Effect of Needle Size on Pain Perception in Patients Treated With Botulinum Toxin Type A Injections: A Randomized Clinical Trial

Importance  Transcutaneous injection through smaller hollow-bore needles may decrease patient discomfort, but current evidence is equivocal.

Objective  To compare injection discomfort in patients treated with botulinum toxin type A with 30- and 32-gauge needles.

Design, Setting, and Participants  Split-face, patient- and injector-blinded randomized clinical trial at the dermatology service of an urban university medical center. The 20 participants were women aged 25 to 70 years in good health and with moderate dynamic forehead and glabellar wrinkles. Data were collected from November 20, 2013, through January 16, 2014. Follow-up was complete on January 16, 2014. Data from the per-protocol population were analyzed from July 1 to July 31, 2014.

Interventions  One side of each patient’sforehead received botulinum toxin type A in saline injected with a 32-gauge needle; the other side received the same treatment injected with a 30-gauge needle. In addition, each patient received randomized injections of saline only to both upper inner arms with the same types of needles.

Main Outcomes and Measures  Primary outcomes included the patient-reported pain rating on a visual analog scale (VAS) on either side of the face and arms and the proportion of patients whose VAS ratings corresponded with more than moderate (ie, clinically significant) pain. The secondary outcome consisted of patient-reported information about the character of the pain at both sites using the expanded and revised version of the Short-Form McGill Pain Questionnaire.

Results  All 20 patients completed the study. Overall, facial and arm injections were nominally but not significantly more painful with 30-gauge needles (mean [SD] VAS ratings for the face, 4.16 [2.55] vs 3.41 [2.31], P = .34; for the arm, 1.66 [2.07] vs 1.21 [1.65], P = .45). For facial injections, the likelihood of clinically significant pain (VAS rating, ≥5.4) was significantly greater with 30-gauge needles, which were associated with such pain in 8 patients (40%) compared with the 32-gauge needles, which were associated with such pain in 3 patients (15%) (odds ratio, 3.80 [95% CI, 1.05-13.78]; P = .04). No difference was found in the character of pain associated with needle bore (P > .05 for all comparisons).

Conclusions and Relevance  For facial injections of neurotoxin in saline, 30-gauge needles were associated with greater incidence of clinically significant pain than 32-gauge needles. For patients prone to experience clinically significant pain with facial injections, use of 32-gauge needles may minimize this discomfort.

Trial Registration  clinicaltrials.gov Identifier: NCT01981174

Neurotoxins, prepackaged fillers, and intralesional local anesthesia are delivered through small-caliber needles. Although injections are safe and well tolerated, one source of associated discomfort for the patient is the sensation of the needle tip piercing the skin and ejecting the material into the surrounding tissues. Mitigation of injection pain has been attempted by pretreatment of the skin with local anesthetic, nerve blocks, and physical modalities, such as ice and vibration; by inclusion of adjuvant injectable substances that reduce discomfort, such as lidocaine hydrochloride with fillers, preserved saline with neurotoxin, and bicarbonate with lidocaine and epinephrine; and by pain-minimizing techniques, such as slow, relatively superficial infiltration.

The external needle gauge or diameter has also been considered to contribute to injection pain. A larger-diameter needle can crush or injure a larger surface volume of skin on insertion, thus leading the patient to experience more discomfort. Because facial injections are common in dermatology, the face is well innervated for sensation, and 30-gauge needles are used routinely, a comparison of facial injections with 30-gauge needles with other modalities may reveal alternative paradigms that may be better tolerated. Results of the few clinical trials designed to test this hypothesis have been uncertain. The purpose of this study is to compare the discomfort associated with superficial cutaneous injections of botulinum toxin type A using 30- and 32-gauge needles.

This was an urban university medical center–based split-face, parallel-group, randomized clinical trial with a 1:1 allocation ratio and a block size of 2 (Figure). The unit of randomization was the individual unilateral forehead. The study was approved by the institutional review board of Northwestern University. The full study protocol can be found in the Supplement.

Patients were recruited from an urban dermatology practice based at Northwestern University and the surrounding community. All participants provided written informed consent. Data were collected from November 20, 2013, through January 16, 2014. Follow-up was complete on January 16, 2014. Eligible patients were aged 25 to 70 years, were in good health, were female, and had moderate dynamic forehead and glabellar wrinkles. The study excluded women who were pregnant or lactating, who had any active infection in the treatment area, who were allergic to cow’s milk protein or albumin, who were currently receiving aminoglycosides, or who were receiving anticoagulation or had a diagnosis of bleeding disorders. In addition, we excluded any patient who had undergone the following treatments to the forehead and glabellar region: botulinum toxin injections, ablative laser procedures, radiofrequency device treatments, ultrasound device treatments, or medium to deep chemical peels in the past 6 months; temporary soft-tissue augmentation material in the past year; semipermanent soft-tissue augmentation material in the past 2 years; or permanent soft-tissue augmentation material at any point in the past. Those patients with prior facial cosmetic surgical procedures, including eyebrow-lifts, blepharoplasties, and rhytidectomies, were also excluded.

Patient screening and enrollment was performed by two of us (A.G. and D.S.). Random sequence generation and concealment were performed (M.N.) and conducted using a toss of the same fair coin, with the outcomes (1 or 2) recorded separately on individual paper cards that were then placed in sealed, opaque, consecutively numbered envelopes. Three envelopes were unsealed consecutively to determine each patient’s assignments. The first number assigned the patient to 1 of 2 groups for facial injections. Those in group 1 were designated to receive 32-gauge needle injections on the right side of the face, and those in group 2 were assigned to receive 32-gauge needle injections on the left side. The second number assigned the patient to 1 of 2 groups for arm injections. Those in group 1 were designated to receive 32-gauge needle injections to the left arm, and those in group 2 were assigned to receive 32-gauge needle injections to the right arm. The third number determined injection order, with assignment to group 1 resulting in the 30-gauge needle side being injected first. Study assignments were overseen by one of us (A.G.). Study treatments were delivered by 2 dermatologists (M.A. and K.M.).

Patients were blinded regarding the needle bore used on each side. The dermatologists who administered the injections were similarly unaware of assignments because the needle hub, the only visually apparent point of difference between the 2 needle types, was obscured with opaque tape by one of us (D.S.) before injection.

Each patient received treatment for dynamic creases of the forehead and glabellar complex with onabotulinum toxin, a biologically distinct form of botulinum toxin type A. The toxin for injection was obtained from 100-U vials of toxin reconstituted on the day of injection with 2.5 mL of normal saline containing benzyl alcohol preservative. Before injection, the patient’s glabellar and forehead area was cleansed with alcohol-impregnated pads. Depending on the patient’s forehead size, muscle size, and degree of muscle activity, 3 to 4 aliquots of 2 to 4 U each were placed on either side of the forehead and glabellar complex, with orientation bilaterally symmetrical relative to the vertical facial meridian. One-milliliter syringes (Luer Lock; Medline Industries) attached to 1.3-cm (½-inch) needles were used for injection. A serial puncture injection technique was used with a vertical angle of incidence (90° relative to the skin surface), and needle tip placement was at the interface of the deep dermis and superficial subcutis during the process of fluid delivery. Patients underwent injection sitting on a surgical table that was raised so that the sites of injection were at eye level for the standing dermatologist. The technique for both sides was identical except for the gauge of needle used.

After each side was injected, the patients were asked to rate the pain they just experienced on a visual analog scale (VAS) ranging from 0 to 10, with 0 representing no pain and 10 the greatest imaginable pain. At the same time, patients were asked to characterize the type of pain they experienced, if any, using 1 or more of the 22 descriptors found in the expanded and revised version of the Short-Form McGill Pain Questionnaire (SF-MPQ-2).¹ A final injection of toxin into the central glabella, which was not on the left or the right side of the face, was delivered after the pain surveys to avoid confounding of the results. Ice packs were then applied to reduce any residual discomfort.

Three injections of saline only were placed 2 cm apart along the midline of each upper inner arm. The technique and procedures were otherwise similar to the facial protocol, with the exception that no additional injections were delivered after administration of the VAS and questionnaires.

The primary outcome measures included the patient-reported VAS pain rating on either side of the face and arms and the proportion of patients whose VAS ratings corresponded with more than moderate (ie, clinically significant) pain.² The secondary outcome measure consisted of patient-reported information about the character of the pain at the several sites using the SF-MPQ-2. The VAS and SF-MPQ-2 are validated scales.^1,3

Assuming an SD of 2.5, a sample of 20 patients had 87% power to detect a 2-U mean difference in pain ratings between treatments. In addition, a sample of 20 patients had 88% power to detect a 25-point difference in the probability of experiencing clinically significant pain between treatments. A 2-sided test, a type I error rate of 5%, and a binomial distribution with a 15% probability of success were assumed.

Data from the per-protocol population were analyzed from July 1 to July 31, 2014. Descriptive analysis and a 1-way analysis of variance were modeled to continuous VAS scores. To allow for valid statistical inference (given that the VAS and SF-MPQ-2 scores were heavily skewed with large numbers of zeroes [ie, no pain]), we dichotomized the outcomes using clinically important pain thresholds. For the VAS, pain ratings of at least 5.4 were considered clinically significant. For the SF-MPQ-2, nonzero ratings were considered clinically significant. Generalized estimating equations were used to fit repeated-measures logistic regression models to the dichotomized outcomes. The predictor of interest in all models was needle gauge (30 vs 32). Additional covariates in the VAS model included patient age and order of injection (first injection with the 32- vs the 30-gauge needle). Models for SF-MPQ-2 ratings were unadjusted. Separate models were fit for each injection location (face and arm) and each pain descriptor.

Data were collected from November 20, 2013, through January 16, 2014. Thirty patients were screened for inclusion; of these, 5 did not meet inclusion criteria and 5 declined to participate. Twenty patients were enrolled, and all of them completed the study per protocol. Demographic characteristics are presented in Table 1. Descriptive statistics for pain ratings using the VAS and differences in mean pain level associated with anatomic site and needle size are displayed in Table 2. Comparison of the number of patients experiencing clinically significant pain and the mean level of such pain by needle size and anatomic site are presented in Table 3. Descriptive results showing the types of pain experienced by patients are presented in Table 4. Four SF-MPQ-2 pain descriptors (cramping, gnawing, tiring or exhausting, and itching) were rated 0 by every patient for each treatment and are not presented.

Mean (SD) injection pain scores at the face (3.41 [2.31] with the 32-gauge needle and 4.16 [2.55] with the 30-gauge needle [P = .34]) and the arm (1.21 [1.65] with the 32-gauge needle and 1.66 [2.07] with the 30-gauge needle [P = .45]) were nominally but not significantly lower with 32-gauge needles. However, significantly more patients receiving facial injections with 30-gauge needles experienced clinically significant pain compared with those receiving facial injections with 32-gauge needles (odds ratio, 3.80 [95% CI, 1.05-13.78]; P = .04). For this analysis, the VAS threshold used as the low boundary of clinically important pain was 5.4 of 10.0 (54 mm on a 100-mm scale), which is a level found to be associated with the onset of severe pain by Dworkin et al.¹ As measured by the VAS, 8 patients (40%) experienced clinically important pain during facial injection with the 30-gauge needle compared with 3 patients (15%) with the 32-gauge needle (Table 3). Arm injections did not exhibit clinically significant differences in pain associated with needle type. Pain descriptors reported by patients were similar for both types of injections, and none of the nominal differences in character of clinically important pain were statistically significant (Table 4).

This study found that 30-gauge needles were associated with nominally more injection discomfort than 32-gauge needles. On the other hand, clinically significant pain, defined as pain greater than an empirically determined threshold, occurred significantly less frequently when the forehead and glabellar complex were injected with 32-gauge needles. That this difference was seen on the face but not the arm may be due to the greater density of sensory structures on the head and neck.

This study is important because it suggests that for a subset of patients prone to experience clinically significant pain on injection with botulinum toxin type A to the face, such pain can be mitigated by selection of smaller-bore needles. The study thus extends the work of prior investigators^4,5 who similarly noted an overall nominal but not significant benefit of 32-gauge needles but did not continue their analysis to show the reduced incidence of clinically significant pain accruing from the use of 32-gauge needles.

As such, this study extends prior work on needle diameter and injection pain by detecting, for the first time, to our knowledge, reduced severe pain when 32-gauge needles were used for neurotoxin injections. Prior split-face investigations of botulinum toxin type A injections performed by Price et al⁴ and Yomtoob et al⁵ also found overall nominal differences in pain that favored 32-gauge needles. Price et al⁴ reported a mean pain level of 4.30 in the 32-gauge needle group and 4.57 in the 30-gauge needle group. Yomtoob et al⁵ noted a level of 3.98 for 32-gauge needles and 4.38 for 30-gauge needles. However, these earlier studies had several methodologic limitations that may have prevented the authors from finding the statistically significant subgroup differences that we uncovered in the present study.

Specifically, Price et al⁴ always treated the left side with 30-gauge needles, and hence did not randomize which side received which needle type. Because manual dexterity, hand strength, and even the degree of photodamage can vary based on laterality, one side of the forehead may be inherently more or less sensitive in any given cohort. In addition, whereas Price et al⁴ appropriately masked patients regarding assignment, the physicians who administered the injections were not blinded. This situation is problematic because minute inadvertent adjustments in injection technique by an unblinded treating physician could bias the outcome; for an operator-dependent outcome such as injection pain, masking of the treating physicians is crucial. Similarly, Yomtoob et al⁵ did not randomize patients as to which side of the face received which needle size, and investigators injecting toxin and those recording pain scores were not blinded. Furthermore, Yomtoob and colleagues⁵ injected only the periocular areas, and their primary indication for injection was treatment of benign essential blepharospasm rather than cosmetic wrinkle reduction.

Other investigations of needle pain have been performed. Skiveren et al⁶ compared the pain associated with 27- vs 30-gauge needles used to inject neurotoxin into the axilla and found a minimal difference that was statistically significant only in a subset of conditions. Flanagan and colleagues⁷ detected no difference in pain when 25- and 27-gauge needles were compared during several types of dental procedures in a cohort of 810 patients. Arendt-Nielsen et al⁸ tested needles of various external diameters on the skin of the thigh and abdomen and found that repeated insertions with 32-gauge needles were 45% less likely to elicit pain than similar insertions with 30-gauge needles (P = .003). However, these investigators did not deliver any injectant through the needles they used to pierce the skin, as did the present study; in addition, they did not compare 30- and 32-gauge needles for facial injections.

We believe that the present study has extended the important prior work of other investigators^4-8 by focusing on methodologic improvements that have significantly mitigated inadvertent bias while allowing more precise measurement of pain. Unlike the prior investigators, we randomized the patient side injected with each needle type and blinded the patients and treating physicians to eliminate unconscious bias in both. Furthermore, to optimally measure the incremental effect of needle-associated discomfort, we deemed it appropriate to reduce other modifiable sources of procedure discomfort so that the resulting needle-related difference could be perceived against the background of a quiet rather than a noisy baseline pain level. Therefore, we reconstituted the toxin with saline containing benzyl alcohol so that the burning sensation associated with unpreserved saline entering the skin did not obscure the relative differences in needle bore–associated discomfort.

This study has certain limitations. First, although we attempted to ensure a uniform technique for both treating dermatologists that was calibrated side-by-side on prestudy test patients, other physicians may use techniques so different that needle bore may not affect patient discomfort. Second, this study did not address the issue of internal needle aperture, which now can be varied in some cases for needles of similar external diameter or gauge.⁹

Evidence now exists that when 32- vs 30-gauge needles are used for facial injections of the forehead and glabellar complex with botulinum toxin type A, the proportion of patients experiencing clinically important pain is reduced significantly. Although this finding may appear intuitive given that 32-gauge needles have an external diameter about 25% smaller than 30-gauge needles (0.23 mm [0.009 inches] vs 0.31 mm [0.012 inches]),¹⁰ the current results are a departure from prior work, which has been more equivocal. To minimize avoidable expense, routine use of costly 32-gauge needles could be restricted to the subset of patients who have in earlier treatments encountered clinically significant pain with larger-bore needles.

Accepted for Publication: June 6, 2015.

Corresponding Author: Murad Alam, MD, MSCI, Department of Dermatology, Feinberg School of Medicine, Northwestern University, 676 N St Clair St, Ste 1600, Chicago, IL 60611 (m-alam@northwestern.edu).

Published Online: September 9, 2015. doi:10.1001/jamadermatol.2015.2232.

Author Contributions: Dr Alam had full access to the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Alam, Sadhwani, Tung, Minkis.

Acquisition, analysis, or interpretation of data: Alam, Geisler, Sadhwani, Goyal, Poon, Nodzenski, Schaeffer, Minkis.

Drafting of the manuscript: Alam, Geisler, Sadhwani, Schaeffer, Tung, Minkis.

Critical revision of the manuscript for important intellectual content: Alam, Goyal, Poon, Nodzenski, Schaeffer, Tung.

Statistical analysis: Nodzenski, Schaeffer, Minkis.

Obtained funding: Alam.

Administrative, technical, or material support: Alam, Poon, Minkis.

Study supervision: Alam, Sadhwani, Tung, Minkis.

Conflict of Interest Disclosures: Dr Alam is employed at Northwestern University; has been a consultant for Amway and Leo Pharma, both unrelated to this research; receives royalties of less than $5000/y from Elsevier for technical books he has edited; and has been principal investigator on studies funded in part by Allergan, Bioform, Medicis, and Ulthera. Northwestern University has a clinical trials unit that receives grants from corporate and governmental entities to perform clinical research. In all cases, grants and gifts in kind have been provided to Northwestern University and not Dr Alam directly, and Dr Alam has not received any salary support from these grants. No other disclosures were reported.

Funding/Support: This study was supported by departmental research funds, Department of Dermatology, Northwestern University.

Role of the Funder/Sponsor: The funding source had a role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.