Although the rate is dropping, circumcision is still a common procedure. In 1984, circumcision was performed on 76.4% of male newborns in the United States.1 In 1989, the American Academy of Pediatrics issued a statement of concern about the necessity of a procedure2 that Kirya and Werthmann called “perhaps the most extensive noxious procedure performed on healthy male infants.”3 Evidence that a majority of physicians use no analgesia is more disturbing.4 Beliefs that newborns have immature nervous systems or are incapable of remembering pain have long ago been dispelled.5–12 Several methods of analgesic delivery are used in newborn circumcision; however, dorsal penile nerve block or local infiltration around the corona initially causes pain and bleeding risks even before circumcision. Topical agents are safer because they have no inherent bleeding-related side effects, produce low systemic absorption, and are easy to administer.8,9
Analgesic agents block the behavioral and physiologic changes of pain associated with neonatal circumcision.6,10–12 These include crying and characteristic facial and body movements,2,5–12 increases in heart rate and blood pressure (BP),2,7–12 and decreases in oxygen saturation.7,9,10,12 Topical 5% lidocaine-prilocaine and 30% lidocaine creams have proved effective in pain control during circumcision.8,9,13 My goal in this study was to change the regional practice of performing circumcisions without analgesia by determining the most effective topical anesthetic for circumcision, thereby preventing the pain and complications associated with injections.
A commercial 5% lidocaine-prilocaine cream (EMLA; Astra USA, Inc., Westborough, MA) was used that is a 1:1 oil/water emulsion of a eutectic mixture of 2.5% lidocaine and 2.5% prilocaine. The safety and efficacy of 5% lidocaine-prilocaine have been well studied.8,13,14 Thirty percent topical lidocaine for circumcision resulted in good pain control with no systemic absorption.2,8,15 Thus, good data suggest adequate pain control using 5% lidocaine-prilocaine and 30% topical lidocaine cream. However, an extensive literature search of MEDLINE from 1966 to 1996, using the terms “anesthetic,” “topical anesthetic,” “circumcision,” “pain,” and “neonatal pain” uncovered no published studies comparing the efficacy of these two agents.
Thirty percent lidocaine for circumcision produces an average serum lidocaine level of 0.27 ± 19 μg/mL.8 Similar levels of 5% lidocaine-prilocaine produce at least 7.5 times lower serum lidocaine and prilocaine levels, based on data of Engberg et al.14 Lidocaine and prilocaine act synergistically, lowering the total dosage of anesthetic necessary to effect a similar response. In the present study, I expected 30% lidocaine to be most effective, followed by 5% lidocaine-prilocaine. I expected less severe response to pain after administration of the topical anesthetic before the procedure, shown by a lesser decrease in oxygen saturation, less increase in heart rate, fewer seconds of total crying time, and less increase in systolic and diastolic BPs.
Materials and Methods
Research approval was obtained from the Institutional Review Board and Departments of Pediatrics and Obstetrics and Gynecology at the hospital where the study was conducted. Study procedures were in accordance with the ethical standards for human experimentation established by the revised Declaration of Helsinki (1983). All healthy male neonates born between October 1995 and September 1996 with no contraindications to circumcision were eligible. I planned to exclude infants with hypospadius or family histories of hemophilia A, but no infants presented with such conditions. One hundred forty-two neonates met the eligibility criteria. Written informed consent was obtained from the parents requesting circumcision for their sons. Sixty-one neonates met the selection criteria with parental consent.
The 61 infants were assigned to one of three groups using a computer-generated random list: 5% lidocaine-prilocaine (n = 20), 30% lidocaine (n = 20), and control (n = 21). The 30% lidocaine cream was formulated by the center's pharmacy and suspended in an acid-mantle base. The control treatment was a plain acid-mantle cream. Topical formulations were placed in identical vials and blindly labeled A (30% lidocaine), B (control), and C (5% lidocaine-prilocaine). The investigator was blinded to the contents of each vial until completion of statistical analysis.
Enrollment criteria were listed on an information sheet given to the parents of potential subjects. Study newborns were required to have had a gestational age between 37 and 42 weeks, an uncomplicated vaginal or cesarean delivery, a 5-minute Apgar score of 7 or more, and a birth weight greater than 2500 g. Postnatal age had to be between 6 and 72 hours, and parental request for circumcision was required.
All infants fasted for at least 1 hour before circumcision. After informed consent was obtained, 1 mL (approximately 1 g) of cream from one of the study groups was applied to the prepuce and covered with a small square of clear plastic wrap to form an occlusive bandage. The bandage was in place for 1 hour. Blood pressure was measured with a pediatric sphygmomanometer before the subjects were restrained and after the circumcisions were completed. A pulse oximeter (Nellcor Symphony 3000; Nellcor, Inc., Haywood, CA) was applied for continuous monitoring. Unrestrained, resting, baseline pulses and oxygen pressure (PO2) levels were measured. The infants were then strapped to a circumcision board (CIRCUMSTRAINT; Olympic Surgical Co., Seattle, WA) and prepared and draped in a sterile manner. Bell-shield clamps (Gomco Inc., St. Louis, MO) were used for circumcision. Pulse and PO2 were measured during each of seven stages of the circumcision: baseline restraint, initial clamping of the foreskin, adhesiolysis, dorsal clamp, application of the bell clamp, tightening of the bell clamp, and removal of the bell clamp. All procedures were videotaped for future analysis. All circumcisions were done by the same operator.
Pulse and PO2 measurements were recorded on videotape by the operator. For each step of the procedure, the pulse recorded was the peak heart rate sustained during or immediately after that step. Oxygen saturation recorded was the nadir sustained during or immediately after each interval. Crying recorded was the time spent crying during each step, timed if the facial characteristics of pain noted by Brazelton (brow bulge, eye squeeze, nasolabial furrow, open lips, horizontal and vertical mouth stretch, and taut tongue9) were present with or without audible cry. Timing ended the second crying ceased. All procedures were coded by the same operator, who was blinded to the analgesics. Time of each step recorded was the beginning of each step to the end of the objective measurements at the end of each step.
The data for pulse, oxygen saturation, time for each step, time spent crying, and BP were analyzed statistically using one-way analysis of variance. Differences were considered significant at two-tailed P < .05. Values were expressed as mean ± standard deviation (SD). A power analysis incorporated the anticipated difference of 10%, SD of 10, intermediate dispersion of means, and alpha of .05 for an effect size of 1 regarding heart rates and BPs. This resulted in a power of 0.79 with 20 subjects in each of the three groups. Isolated data points missed because of gaps in pulse oximeter readouts were excluded. Differences in physiologic outcomes among the groups were analyzed by one-way analysis of variance, and differences between before-and after-procedure data points were analyzed by dependent (matched) Student t test.
Sample characteristics were similar for each study group (Table 1). Differences before the procedure were controlled by randomization. The three groups were not significantly different in birth weight, age, baseline pulse, baseline PO2, baseline BP, time to complete the procedure, and type of delivery. Apgar scores at 1 and 5 minutes were similar across groups (P = .57).
There were no significant differences in oxygen saturation during any phase of the circumcision among the groups (Table 2). There was a uniform decrease in oxygen saturation for all groups during circumcision, with the nadir (average) at the same stage of circumcision (adhesiolysis) for each group.
Heart rate increased significantly in all groups during circumcision (P < .01). The 5% lidocaine-prilocaine group had the smallest increase in heart rate in each of four (of six) active phases of circumcision: placement of the dorsal clamp, application of the bell clamp, tightening of the bell clamp, and removal of the bell clamp (P < .01). Increases in heart rate for the 30% lidocaine group were significantly lower than for the placebo group during the same four phases. Table 3 presents the group means for each phase.
The mean peak heart rate occurred at different phases of circumcision for each group. The mean peak heart rate for the 5% lidocaine-prilocaine group occurred during the restraint phase (146 ± 16 beats per minute), for the 30% lidocaine group during placement of the dorsal clamp (157 ± 10 beats per minute), and for the control group during tightening of the bell clamp (164 ± 16 beats per minute).
Duration of crying was significantly different during tightening of the bell clamp (P < .001), but was statistically similar during the other phases of circumcision (Table 4). Although not statistically significant, both anesthetic groups uniformly showed fewer seconds of crying than the placebo group during each phase of circumcision. The 5% lidocaine-prilocaine and 30% lidocaine groups each produced less crying during three (of six) active phases of circumcision. There were no significant differences in BP across the three groups at baseline or after circumcision (P > .13 and P > .16, respectively).
There was insufficient power to conduct a two-way analysis of variance to compare BPs before and after the procedure. However, it was interesting to note the comparison using a matched Student t test (two-tailed significance). There remained a significant increase in systolic BP from baseline for the 30% lidocaine group (t = 4.8, P = .001) and the control group (t = 2.5, P = .023) after circumcision. There was also a significant difference over time in diastolic BP for the 30% lidocaine (t = 2.9, P = .009) and control (t = 2.3, P = .032) groups. After circumcision, systolic BP (t = 1.6, P = .12) and diastolic BP (t = 1.9, P = .067) in the 5% lidocaine-prilocaine group did not differ significantly from baseline.
Newborns who received 5% lidocaine-prilocaine or 30% lidocaine cream applied 60 minutes before circumcision showed attenuated pain responses, evidenced by smaller increases in heart rate and less crying associated with pain than newborns who received placebo. Newborns who received 5% lidocaine-prilocaine did not have significant increases in systolic and diastolic BPs, unlike the 30% lidocaine and placebo groups. Heart rate seemed to show the most consistent changes across study groups, hypothesized to be one of the most sensitive indicators of the pain response in newborns.2 Five percent lidocaine-prilocaine suppressed heart rate increases better than 30% lidocaine, which in turn was better than placebo (P < .01). Both anesthetics worked better than placebo.
Measured crying time was consistently lower in the anesthetic groups compared with placebo. Five percent lidocaine-prilocaine and 30% lidocaine were each more efficacious during three of six active phases of circumcision. However, infants cried for significantly fewer seconds during tightening of the bell clamp with 5% lidocaine-prilocaine than with 30% lidocaine, which in turn was better than placebo (P < .01). Placement of the bell clamp, as assessed by gross crying time (mean 61 ± 38 seconds) and percentage of time spent crying (mean 66.2%), was the most painful event during circumcision, but did not differ significantly across groups (P = .24). Crying time is a well-documented objective measure of infant pain response when used in conjunction with others.5,6
Increases in BP would be expected during a stressful procedure such as circumcision. One study found that local and general anesthetics blocked such increases.5 Comparing BPs before and after the circumcision, the 5% lidocaine-prilocaine group showed no significant systolic or diastolic increases, unlike the 30% lidocaine and placebo groups (P < .05). Although an interaction with time was possible, the study's sample size was too small to test such an interaction. The relative blockage of this physiologic indicator of pain in the 5% lidocaine-prilocaine group supports my contention of its efficacy.
Expected differences in oxygen saturation across groups were not seen. Mixed results have been shown before.2,7–10,12 In studies in which oxygen saturation was not significantly different, the anesthetic groups were small (n = 20, 15).2,8 The anesthetic groups in the present study were also small (n = 20), so there was a likely lack of power to detect significance for this characteristic. Differences in the partial pressure of oxygen might have been significant, although oxygen saturation was not. The neonatal oxygen-dissociation curve, on which oxygen saturation depends, had a small slope in this study because the data fell in the range of 96–99%.
A surprising result of the present study was the evidence that 5% lidocaine-prilocaine was more effective than 30% lidocaine. If the 5% lidocaine-prilocaine mixture was purely additive, or even moderately synergistic, I would still expect the serum levels of anesthetic to be higher in the 30% lidocaine group. However, the skin can function as a reservoir for lipophilic drugs, and because absorption of drugs applied topically to the skin is slow and incomplete,16 serum levels may not accurately parallel tissue drug levels. Tissue levels of lidocaine and prilocaine could be confirmed in future research by biopsy of anesthetized tissue.
Topical anesthetics for circumcision have several advantages over injectable preparations because they are easy to apply, well tolerated by patients, produce limited systemic absorption, and eliminate the potential vascular complications, stress, and pain associated with dorsal penile nerve block and local injection. A hospital formulation of 30% lidocaine cream has a low per-unit cost, but potential problems arise with homogeneity, shelf life, and standard concentration of each sample. A commercially available 5% lidocaine-prilocaine cream offers standardized dosing, and a long shelf life (36 months) in a prepackaged, premixed form.
The present study supports the use of 5% lidocaine-prilocaine for topical analgesia during neonatal circumcision. The standard dosing of 5–12 mg/kg for both lidocaine and prilocaine can be achieved with most neonates using 1–2 g of 5% lidocaine-prilocaine cream. Both anesthetics have proved safe and effective in previous studies and should not be withheld from neonates because of their perceived lack of pain sensation, memory of pain, or lasting effects. Application of topical anesthetics approximately 1 hour before circumcision can be done quickly and easily by medical or religious personnel with minimal training.
1. Davenport C, Romberg R. 1984 hospital survey of circumcision rates in the United States. Birth 1984;11:247–50.
2. Mudge D, Younger YB. The effects of topical lidocaine on infant response to circumcision. J Nurse Midwifery 1989;34:335–40.
3. Kirya C, Werthmann MW. Neonatal circumcision and penile dorsal nerve block—A painless procedure. J Pediatr 1978;92:998–1000.
4. Wellington N, Reider MJ. Attitudes and practices regarding analgesia for newborn circumcision. Pediatrics 1993;92:541–3.
5. Anand KJ, Hickey PR. Pain and its effects in the human neonate and fetus. N Engl J Med 1987;317:1321–9.
6. Owens ME. Pain in infancy: Conceptual and methodological issues. Pain 1984;3:212–27.
7. Rabinowitz R, Hulbert WC Jr. Newborn circumcision should not be performed without anesthesia. Birth 1995;22:45–6.
8. Weatherstone KB, Rasmussen LB, Erenberg A, Jackson EM, Claflin KS, Leff RD. Safety and efficacy of a topical anesthetic for neonatal circumcision. Pediatrics 1993;92:710–4.
9. Benini F, Johnston C, Faucher D, Aranda JV. Topical anesthesia during circumcision in newborn infants. JAMA 1993;270:850–3.
10. Masciello AL. Anesthesia for neonatal circumcision: Local anesthesia is better than dorsal penile nerve block. Obstet Gynecol 1990;75:834–8.
11. Holve RL, Bromberger PJ, Groveman HD, Klauber MR, Dixon SD, Snyder JM. Regional anesthesia during newborn circumcision. Effect on infant pain response. Clin Pediatr 1983;22:813–8.
12. Williamson PS, Williamson ML. Physiologic stress reduction by a local anesthetic during newborn circumcision. Pediatrics 1983;71:36–40.
13. Taddio A, Stevens B, Craig K, Rastogi P, Ben-David S, Shennan A, et al. Efficacy and safety of lidocaine-prilocaine cream for pain during circumcision. N Engl J Med 1997;336:1197–201.
14. Engberg G, Danielson K, Henneberg S, Nilsson A. Plasma concentrations of prilocaine and lidocaine and methaemoglobin formation in infants after epicutaneous application of a 5% lidocaine-prilocaine cream (EMLA). Acta Anaesthesiol Scand 1987;31:624–8.
15. Lubens HM, Ausdenmoore RW, Shafer AD, Reece RM. Anesthetic patch for painful procedures such as minor operations. Am J Dis Child 1974;128:192–4.
16. Franz TJ. Kinetics of cutaneous drug penetration. Int J Dermatol 1983;22:499–505.