Accuracy of Carpal Tunnel Injection: A Prospective Evaluation of 756 Patients

Abstract

Background: Corticosteroid injection into the carpal tunnel is both a diagnostic test and a therapeutic modality in the treatment of carpal tunnel syndrome. Many injection techniques are described in the literature. Improper placement of injection may result in damage to neurovascular structures in the carpal canal or decrease efficacy of the test and/or therapy. The purpose of this study is to determine if carpal tunnel injection using anatomic landmarks is reproducible and safe. A review of the senior author’s injection technique is presented. Methods: Over 8 years, there were 756 attempted placements of a 25-gauge needle into the carpal tunnel in a simulated carpal tunnel injection prior to open carpal tunnel release. The needle was inserted at the wrist crease, just ulnar to palmaris longus. Open carpal tunnel release was subsequently performed, and position of the needle was recorded. Results: In 572 patients (75.7%), the needle was found to be in the carpal tunnel without penetration of contents. The needle was in the carpal tunnel but piercing the median nerve in 66 attempts (8.7%). The carpal tunnel was missed in 118 attempts (15.6%). Conclusions: This is the largest study looking at accuracy of carpal tunnel injection using anatomic landmarks. Our injection accuracy (75.7%) is less than reported in previous studies, which note 82% to 100% accuracy using the same injection technique. This may indicate that carpal tunnel injection is less reliable than previously thought. Safety of carpal tunnel injection remains an important concern. The median nerve was penetrated in 8.7% of attempts.

Introduction

Carpal tunnel syndrome is the most common compressive neuropathy affecting the upper extremity.^{2,3,5,11,17,27,29} Unless motor function or severe sensory loss is present, the initial treatment for the disease usually consists of carpal tunnel injection with steroid.^{5,19,25,29,31,35,37} The senior author previously reported both the therapeutic and diagnostic value of carpal tunnel injection.⁹ As a therapeutic modality, 81% of patients obtained good or complete relief, lasting an average of 3.3 months after injection of the carpal tunnel before recurrence of symptoms. Only 46% of recurrences were severe enough to warrant surgical intervention. As a diagnostic test, good relief from a previous carpal tunnel injection correlated with a good result from surgical intervention. Ninety-six percent of patients obtained good results from carpal tunnel release after previously receiving good results from a carpal tunnel injection.

Various techniques have been used for carpal tunnel injections.^{1,4,6-11,13-16,18-20,23-25,27-32,35-37} Damage to neurovascular structures during a carpal tunnel injection can occur and has been reported in the literature.^{6,8,12,16,18,21-23,26,29,33} Also, steroid preparations can have a neurotoxic effect.²² Given the diagnostic and therapeutic value of carpal tunnel injections, and the harm that can occur from misplacement of the needle, correct technique is critical in the performance of the injection.

The senior author has performed a teaching study over 8 years to determine the accuracy of carpal tunnel injections using his injection technique, which is reviewed in the discussion section. The idea was first presented by Michael R. Wood in 1980.³⁷

Materials and Methods

Over an 8-year period from 1999 to 2006, 31 hand fellows (trained in orthopaedic or plastic surgery) performed a total of 756 simulated injections. Surgeries were conducted at the affiliated surgical centers in hospitals for an academic private practice of the senior partner.

Patients recruited for the study were those scheduled for carpal tunnel release in the senior partner’s practice. Ages ranged from 16 to 89, with approximately one-third men and two-thirds women. Patients were consented to needle placement and study inclusion as part of informed consent for carpal tunnel release.

Prior to open carpal tunnel release, a simulated carpal tunnel injection was performed. A 25-gauge, 1.5-inch needle was placed in the carpal tunnel without injection of lidocaine or steroid. The needle was left in situ. After opening of the carpal tunnel, the location of the needle was determined and recorded (Figure 1).

Figure 1.

A simulated carpal tunnel injection was performed and the needle was left in situ. After opening of the carpal tunnel, the location of the needle was determined and recorded.

Results

Needle placement was accurate in 572 (75.7%). Inaccurate placement included 66 (8.7%) penetrations of the median nerve, 44 (5.8%) in Guyon’s canal, 55 (7.3%) in the ulnar wall of the carpal tunnel, 14 (1.9%) in or superficial to the transverse carpal ligament, and 5 (0.7%) in the radial wall (Figure 2). The accuracy of injections per fellow ranged from 53% to 100%. Only 3 fellows correctly placed the needle 100% of the time.

Figure 2.

Penetration of the median nerve was the most common location of inaccurate needle placement. Other locations included Guyon’s canal, the ulnar and radial walls of the carpal tunnel, and within or superficial to the transverse carpal ligament. FCR = flexor carpi radialis; FCU = flexor carpi ulnaris; PL = palmaris longus; H = hamate; P = pisiform.

Discussion

Accurate placement of the needle during carpal tunnel injection is more difficult than it appears, as learned by each of the hand fellows participating in this study. Other studies have noted accuracy rates between 72% and 100% with various injection techniques utilizing anatomic landmarks or ultrasound guidance.^{9,10,29,34,27} Gutierrez et al found accurate placement in 23 of 28 cadavers (82%) when injecting through a distal palmar site (Rankin’s zone) in a cadaver model.¹⁰ Ozturk et al also used a cadaver model, comparing the accuracy of needle placement between 3 different techniques: (1) Radial entry 1 cm proximal to the wrist crease and directed in an ulnar direction by roughly 45° through the FCR (96% accuracy); (2) our senior author’s technique (82% accuracy); and (3) ulnar entry just distal to wrist crease in line with the fourth ray, then directed in a radial direction by roughly 45° (72% accuracy).²⁹

This wide range includes significant differences in accuracy using the same technique, suggesting surgeon skill plays a large role. Practicing needle placement (without injection) at the time of open carpal tunnel release has proven to be an excellent way to enhance one’s skill.^9,37 Having performed several thousand carpal tunnel injections during his career, the senior author admits that he does not always place the needle in the correct position. Knowledge of pertinent anatomy and consistent technique are important to ensuring accurate needle placement during carpal tunnel injection.

Anatomy

There is a narrow window for placement of the needle during injection of the carpal tunnel. The goal is to place the needle tip between the median nerve and the hook of the hamate in the region of the flexor tendons. Placement of the needle too far radial may risk damage to the median nerve. Placement too far ulnar will position the needle in the ulnar wall or Guyon’s canal. Superficial placement puts the needle in or above the transverse carpal ligament.

Injection Technique

After aseptic preparation of the skin, the needle is inserted into the carpal canal from a point in line with the axis of the fourth ray at the wrist flexion crease (Figure 3). This point corresponds to a position just ulnar to the palmaris longus tendon in most patients. With the wrist in slight extension, the needle is angled approximately 30° from the skin. The hook of the hamate is palpated, and the needle is directed just radial to this bony landmark.

Figure 3.

The needle is inserted at the wrist flexion crease at a point in line with the axis of the fourth ray.

A 2- to 3-mL bolus of lidocaine can then be injected beneath the transverse carpal ligament (Figure 4). If the patient experiences paresthesias in the median nerve distribution, the needle should be withdrawn slightly and redirected more ulnarly. If there is difficulty with the injection, the needle may be positioned in the ulnar wall or the transverse carpal ligament. Slight repositioning of the needle should allow a free flowing injection. The senior author prefers to inject lidocaine separately from the steroid preparation. This ensures that the needle is in the correct position prior to the injection of the potentially harmful steroid.

Figure 4.

With the wrist in slight extension, the needle is angled approximately 30° from the skin and inserted beneath the transverse carpal ligament. (A) As lidocaine is injected, a palpable bulge at the distal edge of the transverse carpal ligament (B) ensures correct placement of the injection.

Complications

Incorrect placement of the needle is the most common complication that can occur during carpal tunnel injections.^{6,18,26,29,31} The physician can utilize 3 methods to determine if the injection is located in the correct position. First, the bolus of lidocaine should be palpable just distal to the transverse carpal ligament during the injection.⁹ Second, a change in position of the needle proximally should be noticed with gentle active flexion of the patient’s fingers.¹⁸ Third, the patient should develop numbness in the distribution of the median nerve.¹⁹ This is probably the most important sign that the injection has been placed within the carpal tunnel.

Several cases of the intraneural injection of steroid have been documented in the literature.^8,21,26,33 Mackinnon et al previously documented the neurotoxic effect of different steroid preparations.²² The damage occurs only when the steroid is injected intrafascicularly. There is a wide range of damage that can be done, depending on the type of agent used. Dexamethasone caused the least damage, while hydrocortisone and triamcinolone hexacetonide caused the most damage. Injection of lidocaine intrafascicularly resulted in no damage to the nerve.

Conclusion

Carpal tunnel injection is both diagnostic and therapeutic in the treatment of carpal tunnel syndrome; therefore, proper placement of the needle is critical. Damage can result from inadvertent injection of steroid into the median nerve, and injection into the walls or outside of the carpal canal will obviously provide neither relief of symptoms nor diagnostic information. This study suggests that accuracy of injection may be less than previously believed. Future studies should investigate whether practicing simulated carpal tunnel injections prior to opening the transverse carpal ligament helps physicians improve the accuracy of their injections.