number of cadaveric specimens varied across studies, ranging from a single cadaver to a maximum of 25 (Table 1). While two studies conducted their procedures on a single specimen [17,27], others included larger sample sizes, such as Mesa-Jiménez et al. [23] with two cadavers and Belón-Pérez et al. [15] with three. Hannah et al. [24] utilized four embalmed cadavers, whereas studies by Fernández de las Peñas et al. [10], Mesa-Jiménez et al. [11], and Calderón-Díez et al. [14] each involved five. Larger sample sizes were observed in Kearns et al. [28] with seven cadavers, while groups of ten were included in Kearns et al. [29], Mansfield et al. [21], Pérez-Bellmunt et al. [19], and Calderón-Díez et al. [16]. Slightly larger studies included Pérez-Bellmunt et al. [12] and Rodríguez-Sanz et al. [13], both of which analyzed 11 cryopreserved cadavers. The largest study was conducted by Kearns et al. [29], utilizing 25 embalmed cadavers.

To ensure anatomical integrity and methodological rigor, various corpse preservation techniques were employed. In eleven studies, fresh cryopreserved cadavers were used [11–13,16,18–20,22,26,28,30]. In other studies, different techniques for preserving cadavers were used. Williams et al. [17] and McConnel et al. [27], used light embalming. In the studies by Calderón-Díez et al. [14] and Belón-Pérez et al. [15] specimens embalmed with the Thiel technique were used. In the studies by Hannah et al. [24] bodies embalmed with a Maryland embalming solution were used, which contains 5.6% formaldehyde, 27.8% phenol, 33.3% methanol, and 33.3% glycerin, mixed with 2 parts of water. However, the studies by Fernández-de-las-Peñas et al. [10] and Mesa-Jiménez et al. [23] used cadavers fixed in a mixture of formalin, alcohol, and lysol. In isolation, the study by Mansfield et al. [21] used cadavers embalmed with formalin, and Kearns et al. [29] used bodies embalmed with a mixture of water, ethanol, formaldehyde, ethylene glycol, phenol, and glutaraldehyde.

3.3. Needle characteristics and insertion methods

3.3.1. Needle type and length

In all the studies, a wide range of needle types and lengths were used, reflecting the anatomical structures they targeted and the required penetration depth. Sterile stainless steel dry needling needles were used in fifteen studies [10–15,18–20,22,23,26,28–30], in the study by McConnel et al. [27], Calderón-Díez et al. [16] and Williams et al. [17] it was specified that Agupunt brand dry needling needles were used. In the study byMansfield et al. [21] the use of acupuncture needles was reported. Finally, Dritz® hypodermic needles appeared in Hannah et al. [24] y Hannah et al. [25].

Needle lengths across studies ranged from 25 mm to 100 mm, depending on the depth and location of the target muscle. Shorter needles (25–40 mm) were used for superficial structures such as the pronator quadratus muscle [19] and supinator muscle [26]. Medium-length needles (50–70 mm) were selected for moderately deep structures like the multifidus [17] and popliteus [13], while the longest needles (75–100 mm) were used for deep muscle insertions, particularly in studies targeting the piriformis muscle [20] and lumbar multifidus [24] (Table 1).

3.3.2. Needle insertion techniques

Different needle insertion techniques were reported depending on the anatomical region and study goals (Table 1). Manual palpation and anatomical landmark-based approaches were the most frequently employed techniques, found in eight studies [13,19–25]. These methods were particularly useful for targeting deep musculature such as the piriformis and multifidus.

Ultrasound-guided insertions were performed in six studies [10,14–17,30], enhancing accuracy and ensuring the avoidance of neurovascular structures in high-risk areas such as the biceps tendon, Achilles tendon, and patellar tendon. Additionally, five studies confirmed needle placement through cadaveric dissection [11,20,22,27,29]. Two studies employed electrolysis guidance with ultrasound validation to optimize percutaneous electrolysis procedures [15,16,30].

3.4. Procedural safety and validation techniques

Several studies explicitly reported successful needle insertions without compromising neurovascular structures, including Fernández De Las Peñas et al. [18] in the pronator teres muscle, Fernández De Las Peñas et al. [26] in the supinator muscle, Pérez-Bellmunt et al. [19] in the pronator quadratus, Kearns et al. [20] in the piriformis, Williams et al. [17] in the multifidus, and Calderón-Díez et al. [16] in the Achilles tendon (Table 1).

However, some studies highlighted potential risks requiring additional precision and imaging support. Williams et al. [17] reported cases where needle penetration into the multifidus muscle extended into the spinal canal, while Hannah et al. [25] found that medial piriformis insertions resulted in sciatic nerve penetration in 11 out of 14 cases.

To enhance procedural accuracy and safety, various validation methods were employed across studies. Cadaveric dissection, used in thirteen studies, allowed for direct anatomical verification of needle placement [10,11,14–16,18,19,21–24,26,29]. Ultrasound confirmation, employed in six studies, provided real-time visualization and minimized risks associated with blind insertions [10,14,15,17,27,30]. Lastly, electrolysis-guided verification was performed in two studies, ensuring precise needle positioning for therapeutic applications [14,16].

3.5. QUACS synthesis

The analysis of the included studies with the QUACS scale revealed a variation in scores, with the highest score being 12 [15]; and the lowest score 4 [29]. The mean score across all studies was 9.19 with a standard deviation of 2.27, reflecting variability in adherence to reporting standards. The mean percentage of the total score across all studies was 70.70%, indicating overall moderate compliance with quality criteria.

Regarding specific criteria, the item with the most zeros (i.e. the least frequently fulfilled criterion) was ‘Findings are observed by more than one researcher,’ suggesting that many studies did not include multiple observers to validate their results. Conversely, the item with the most ones was ‘Objective stated,’ indicating that nearly all studies clearly articulated their research objectives (Table 2). The Cohen’s kappa was 0.76, indicating substantial inter-rater reliability.

Table 2.

Methodological quality assessment of included studies using the QUACS scale.

Study	Muscle/Structure	Objective stated	Basic information about sample is included	Applied methods are described comprehensibly	Study reports condition of the examined specimen	Education of dissecting researchers is stated	Findings are observed by more than one researcher	Results presented thoroughly and precise	Statistical methods appropriate	Details about consistency of findings are given	Photographs of the observations are included	Study is discussed within the context of the current evidence	Clinical implications of the results are discussed	Limitations of the study are addressed	Score/13	%
Fernández-de-las-Peñas et al. 2019 [10]	Obliquus capitis inferior muscle	1	0	1	1	1	0	1	0	1	1	0	1	0	8	61.53
Mesa-Jiménez et al. 2020 [11]	Medial pterygoid muscle	1	1	0	1	1	1	0	0	0	1	1	1	1	9	69.23
Pérez-Bellmunt et al. 2022 [12]	Tibialis posterior muscle	1	1	1	1	1	0	1	1	1	1	1	1	0	11	84.61
Rodríguez-Sanz et al. 2021 [13]	Popliteus muscle	1	1	0	1	1	0	1	1	1	0	1	1	1	10	76.92
Calderón-Díez et al. 2022 [14]	Distal biceps tendon (electrolysis)	1	1	1	1	0	0	1	1	1	1	0	0	0	8	61.53
Belón-Pérez et al. 2022 [15]	Supinator muscle – Arcade of Frohse	1	1	1	1	1	0	1	1	1	1	1	1	1	12	92.30
Calderón-Díez et al. 2022 [16]	Achilles tendon	1	1	1	1	0	0	1	1	1	1	1	1	0	10	76.92
Williams et al. 2023 [17]	Multifidus – Thoracolumbar region	1	0	0	1	1	0	0	0	0	0	1	1	0	5	38.46
Fernández-de-las-Peñas et al. 2021 [18]	Pronator teres muscle	1	1	1	1	1	0	1	1	1	1	1	1	1	12	92.30
Pérez-Bellmunt et al. 2023 [19]	Pronator quadratus muscle	1	1	1	1	1	0	1	1	1	1	1	1	1	12	92.30
Kearns et al. 2018 [20]	Piriformis muscle	1	1	0	0	0	0	1	0	1	1	1	1	0	7	53.84
Mansfield et al. 2020 [21]	Upper lumbar spine (L1-L3)	1	1	0	1	0	1	1	1	0	1	1	1	1	10	76.92
Fernández-de-las-Peñas et al. 2017 [22]	Cervical multifidus muscle	1	0	1	1	1	0	1	0	0	1	1	1	0	8	61.53
Mesa-Jiménez et al. 2015 [23]	Lateral pterygoid muscle	1	0	1	1	1	0	1	0	0	1	1	1	1	9	69.23
Hannah et al. 2016 [24]	Lumbar multifidus muscle	1	0	0	1	0	0	0	0	1	1	1	1	1	7	53.84
Hannah et al. 2023 [25]	Piriformis muscle	1	1	1	1	1	0	1	1	1	0	1	1	1	11	84.61
Fernández de las Peñas et al. 2021 [26]	Supinator muscle	1	1	0	1	0	0	1	1	1	0	1	1	1	9	69.23
McConnel et al. 2024 [27]	Rectus abdominis muscle	1	1	1	1	1	1	1	0	1	0	1	1	1	11	84.61
Kearns et al. 2021 [28]	Inferior oblique muscle of the head	1	1	1	1	1	1	1	1	1	0	1	1	1	12	92.30
Kearns et al. 2025 [29]	Rectus capitis posterior major	0	1	1	1	1	0	0	0	0	0	0	0	0	4	30.76
Arias-Buria et al. 2023 [30]	Patellar tendon	1	0	1	0	1	1	1	1	0	0	0	1	1	8	61.53

3.6. Risk of bias assessment based on JBI criteria

The included studies demonstrated varying levels of methodological rigor, with notable differences in the clarity of inclusion criteria, the description of sample characteristics and study settings, the control of confounding factors, and the validity and reliability of measurement procedures (Table 3).

Table 3.

Risk of bias assessment of included studies based on Joanna Briggs Institute (JBI) criteria.

STUDY	MUSCLES/STRUCTURE	Were the criteria for inclusion in the sample clearly defined?	Were the study subjects and the setting described in detail?	Was the exposure measured in a valid and reliable way?	Were objective. standard criteria used for measurement of the condition?	Were confounding factors identified?	Were strategies to deal with confounding factors stated?	Were the outcomes measured in a valid and reliable way?	Was appropriate statistical analysis used?
Fernández-de-las-Peñas et al. 2019 [10]	Obliquus capitis inferior muscle	YES	UNCLEAR	YES	YES	YES	NO	YES	YES
Mesa-Jiménez et al. 2020 [11]	Medial pterygoid muscle	YES	UNCLEAR	UNCLEAR	NO	NO	NO	UNCLEAR	YES
Pérez-Bellmunt et al. 2022 [12]	Tibialis posterior muscle	YES	YES	YES	YES	NO	NO	YES	YES
Rodríguez-Sanz et al. 2021 [13]	Popliteus muscle	NO	NO	YES	NO	NO	NO	YES	YES
Calderón-Díez et al. 2022 [14]	Distal biceps tendon (electrolysis)	YES	YES	NO	NO	NO	NO	YES	YES
Belón-Pérez et al. 2022 [15]	Supinator muscle – Arcade of Frohse	UNCLEAR	UNCLEAR	NO	NO	NO	NO	UNCLEAR	NO
Calderón-Díez et al. 2022 [16]	Achilles tendon	YES	YES	NO	UNCLEAR	NO	NO	YES	YES
Williams et al. 2023 [17]	Multifidus – Thoracolumbar region	NO	NO	NO	NO	NO	NO	YES	NO
Fernández-de-las-Peñas et al. 2021 [18]	Pronator teres muscle	NO	YES	NO	NO	NO	NO	NO	UNCLEAR
Pérez-Bellmunt et al. 2023 [19]	Pronator quadratus muscle	YES	UNCLEAR	YES	YES	NO	NO	YES	YES
Kearns et al. 2018 [20]	Piriformis muscle	NO	UNCLEAR	YES	YES	NO	NO	YES	NO
Mansfield et al. 2020 [21]	Upper lumbar spine (L1-L3)	YES	YES	YES	NO	NO	NO	YES	YES
Fernández-de-las-Peñas et al. 2017 [22]	Cervical multifidus muscle	YES	YES	YES	YES	NO	NO	YES	NO
Mesa-Jiménez et al. 2015 [23]	Lateral pterygoid muscle	NO	NO	NO	NO	NO	NO	NO	NO
Hannah et al. 2016 [24]	Lumbar multifidus muscle	NO	NO	NO	NO	NO	NO	NO	NO
Hannah et al. 2023 [25]	Piriformis muscle	NO	YES	YES	NO	NO	NO	YES	YES
Fernández de las Peñas et al. 2021 [26]	Supinator muscle	NO	YES	YES	NO	NO	NO	NO	NO
McConnel et al. 2024 [27]	Rectus abdominis muscle	NO	YES	YES	NO	YES	NO	YES	NO
Kearns et al. 2022 [28]	Inferior oblique muscle of the head	NO	NO	YES	YES	YES	UNCLEAR	YES	YES
Kearns et al. 2025 [29]	Rectus capitis posterior major	YES	YES	YES	YES	YES	NO	YES	YES
Arias-Buria et al. 2023 [30]	Patellar tendon	UNCLEAR	YES	UNCLEAR	NO	NO	NO	UNCLEAR	YES

Studies classified as having a high risk of bias exhibited the greatest methodological weaknesses, particularly in terms of sample description, management of confounding factors, and the validity of outcome measurements. These studies often lacked sufficient detail regarding the study setting and participants, leading to a lack of transparency in their methodology. Furthermore, they did not clearly define or control for confounding variables, which may have impacted the reliability of their findings. The absence of standardized and objective measurement criteria further compromised their methodological robustness [11,15,17,18,23,24].

Studies with a moderate risk of bias demonstrated a mixed adherence to methodological quality standards. While they met some of the JBI criteria, notable shortcomings remained, particularly in the identification and management of confounding factors. Many of these studies lacked sufficient details on how exposures and outcomes were measured, limiting their reproducibility and increasing the potential for bias. While their conclusions are likely to be more reliable than those in the high-risk category, some degree of methodological concern remains [15,18,23,24,30].

Studies with a low risk of bias exhibited strong methodological rigor and adhered to the majority of the JBI quality assessment criteria. These studies provided clear inclusion criteria, detailed descriptions of study subjects and settings, and reliable outcome measurements. Additionally, they demonstrated appropriate use of statistical analysis, contributing to the robustness of their findings [12,14,16,21,25,29,30]. The Cohen’s kappa was 0.90, indicating near perfect inter-rater agreement.

4. Discussion

This systematic review evaluated 21 cadaveric studies investigating the anatomical accuracy and safety of deep dry needling performed without ultrasound guidance across various anatomical regions, including the craniocervical, thoracolumbar, abdominal, upper and lower extremities, as well as tendon structures. The review considered cadaveric sample sizes and preservation methods; needle insertion techniques and characteristics; the validity of the procedures in terms of safety, particularly the avoidance of puncturing anatomical ‘red flags’ such as neurovascular bundles; and the methodological quality and risk of bias of the included studies.

The observed wide variability in needle types and lengths across studies, aligned with the targeted anatomical structures and required penetration depths highlights a key challenge. The lack of standardized recommendations regarding needle selection for specific regions and clinical scenarios. This heterogeneity mirrors findings in clinical practice, where substantial variation in needle choice and technique has been reported across practitioners and settings [31]. Establishing consensus on optimal needle parameters for different anatomical targets is essential to enhance procedural reproducibility, safety, and training standards.

Safety outcomes from these studies generally confirmed the avoidance of neurovascular complications in several muscles, including pronator teres [18], pronator quadratus [19], and Achilles tendon [16]. Nonetheless, critical safety concerns were also reported, such as inadvertent spinal canal penetration observed by Williams et al. [17] and frequent sciatic nerve penetration in piriformis insertions highlighted by Hannah et al. [25]. In this regard, the implementation of standardized protocols for the description of invasive approaches becomes essential. Explicitly defining anatomical landmarks used to determine needle entry points, specifying distances to target structures, insertion angles, and needle depth relative to its size may contribute to improving both the precision and reproducibility of procedures [31]. Furthermore, reporting additional cadaver-specific variables that may influence needle selection and insertion angle, such as the circumference of the target region or the thickness of subcutaneous tissue, has been recommended to enhance accuracy and safety in both cadaveric research and clinical applications [31,32]. It should also be noted that ultrasound guidance, while invaluable for increasing procedural accuracy in cadaveric validation studies, is equally relevant for improving safety and precision in clinical practice with living patients, particularly when targeting deep or anatomically complex structures [33].

The sample sizes varied considerably, from single-cadaver studies [17] to those involving up to twenty anatomical specimens, as observed in Fernández-De-Las-Peñas et al. [18]. Preservation techniques employed in the studies predominantly featured fresh cryopreserved cadavers [11–13,16,18–20,22,26,28,30], with additional use of various embalming methods, including Thiel [14,15] and Maryland [24,25]. However, the heterogeneity in preservation protocols, ranging from differences in embalming solutions to the timing and frequency of freezing and thawing cycles, may significantly alter tissue integrity. This variability can impact needle-tissue interaction, potentially introduce bias when comparing accuracy and safety outcomes across studies.

In relation to the methodology employed and considering the QUACS scale, certain studies lacked a clear and reproducible methodology, such as those by Mesa-Jiménez et al. [11], Rodríguez-Sanz et al. [13], Williams et al. [17], Mansfield et al. [21], Hannah et al. [24], and Fernández De Las Peñas et al. [26]. Additionally, although numerous articles included photographs of dissections, only Belón-Pérez et al. [15] provided a detailed description of both the procedure and the professionals performing it. Moreover, according to the QUACS checklist, only 5 out of the 21 studies involved more than one observer in the assessment of outcomes, increasing the risk of bias due to the lack of independent verification of adherence to the defined invasive approach protocols.

4.1. Clinical applications

Results from this systematic review support the critical importance of precise anatomical landmark identification and procedural standardization to ensure patient safety during dry needling. The development of standardized clinical protocols based on cadaveric findings will enhance the reliability of needle insertion techniques. Moreover, educational programs for physical therapists and related healthcare professionals should incorporate cadaver-based anatomical training and simulation-based practice into their curriculum. Hands-on training with cadaveric models allows practitioners to develop detailed anatomical knowledge and to safely practice needle insertion techniques before performing them on patients. This approach can reduce the risk of adverse events and enhance clinical confidence. Simulation-based training models for dry needling and invasive procedures, which integrate cadaver-based practice and ultrasound-guided feedback, could be particularly valuable for improving procedural accuracy and patient safety.

Clinicians should consider adopting conservative needle insertion approaches, especially in high-risk and deep anatomical areas. Integration of preliminary screening tools, including patient-specific anatomical variations (e.g. scoliosis, obesity), could further improve procedural safety and effectiveness. Findings from this systematic review can be used to inform future guidelines and consensus documents regarding best practices for deep dry needling, particularly in settings where ultrasound guidance is not feasible or available.

4.2. Limitations

This systematic review presents several limitations that should be considered when interpreting the results. The heterogeneity in cadaver preservation methods (cryopreservation, light embalming, Thiel embalming, and formalin fixation), combined with the lack of standardization in key variables (e.g. time elapsed from death to preservation, number of freeze-thawing cycles), may impact tissue integrity and alter needle-tissue interactions, thus affecting the accuracy and safety of the procedures assessed. In addition, some studies had limited sample sizes, occasionally involving as few as a single cadaver, which restricts external validity and the generalizability of findings. Such limited anatomical representativeness fails to adequately capture individual variability typically encountered in clinical practice, including anatomical differences or gender-related variations. Another significant limitation identified through the QUACS scale is the widespread lack of multiple observers in result validation. The reliance on a single investigator introduces assessment bias due to lack of blinded or multiple observers, thereby diminishing the reproducibility and reliability of the conclusions. Methodological shortcomings were noted in the procedural descriptions across multiple studies, complicating exact replication. More comprehensive documentation regarding procedural details, technical needle specifications, and investigator training background would be crucial for enhancing the clinical applicability of these techniques. Future research should prioritize the development of reproducible, standardized protocols with well-documented procedural details, and should include needle insertions performed by different practitioners to assess intra- and inter-examiner reliability. This would contribute to improving the overall methodological robustness and clinical applicability of these techniques.

5. Conclusions

This systematic review demonstrates that non-ultrasound-guided deep dry needling can be performed accurately and safely across various anatomical regions when standardized protocols are followed. However, in certain high-risk or anatomically complex regions, the use of ultrasound guidance may provide important benefits in enhancing procedural accuracy and patient safety. Inconsistencies in cadaver preservation methods, limited use of multiple observers, and insufficient methodological reporting in some studies constrain the generalizability of findings. Future research should prioritize the use of fresh-frozen cadavers, appropriate sample sizes, and rigorous anatomical validation, as well as the systematization of invasive procedures aimed at assessing intra- and inter-examiner reliability, to enhance the overall reliability and clinical safety of these techniques.

Biographies

Marta Sánchez Montoya, PT, MSc, is a physiotherapist and adjunct professor in the Department of Medicine at the European University of Madrid, Madrid, Spain. Her areas of expertise include human anatomy and anatomical dissection.

Jaime Almazán Polo, PhD, PT, is a physiotherapist and faculty member in the Department of Physiotherapy at the Universidad Complutense de Madrid (Faculty of Nursing, Physiotherapy and Podiatry), Madrid, Spain. His areas of expertise include musculoskeletal ultrasound, invasive physiotherapy approaches, and anatomical research.

Ángel González de la Flor, PhD, PT, is a physiotherapist and faculty member in the Department of Physiotherapy at the School of Medicine, Health and Sports, European University of Madrid, Madrid, Spain. His areas of expertise include physical and functional assessment, human anatomy, and therapeutic exercise.

Funding Statement

The author(s) reported there is no funding associated with the work featured in this article.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/10669817.2025.2536818

Highlights

• This systematic review synthesizes anatomical safety data from 21 cadaveric studies on non-ultrasound-guided dry needling techniques, revealing significant heterogeneity in methodological quality and reporting standards.

• The review underscores the need for standardized, reproducible procedural frameworks in invasive physiotherapy, recommending the use of fresh-frozen cadavers, multiple observers, and comprehensive protocol reporting to improve accuracy, safety, and clinical applicability.

• Future research should focus on intra- and inter-examiner reliability through the implementation of structured needling protocols performed by different practitioners, supporting evidence-based integration of these techniques into clinical practice.