Impact of Anesthetic Cream and Vibration on Injection Pain in Lymphedema

Abstract

Background:

Although lymphaticovenous anastomosis (LVA) is commonly performed under general anesthesia, some centers perform it under awake, local anesthesia. Reducing injection pain is crucial in patients undergoing LVA for lymphedema under local anesthesia. This study examined the effects of a topical anesthetic cream (eutectic mixture of local anesthetics [EMLA]) and vibratory stimulation on injection pain, assessed using a visual analog scale (VAS) in patients with lymphedema. The association between pain and the severity of lymphedema (as classified by the International Society of Lymphology stage) was assessed.

Methods:

We enrolled 47 patients with lymphedema who underwent LVA at our institution. In total, 197 sites (394 injections) were analyzed. For each patient, up to 5 sites were designated by varying EMLA application time and the use of vibratory stimulation. Immediately following the local anesthetic injection, we measured the VAS score at each site and analyzed the data.

Results:

Patients with severe lymphedema had low VAS scores (P < 0.001), suggesting a reduced pain sensitivity. Prolonged EMLA application was associated with lower VAS scores (P < 0.05). Vibratory stimulation did not significantly impact the VAS scores (P = 0.59). Age, as well as history of chemotherapy and radiotherapy, did not significantly influence the VAS scores. No adverse events were reported.

Conclusions:

Higher lymphedema severity was associated with lower VAS scores, suggesting reduced pain sensitivity in such patients. EMLA cream application time was associated with a reduction in injection pain in patients with lymphedema. In contrast, the vibratory stimulation device demonstrated no significant benefits, warranting further investigation.

Takeaways

Question: Does topical anesthetic cream (EMLA) or a vibratory stimulation device reduce injection pain in lymphedema patients, and how is pain related to disease severity?

Findings: In a retrospective study of 47 patients (394 injections), EMLA applied for 60 minutes or more modestly reduced visual analog scale pain, vibratory stimulation had no significant effect, and higher International Society of Lymphology stages were associated with lower pain scores.

Meaning: For lymphedema patients, EMLA can modestly reduce injection pain, whereas vibratory stimulation provides no consistent benefit. Pain perception may also be lower in patients with more advanced International Society of Lymphology stages.

INTRODUCTION

Although lymphaticovenous anastomosis (LVA) is commonly performed under general anesthesia in many centers, our findings pertain to an outpatient, awake local anesthesia pathway. LVA is a viable surgical option for the treatment of lymphedema. Recent studies have reported the usefulness of LVA.¹ To optimize therapeutic outcomes, multiple anastomoses should be performed at various sites on the affected limb.² During surgery, local anesthesia was administered at each site. Therefore, effective pain management during LVA is important for reducing patient discomfort, ensuring treatment continuity, and improving patient satisfaction. Injection-related pain typically comprises 2 components: pain from needle insertion (needle puncture pain) and pain from infiltration of the anesthetic solution into the tissue (injection pain). Needle puncture pain occurs when a needle stimulates nociceptors in the skin. Conversely, injection pain arises from tissue distension or the characteristics of the anesthetic solution.

Various methods have been used to reduce pain during injections.³ Eutectic mixture of local anesthetics (EMLA), a topical anesthetic cream containing 2.5% lidocaine and 2.5% prilocaine, is widely used to minimize pain by numbing the skin during venipuncture and injections. EMLA can improve visual analog scale (VAS) scores by more than 30%.^4,5 Numerous studies have confirmed the analgesic efficacy of EMLA on intact healthy skin. Moreover, EMLA blocks sensory nerves in the epidermis and superficial dermis, demonstrating effectiveness 60 minutes after application.^6,7 An even prolonged application time is required to achieve analgesia at tissue depths exceeding 5 mm.⁸ The implementation of gate-control theory using vibratory stimulation^9,10 has also garnered significant attention. Small vibrating devices are used to alleviate injection-related pain. In a randomized controlled trial by Kazi et al,¹¹ the use of a vibratory device during local anesthetic injection in dermatologic surgery significantly reduced VAS scores. A systematic review of pediatric dentistry also reported that vibratory stimulation significantly reduced pain scores during local anesthesia.¹² The effectiveness of the vibratory stimulation has been confirmed across various clinical fields.¹³

However, in patients with lymphedema, chronic lymph fluid stasis and inflammation lead to skin thickening, fibrosis, and fat deposition.^14–17 Consequently, the structure of the skin and subcutaneous tissue in these patients differs from that of healthy adults. Sensory changes have also been reported previously. In a before-and-after study by Baran et al,¹⁸ sensory decline was assessed using a Semmes–Weinstein monofilament, 2-point discrimination, and pressure pain threshold, all of which correlated with the severity of lymphedema. In a large-scale analysis by Li et al,¹⁹ subjective symptom scores, including chronic pain and skin tightness, correlated with lymphedema severity. Conversely, Lee et al²⁰ reported that the sensory indicators exhibited little association with International Society of Lymphology (ISL) staging. Therefore, no consensus exists regarding pain sensitivity in patients with lymphedema.

The impact of skin and sensory changes in patients with lymphedema on pain during local injection, as well as the extent to which EMLA or vibratory devices can alleviate this pain, remains unclear. Improving pain management under local anesthesia may enable more patients to undergo LVA without general anesthesia.

This study aimed to assess the effectiveness of pain relief methods during local anesthesia in patients with lymphedema. Pain during local anesthetic injection at the time of LVA in patients with lymphedema was evaluated using the VAS. We examined the effects of prolonged EMLA application and the efficacy of using a vibratory device. We also examined whether pain relief varied according to lymphedema severity, as classified by the ISL staging system.

METHODS

Participants

We included 47 patients with limb lymphedema who underwent LVA under local anesthesia at our hospital between February and July 2025. We retrospectively reviewed the surgical records and medical charts to obtain data on patient characteristics, injection techniques, and pain intensity. The ethics committee of Aichi Medical University School of Medicine approved this study (approval number 2024-303). Protocol amendments approved in February and July 2025 covered the present retrospective analysis. Informed consent was waived by the institutional review board due to the retrospective nature of this study.

Intervention and Measurement

Pain during local anesthesia with 1% lidocaine was assessed intraoperatively during LVA using the following protocol. For each patient, local anesthetic injections were administered at up to 5 distinct surgical sites per operation (eg, left lower leg, left thigh, and right lower leg), as shown in Figure 1. At each site, 2 consecutive local anesthetic injections were administered, 1 with and 1 without a vibratory stimulation device. Pain VAS scores were compared at identical sites within the same patient. VAS (0–10) was obtained immediately after each injection using a 10-cm line (0 = no pain, 10 = worst pain). The sequence of vibratory stimulation (first or second) was predetermined for each patient: 11 received vibratory stimulation during the first injection, whereas 36 received it during the second. The vibratory device used was a commercially available small pen-shaped vibrating instrument with a frequency of approximately 100 Hz. Based on the gate-control theory, the device was positioned near the central aspect of the injection site. The device was activated approximately 5 seconds before insertion and applied throughout the injection, either continuously or intermittently. The device used was the “BullBull pen” (JMEC Co., Ltd.), which lacked a cooling function; therefore, vibration was applied without concurrent cooling.

Fig. 1.

Injection sites and vibration placement. Triangular markers indicate local anesthetic injection sites placed more than 2 cm apart, and the arrow indicates the location of vibratory stimulation applied approximately 5 cm proximal to the injection site.

At each injection site, the effects of varying EMLA application durations were evaluated before local anesthesia administration. No cream was applied at the first site (initial injection site). For the second site, local anesthesia was administered 30 minutes after cream application; for the third site, after 60 minutes; for the fourth site, after 90 minutes; and for the fifth site, after 120 minutes. A thick layer of cream was applied and covered with a film dressing. The cream was removed immediately before injection.

Immediately following each injection, patients rated their pain intensity using the VAS, and the scores were recorded. For each injection, we documented the EMLA application duration, use of vibratory stimulation, ISL severity stage, injection site (upper or lower limb), underlying conditions, and history of chemotherapy or radiotherapy. However, due to the uneven distribution of cases across injection sites and underlying diseases, these variables were excluded from the subsequent statistical analyses. In our protocol, EMLA time was coupled with the within-session injection order, and thus, residual order-related confounding cannot be fully excluded.

Statistical Analysis

Statistical analyses were performed using EZR (ver. 1.75; Saitama Medical Center, Jichi Medical University, Saitama, Japan), a graphical user interface based on R Commander. EZR is a free software package that operates within the R statistical environment (The R Foundation for Statistical Computing, Vienna, Austria).²¹ As VAS scores represented repeated measurements within individuals, we applied a linear mixed-effects model (lme4 package) using participant identification as a random effect (random intercept). For fixed effects, we included the EMLA application time (0, 30, 60, 90, and 120 min), use of vibratory stimulation, and ISL classification (1, 2a, 2b, and 3). Furthermore, sex, age, and history of chemotherapy or radiotherapy were included as covariates in the model. We examined the timing of vibratory stimulation (stimulation first versus stimulation second), as well as the 2-way interactions between the duration of EMLA application and ISL classification, and between the use of vibratory stimulation and ISL classification. The P values for fixed effects were calculated using the Satterthwaite approximation (lmerTest package), and statistical significance was defined as a 2-sided probability of less than 0.05. Due to the uneven distribution of cases across underlying diseases and injection sites, these variables were excluded as fixed effects in the model. Model parsimony and fit were assessed using Akaike information criterion and likelihood-ratio tests under maximum likelihood; the main-effects model is presented primarily, with interaction results provided in Supplemental Digital Content 1. (See figure, Supplemental Digital Content 1, which displays the interaction effects [linear mixed-effects model, random intercept for patient], https://links.lww.com/PRSGO/E803.)

RESULTS

Patient Background

Table 1 presents the patients’ background characteristics, whereas Table 2 summarizes the injection site characteristics.

Table 1.

Patient Characteristics

	Value
Age, median (range), y	60 (31–81)
Mean number of injections per patient	8.4
Sex, n (%)
Female	45 (95.7)
Male	2 (4.3)
Chemotherapy, n (%)
No	22 (42.6)
Yes	25 (57.4)
Radiotherapy, n (%)
No	26 (55.3)
Yes	21 (44.7)
Diabetes, n (%)
No	46 (97.9)
Yes	1 (2.1)

Table 2.

Injection Details

	n (Injections)
Location, n (%)
Thigh	186 (47.2)
Lower leg	166 (42.1)
Forearm	26 (6.6)
Upper arm	16 (4.1)
ISL stage, n (%)
Stage 1	62 (15.7)
Stage 2a	72 (18.3)
Stage 2b	190 (48.2)
Stage 3	70 (17.8)
EMLA application time, min, n (%)
0	96 (24.3)
30	94 (23.9)
60	84 (21.3)
90	68 (17.3)
120	52 (13.2)
Vibratory stimulation, n (%)
No	197 (50.0)
Yes	197 (50.0)
Timing of vibratory stimulation, n (%)
First injection	48 (12.2)
Second injection	149 (87.8)

Effect of Interventions on Pain Scores

We analyzed the influence of the EMLA application duration and vibratory stimulation on pain scores (VAS) during local anesthesia administration using a linear mixed model. The results are summarized in Table 3.

Table 3.

Results of Linear Mixed Model Analysis for Pain Scores (VAS) During Local Anesthetic Injection (Main-effects Model)

Variable	Estimate (β)	SE	t	P
Intercept	3.21	1.38	2.33	0.024
EMLA application time, min
30	−0.14	0.19	−0.77	0.44
60	−0.55	0.20	−2.82	0.005
90	−1.01	0.23	−4.37	<0.001
120	−1.13	0.26	−4.42	<0.001
Vibratory stimulation
Yes	−0.14	0.26	−0.54	0.59
Timing of vibratory stimulation
Second	0.51	0.29	1.74	0.08
ISL stage
2a	−1.39	0.33	−4.22	<0.001
2b	−1.40	0.30	−4.65	<0.001
3	−2.20	0.48	−4.57	<0.001
Age, y	0.03	0.02	1.35	0.18
Chemotherapy (yes)	−0.99	0.53	−1.86	0.07
Radiotherapy (yes)	0.62	0.54	1.16	0.25

Reference categories: EMLA 0 minutes, ISL stage 1, vibration absent, vibration timing “first,” chemotherapy and radiotherapy absent.

Legend: Estimates represent the change in VAS score compared with the respective reference group, as estimated by the linear mixed model with random intercepts for patient identification.

SE, standard error; β, estimated coefficient.

In the interaction model, none of the interactions (EMLA application time × ISL severity, vibratory stimulation × ISL severity, or vibratory stimulation timing) were statistically significant (all P > 0.05). Model comparison under maximum likelihood favored the main-effects specification (Akaike information criterion 1450.6 versus 1467.1; likelihood-ratio test χ²(15) = 10.68, P = 0.55); accordingly, we present the main-effects model as a primary (Table 3), with interaction results in Supplemental Digital Content 1.

In the main-effects model, EMLA application times of 60 minutes or longer were associated with a significant reduction in VAS pain scores (60 min: β = −0.552, P = 0.005; 90 min: β = −1.013, P < 0.001; 120 min: β = −1.127, P < 0.001). In contrast, the application of vibratory stimulation did not significantly affect VAS pain scores (β = −0.138, P = 0.590).

However, higher ISL severity was significantly associated with lower VAS pain scores (ISL 2a: β = −1.385, P < 0.001; ISL 2b: β = −1.404, P < 0.001; ISL 3: β = −2.204, P < 0.001). The timing of vibratory stimulation (first versus second injection) did not result in a statistically significant difference (β = 0.500, P = 0.091). Age (β = 0.029, P = 0.183), sex (male: β = −1.606, P = 0.179), history of chemotherapy (β = −0.990, P = 0.069), and history of radiotherapy (β = 0.618, P = 0.254) did not exert a statistically significant effect on VAS pain scores.

DISCUSSION

In this study, we examined the effects of EMLA and a vibratory stimulation device on pain during local anesthetic injection in patients with lymphedema undergoing LVA surgery. We observed that patients with higher severity exhibited lower VAS pain scores during the injection. Prolonged EMLA application (≥60 min) was associated with lower VAS scores. By contrast, the vibratory stimulation device did not demonstrate significant analgesic effects. Neither intervention demonstrated a differential effect based on lymphedema severity.

Lymphedema Severity and Pain Sensitivity

In this study, we discovered an inverse correlation between lymphedema severity and the VAS scores during local anesthetic injection. The mean VAS score in patients with stage 3 lymphedema was 2.2 points lower than that of patients with stage 1 (P < 0.001). To the best of our knowledge, this is the first study to describe the aforementioned relationship. Pain during local anesthetic injection arises from both superficial needle puncture and acute discomfort caused by pressure or chemical irritation from the injected agent. This mechanism may differ from the chronic symptoms commonly reported in the literature. As previous studies have demonstrated a correlation between lymphedema severity and reduced superficial skin sensation,¹⁵ the reduced sensation in the skin and subcutaneous tissue, which was also correlated with severity in our study, may have contributed to the altered perception of nociceptive pain.

Effect of EMLA and Its Application in Lymphedema

We hypothesized that the analgesic effect of EMLA would diminish in patients with lymphedema due to thickening of the skin and subcutaneous tissues. However, our study did not identify a reduced effectiveness of EMLA, even in patients with severe lymphedema. One possible explanation is that, although skin thickening and fibrosis in lymphedema may prevent deep anesthetic penetration, reduced clearance could permit the drug to persist at high concentrations locally, thereby maintaining an adequate analgesic effect. In our protocol, EMLA application time was coupled with the within-session injection order, and thus, residual order-related confounding due to habituation to the injection procedure cannot be fully excluded. Nevertheless, VAS decreased monotonically with longer application times, which may suggest a time–response relationship that is not readily explained by order habituation alone, even when modified by habituation and accumulating anxiety.

The reduction in VAS did not achieve the minimum clinically important difference of approximately −1.3, as proposed by Todd et al.²² A possible explanation is that the mean VAS score in this study was already low (3.21), potentially limiting the magnitude of further reduction. However, no plateau in the analgesic effect was observed even after 120 minutes of application. This suggests that the onset of EMLA’s analgesic effect may be delayed in lymphedematous tissues compared with that in healthy skin. Some reports have indicated that the peak analgesic effect occurs 1–2 hours after the removal of EMLA.⁸ Accordingly, instituting some waiting period after cream removal—even when application has already reached the conventional 120-minute maximum—may further reduce injection pain and, in some patients, help the reduction reach or exceed the approximately 1.3-point minimum clinically important difference, yielding a clinically perceptible benefit.

Another consideration is the potential influence of systemic absorption of local anesthetics. In this study, the maximum total dose of 1% lidocaine was approximately 200 mg and was administered by local infiltration. Because subcutaneous infiltration results in gradual systemic uptake with low peak plasma concentrations, and edematous tissue may further delay absorption, systemic analgesic effects of lidocaine or EMLA are unlikely to explain the observed pain reduction.

Effect of Vibratory Stimulation Device and Its Application in Lymphedema

In this study, the use of a vibratory stimulation device did not reduce pain during the injection. Additionally, pain may exacerbate in some patients. The analgesic effect of the vibratory stimulation may have been diminished due to the distinctive skin characteristics observed in patients with lymphedema. However, in this study, the interaction between vibratory stimulation and lymphedema severity was not statistically significant. Previous studies have indicated that skin manifestations in patients with lymphedema do not consistently correlate with disease severity.²⁰ The physical sensation of applying a vibratory device to the affected skin may cause patient discomfort, thereby lowering the pain threshold. The timing of vibratory stimulation (first versus second injection) did not result in a statistically significant difference (β = 0.500, P = 0.091); however, pain scores tended to be higher when vibration was applied during the second injection. In this study, the order of vibratory stimulation was predetermined for each patient; however, the sample sizes were unequal, with 11 patients receiving stimulation first and 36 receiving it second. Consequently, patients may have experienced greater apprehension before the second injection, potentially increasing anxiety and lowering the pain threshold. This potential order effect may have obscured the analgesic impact of vibratory stimulation, resulting in no overall difference between the 2 conditions. In addition, in this study, vibratory stimulation was applied for approximately 5 seconds before injection. It is possible that a longer duration of stimulation, as described in some discussions of gate-control theory, may have allowed detection of an analgesic effect.

Effects of Other Factors

Age, sex, and history of radiotherapy did not affect the pain scores. Patients with a history of chemotherapy tended to report lower pain scores on VAS. Although peripheral neuropathy may have contributed to decreased sensation in these cases,²³ the number of cases was limited, and the difference was not statistically significant.

Study Limitations

This study has several limitations. First, pain scores were not separated into puncture pain and injection pain. Second, for vibration, psychological and attentional influences could not be isolated, and the timing was nonrandomized and imbalanced (11 first versus 36 second), introducing potential allocation bias. Third, EMLA application time was coupled with within-session injection order; thus, residual order-related confounding cannot be excluded, although a monotonic decrease in VAS with longer application times was observed. These issues merit future study, including counterbalancing order and separating puncture from injection pain.

CONCLUSIONS

Patients with greater ISL severity exhibited significantly lower VAS scores during injection, suggesting that sensory attenuation in response to local anesthetic injections may progress with increasing ISL severity. The application time of the topical anesthetic cream (EMLA) was associated with a reduction in pain during local anesthetic injections in patients with lymphedema, similar to the reductions observed in healthy individuals. However, because the application time was coupled with the injection order, the observed reduction could be influenced by habituation, and this limitation should be considered when interpreting the results. In contrast, the vibratory stimulation device did not exhibit a significant analgesic effect, warranting further investigation into its clinical utility.

Evidence regarding analgesic strategies during local anesthetic injection in patients with lymphedema—and how pain perception may vary by disease severity—remains limited. The present findings may contribute to safer, more accessible LVA under local anesthesia and support the development of minimally invasive protocols that reduce reliance on general anesthesia.

DISCLOSURE

The authors have no financial interest to declare in relation to the content of this article.

ACKNOWLEDGMENTS

The authors used ChatGPT, a generative artificial intelligence tool developed by OpenAI, to assist in language editing and improve the clarity of the article text. In addition, the authors used Editage (Cactus Communications) for professional English language editing. All content, data, and references were generated by the authors and verified for accuracy by the authors before submission. EMLA cream (lidocaine 2.5%/prilocaine 2.5%) (Sato Pharmaceutical Co., Ltd., Tokyo, Japan) and BullBull pen (JMEC Co., Ltd., Tokyo, Japan) were also used in the study.