Trigger Finger Corticosteroid Injection With and Without Local Anesthetic: A Randomized, Double-Blind Controlled Trial

Abstract

Background: The first-line treatment for trigger finger is a corticosteroid injection. Although the injectable solution is often prepared with a local anesthetic, we hypothesize that patients receiving an injection with anesthetic will experience more pain at the time of injection. Methods: C Patients with trigger finger were prospectively randomized into 2 cohorts to receive triamcinolone (1 mL, 40 mg) plus 1% lidocaine with epinephrine (1 mL) or triamcinolone (1 mL, 40 mg) plus normal saline (1 mL, placebo). Both patient and surgeon were blinded to the treatment arm. The primary outcome was pain measured using a (VAS) immediately following the injection. Results: Seventy-three patients with a total of 110 trigger fingers were enrolled (57 lidocaine with epinephrine and 53 placebo). Immediate postinjection pain scores were significantly higher for injections containing lidocaine with epinephrine compared with placebo (VAS 3.5 vs 2.0). Conclusions: In the treatment of trigger finger, corticosteroid injections are effective and have relatively little associated pain. This study shows there is more injection-associated pain when lidocaine with epinephrine is included with the corticosteroid. Therefore, surgeons looking to decrease injection pain should exclude the anesthetic, but they should discuss the trade-off of foregoing short-term anesthesia with patients. Using only a single drug (ie, corticosteroid alone) is not only less painful but is also more simple, efficient, and safe; this has therefore become our preferred treatment method.

Introduction

Stenosing tenosynovitis (trigger finger) is present in 2% to 5% of the adult population and in up to 10% of patients with diabetes mellitus.^1-4 This common hand pathology results from impingement of digital flexor tendons as they glide through a hypertrophied tendon sheath at the first annular (A1) pulley.^3-5 Treatment of trigger finger includes a corticosteroid injection or surgical division of the A1 pulley—the former being pursued initially because it is less invasive and more cost-effective.⁶

Corticosteroids decrease inflammation and are an effective first-line treatment for trigger finger, resulting in resolution of symptoms in 50% to 90% of patients.^6-11 Given the regularity of this clinical procedure for hand surgeons, various injection preparations have been described, with many surgeons administering corticosteroids mixed with a local anesthetic.^5,11-13 Although the anesthetic provides pain relief after injection, it does not diminish needle pain and is known to cause a burning sensation during infiltration. Therefore, although local anesthetic is given to increase patient comfort, it may paradoxically make the experience of an injection worse.

Given this possibility, we sought to investigate whether an anesthetic additive increases injection pain for trigger finger. We performed a randomized, double-blind, placebo-controlled trial to test our hypothesis that there is no difference in pain at the time of injection in patients receiving injections of corticosteroids with and without a local anesthetic for the treatment of trigger finger.

Materials and Methods

Study Cohort

After obtaining prestudy approval from the institutional review board, patients were prospectively enrolled if they were willing to provide informed consent and had symptoms of 1 or more trigger digits. Patients were excluded if they elected for open trigger finger release or had contraindications to corticosteroids or lidocaine with epinephrine. Trigger finger was diagnosed by patient-reported history and physical examination (as determined by a fellowship-trained hand surgeon) demonstrating locking or catching of the digit and pain to palpation at the A1 pulley. Demographic data, history of trigger finger, history of corticosteroid injections, and comorbid conditions (diabetes mellitus, carpal tunnel syndrome, smoking, renal disease, thyroid disease, chronic opioid use) were recorded.

Intervention

Using a coin flip, patients had a 50% chance of being randomly assigned to 1 of the 2 treatment arms to receive a 2-mL injection of either: (1) triamcinolone (1 mL of 40 mg/mL) plus 1% lidocaine with 1:100 000 epinephrine (1 mL); or (2) triamcinolone (1 mL of 40 mg/mL) and normal saline (1 mL). Normal saline was a placebo added to control for volume. Although many variations exist in volume and in steroid and anesthetic selection, these mixtures were chosen based on our standard clinical practice and evidence supporting its common use by hand surgeons.^11-13 A study coordinator prepared the injection in a 5-mL syringe with a 27-gauge needle, therefore blinding both patient and surgeon to randomization. A single fellowship-trained hand surgeon performed all injections in a standard technique. Using the intercrease line as a landmark, the 2-mL solution was injected at a steady rate outside the tendinous sheath at the level of the A1 pulley (intrathecal or intratendinous injections were avoided).¹⁴ A small dressing was applied, and no postprocedural analgesic medications were prescribed or advised.

Outcomes

The primary outcome was pain associated with the injection. Therefore, immediately after treatment, the surgeon asked the patient to rate the pain experienced during injection using a Visual Analog Scale (VAS) with an image of the Wong-Baker Pain FACES Pain Rating Scale.^15,16 This technique has been validated for accurately assessing patient pain after minor procedures including in hand surgery.^15-17 Infection, hypopigmentation, and tendon rupture were the defined adverse outcomes and were monitored throughout the study period. Cost of care was calculated based on institutional cost of purchasing the equipment required for treatment (eg, syringes, needles, and medication).

Analysis

To obtain 80% power with an α of .05, it was necessary to perform 73 injections to reject the null hypothesis. Based on prior studies of trigger finger injection, a standard deviation of 0.3 was used for the sample size calculation, and a VAS score difference of 1.2 was set as a minimally clinically relevant difference.^18,19 Statistical significance was set as P < .05. The VAS scores were statistically compared using a Wilcoxon rank sum test. Finally, costs of purchasing supplies were collected at the institutional level. Consolidated Standards of Reporting Trials (CONSORT) guidelines were reviewed and followed.

Results

Seventy-five consecutive patients were screened for enrollment into the study, with 2 declining injections in favor of open treatment. All remaining patients met inclusion criteria, leading to 73 patients enrolled with a total of 110 injections. There were 57 injections using triamcinolone plus lidocaine with epinephrine (intervention) and 53 using triamcinolone with normal saline (placebo). The 2 treatment groups had similar demographics and comorbidities (Table 1). Most patients were men (95%), and the average age was 64 years. The treated digit was most frequently on the right hand (61%), and the middle finger was most commonly injected, followed by ring, index, and thumb. Overall, the most common comorbidity was diabetes mellitus (29%). There were no adverse outcomes in either cohort.

Table 1.

Cohort Demographics.

	Steroid and normal saline	Steroid and lidocaine/epinephrine	Total
Total injections	53	57	110
Age
Mean	66.0	62.7	—
Sex
Male	50	55	105
Female	3	2	5
Affected hand
Right	32	35	67
Left	21	22	43
Affected digit
Thumb	5	8	13
Index	7	7	14
Middle	22	22	44
Ring	16	20	36
Small	3	0	3
Recurrence
Prior corticosteroid injection	18	10	28
Comorbidities
Diabetes mellitus	16	16	32
Renal disease	5	4	9
Thyroid disease	4	1	5
Current smoker	4	5	9
History of smoking	8	13	21
Chronic opioid use	6	2	8
Carpal tunnel syndrome	15	11	26

Note. There were no statistically significant differences in this table.

Table 2 compares the patient-reported VAS between cohorts. The immediate postinjection VAS scores were significantly lower in the steroid-only treatment arm (2.0 vs 3.5, P < .01). Based on our minimally clinically relevant difference of 1.2, there was a clinically and statistically significant increase in pain experienced by patients receiving an injection that included lidocaine with epinephrine. There were no adverse outcomes at 6 weeks. Not surprisingly, the addition of lidocaine with epinephrine increased the cost of trigger finger injections by 185% compared with corticosteroid alone.

Table 2.

Postinjection VAS Pain Scores.

	Steroid and normal saline	Steroid and lidocaine/epinephrine	P value
Total injections	53	57	—
Postinjection VAS score
Immediate	2.0	3.5	.00024
Adverse outcomes
Infection	0	0	—
Hypopigmentation	0	0	—
Tendon rupture	0	0	—

Note. VAS = Visual Analog Scale.

Discussion

Corticosteroid injection is an effective treatment for trigger finger, which is one of the most frequent clinical procedures for many hand surgeons. Commonly, a local anesthetic is added to the corticosteroid solution, presumably to improve patient comfort. However, whether patients actually benefit is not clear. In this randomized, double-blind, placebo-controlled trial, we found that patients experience more pain during treatment when the injectable solution contains lidocaine with epinephrine. Although the patients have short-term pain relief with anesthetic (depending on the choice of anesthetic), they are less comfortable during the injection of it. Therefore, surgeons should discuss with patients whether they would prefer a less painful injection (without anesthetic) or a period of temporary anesthesia.

This study has several important limitations, primarily due to the variety of methods used to treat trigger finger by corticosteroid injection. In preparation of the injectable solution, literature indicates that many hand surgeons use a local anesthetic (with or without epinephrine) prior to or simultaneously (ie, mixed solution) during treatment. Lidocaine with epinephrine is frequently used, as it has a rapid onset of action (less than 2 minutes) and longer duration than plain lidocaine. However, patients will experience an intense burning during its injection due to acidic pH.^20,21 Our results support this theory, as the cohort receiving a mixture of lidocaine with epinephrine reported immediate pain from the injection. Although some surgeons advocate using sodium bicarbonate to neutralize this sensation, this increases both cost and time of preparation, and there is significant debate over its efficacy.⁵ More importantly however, local anesthetic (even with bicarbonate) cannot affect the noxious component associated with needle placement, which in our experience is a particularly painful aspect of trigger finger injection as the glabrous palmar skin has a high density of nociceptors.

Removing anesthetics from the trigger injections may lessen pain by decreasing the volume of injectate. Injecting larger volumes causes the nociceptors to stretch, thereby activating pain fibers.^22,23 Therefore, including even a small volume of anesthetic in an injection will be noxious until the medication induces local anesthesia. An alternative approach to minimize the injectable volume is via a needle-free jet administration of lidocaine. Earp et al²⁴ demonstrated in a randomized controlled trial that jet administration is less painful than injecting lidocaine.

To reduce acute injection pain, it could be more advantageous if the surgeons injected the anesthetic prior to steroid treatment using a separate syringe, but we do not find this either practical or pain-free. Instead, our study demonstrates that lidocaine with epinephrine causes greater pain than corticosteroid alone. Therefore, doing a block before a trigger injection is ostensibly more painful than a plain corticosteroid injection, and we cannot support this approach.

Finally, there are several relevant drawbacks to including lidocaine with epinephrine with trigger injections. First, additional medications and injection supplies increase costs. Our conservative sample calculations show additional injections cost nearly twice as much. Moreover, there have been regular national shortages of lidocaine with epinephrine, which has directly affected the availability and unit cost of this preferred medication.²⁵ Second, adding lidocaine increases the risk of a rare, but not irrelevant, possibility of adverse medication reaction, allergy, or interaction.

There are other notable limitations to this study. The results may not be generalizable because of our patient population at the Veterans Affairs Health System, who are predominantly men. In addition, there was a relatively high incidence of insulin-dependent diabetes mellitus, a population familiar with injections. Interviewer and response biases are inherently a risk when soliciting a pain score response, but we conducted each injection and VAS evaluation with a standardized approach by 1 surgeon, and researcher bias was protected by the blinding process. Regarding the study design, we used lidocaine with epinephrine (pH 4.5) based on our standard practice, but using an anesthetic with a more neutral pH (eg, plain lidocaine, pH 6.5) or buffering the solution using bicarbonate may have decreased the burning sensation associated with injection. Nevertheless, this change would not have removed the noxious stimulus of the needle and therefore only has a potentially short-term anesthetic benefit.

There are many variations in treating trigger finger with corticosteroids. In the decision-making process with patients, surgeons should discuss the short-term benefits of added a local anesthetic and contrast this with increased injection pain. It is now our practice to include patients in this shared decision-making when we administer trigger injections. Given our findings, there may be precedence to remove local anesthetics from corticosteroid mixtures in the injection to other sites (eg, intralesional or intra-articular), and this is the subject of future research.