Volume 278 No. 24, pages 2157-2162, December 24/31, 1997

Comparison of Ring Block, Dorsal Penile
Nerve Block, and Topical Anesthesia
for Neonatal Circumcision

A Randomized Controlled Trial

Janice Lander, PhD; Barbara Brady-Freyer, MN; James B. Metcalfe, MD, FRCSC; Shermin Nazerali, MPharm; Sarah Muttit, MD, FRCPC.

Context. - Beliefs about the safety and effectiveness of current anesthetics have resulted in many newborns being circumcised without the benefit of anesthesia.

Objective. - To compare ring block, dorsal penile nerve block, a topical eutectic mixture of local anesthetics (EMLA), and topical placebo when used for neonatal circumcision. The placebo represented current practice, with no anesthetic for neonatal circumcision.

Design. - A randomized controlled trial.

Setting. - Antenatal units in 2 tertiary care hospitals in Edmonton, Alberta.

Participants. - A consecutive sample of 52 healthy, full-term male newborns, aged 1 to 3 days.

Interventions. - Physiological and behavioral monitoring occurred in a series of trials: baseline, drug application, preparation, circumcision, and postcircumcision. Surgical procedures defined the following 4 stages of the circumcision: cleansing, separation, clamp on, and clamp off. Methemoglobin level was assessed 6 hours after surgery.

Main Outcome Measures. - Heart rate, cry, and methemoglobin level.

Results. - Newborns in the untreated placebo group exhibited homogeneous responses that consisted of sustained elevation of heart rate and high pitched cry throughout the circumcision and following. Two newborns in the placebo group became ill following circumcision (choking and apnea). The 3 treatment groups all had significantly less crying and lower heart rates during and following circumcision compared with the treated group. The ring block was equally effective through all stages of the circumcision, whereas the dorsal penile nerve block and EMLA were not effective during foreskin separation and incision. Methemoglobin levels were highest in the EMLA group, although no newborn required treatment.

Conclusions. - The most effective anesthetic is the ring block; EMLA is the least effective. It is our recommendation that an anesthetic should be administered to newborns prior to undergoing circumcision.
JAMA 1997;278:2157-2162

CIRCUMCISION OF male newborns is a controversial procedure. Some favor the procedure as a means of reducing problems such as sexually transmitted diseases and urinary tract infection.1-3 Others view circumcision as unnecessary or disfiguring.4-5 This debate has overshadowed an important problem. Many male newborns who are circumcised in North America do not receive anesthetics6 (64%-96% in some areas7,8).

Researchers have assessed several anesthetics for circumcision. Dorsal penile nerve block, which was first described by Kirya and Werthmannn,9 is the procedure that has been most investigated and recommended. Clinicians and other health care workers, however, have not been swayed by evidence demonstrating its benefits9-11 and safety.12,13 Indeed some physicians maintain that circumcision causes minimal pain, that newborns do not remember pain, or that a quick and efficient circumcision causes less pain than performing a block.7

Eutectic mixture of local anesthetics (EMLA) is a topical anesthetic that is used for minor procedures. It has not yet been recommended for use with neonates because of the risk of acquired methemoglobinemia caused by the newborns immature methemoglobin (metHB) reductase system. Efficacy of EMLA for neonatal circumcision has been evaluated, although its studies have design flaws.14,15 In both studies distress from being restrained coincided with distress from circumcision, and this introduced ambiguity and increased error variance. Another problem was that no established intervention, such as a dorsal penile nerve block or a ring block, was included to compare EMLA's efficacy.

Ring block has been used for children's post circumcision analgesia,16 and there may be merit in using it for circumcision anesthesia in newborns. We were unable to find evidence of the efficacy of this approach when used for surgery among newborns.

Therefore, we planned a randomized, placebo-controlled trial to compare dorsal penile nerve block, subcutaneous ring block and EMLA. Our primary hypotheses were as follows:
1. Newborns receiving placebo will have greater distress during and following circumcision than newborns receiving EMLA, dorsal penile nerve block, or ring block.
2. Application of a topical agent will cause less distress among newborns than infiltration of anesthetic for dorsal penile nerve block or ring block.
3. Distress caused by an unanesthetized circumcision will be greater than that caused by infiltration of the anesthetic.

Time in study,       Topical              Infiltration
min.                 Groups               Groups

  0                  Baseline A           Baseline A

 15                  Drug Application

105                  Baseline B           Baseline B

110                  Preparation          Preparation, then
                                          Drug Infiltration

118                  Circumcision         Circumcision
                     (Gomco Clamp         (Gomco Clamp
                      Procedure)           Procedure)

123                  Postcircumision      Postcircumcision

6 h                  Methemoglobin        Methemoglobin

[Note: This figure has been rearranged for electronic presentation.]

Figure 1. - Progress through phases or trials of the study.



At the time of the study, most neonatal circumcisions were being performed without anesthetics or analgesics in the region in which this study was conducted. The population in the area is multicultural, and newborns enrolled in the study represented diverse ethnic backgrounds. The sample was drawn from among healthy male newborns with a gestational age of 37 weeks or more scheduled for circumcision in Edmonton, Alberta. The study took place from September 1994 to October 1996. University and hospital institutional review boards approved the study protocol. Parents who requested circumcision for their newborn sons were asked for informed consent to participate.

Exclusion criteria included (1) that the newborn was unable to remain in hospital for 6 hours after circumcision, (2) that the newborn was given drugs other than routine phytonadione and erythromycin eye ointment; and (3) that breast-feeding mothers received sulfonamides or salicylates, since both have been linked to methemoglobinemia and can be excreted in breast milk.

Outcome Measurements.

Physiologic measure of newborn distress included heart rate, respiratory rate, oxygen saturation, and palmar sweat.17-19 Behavioral measures of pain and newborns consisted of sleep-wake states and cry and motor responses, particularly facial expression.20-21 Heart rate, however, may be among the single most sensitive measure for acute distress among newborns and infants because it increases within seconds following a nociceptive stimulus. It was, therefore, identified as our primary outcome. In this study, continuous physiological data were collected by a Merlin monitor (Hewlett Packard Co, Andover, Md) and transmitted to a computer. A video recorder collected data on behavior among the newborns.

Specimens were obtained for metHB via a heel stick 6 hours after circumcision, usually at the same time that other blood was collected. Approximately 0.1 to 0.3 mL of blood was collected in a capillary tube and tested for metHB on the Co-oximeter 278 Blood Gas System (Chiron Inc, Markham, Ontario).


Topical Drugs. - Two-gram doses of EMLA and doses of placebo were prepared by the pharmacy and placed in 3-mL syringes. The placebo was an inactive cream that could not be distinguished from EMLA either by texture, color, or volume. The pharmacist placed the prepared syringe into an opaque envelope marked with a coded case number.

Correct administration of EMLA is known to be important for the drug to be effective22; however, no information was available at the start of the study about the application technique for circumcision, especially the correct waiting time. On the basis of earlier research about use of EMLA with children,22 we chose a 90-minute application time and the following procedure.

In preparation for drug application, the newborn's diaper was removed and his legs restrained. The drug was applied in a thick mound over the distal half of the penis by using the syringe as an applicator. The penis was then wrapped with a semipermeable dressing (Tegaderm). Afterward, the newborn was returned to his bassinet to await the circumcision. The dressing was checked every 10 minutes, and a new dose and dressing were applied if the dressing had leaked or the newborn urinated.

Parenteral Drugs.-The pharmacist placed empty syringes in the opaque envelope along with instructions about the block to be done when preparing drugs for the infiltration groups. This envelope was placed in another envelope and coded in a way identical to topical cases.

The drug was administered with a tuberculin syringe and a 1.2-cm, 27-gauge needle. For both dorsal penile nerve and ring blocks, 0.8 mL of 1% lidocaine without epinephrine was used. Newborns who had a dorsal penile nerve block had 0.4 mL of lidocaine infiltrated at the 2 dorsolateral injection sites, using the method of Kirya and Werthmann.9 Newborns who had a ring block had 0.8 mL of lidocaine without epinephrine infiltrated in a band or ring around the penis, halfway along the shaft of the penis, following the procedure Broadman et al16 used for children's postcircumcision analgesia.

Our pilot work indicated that newborns settle if allowed to rest 6 to 8 minutes after injection, thus allowing the distress associated with circumcision to be distinguished from other distress. Accordingly, newborns in this study were allowed to rest for 8 minutes at the conclusion of the application of the dorsal penile nerve or ring blocks.

Study Design

The newborn was taken to the procedure room 2 hours prior to the scheduled circumcision time. Data were collected in a series of trials: baseline A, drug administration, baseline B, preparation, circumcision, and postcircumcision. Video recording was continuous during all trials. Physiological monitoring occurred during all trials, with exceptions noted. Figure 1 summarizes the progress of the topical and infiltration groups through the trials or phases of the procedure.

Baseline A.-This trial was completed at least 110 minutes before the circumcision, prior to drug administration. Newborns were placed on the padded circumcision mold, and electrodes were attached. Heart rate was recorded for 5 minutes. Newborns in the 2 infiltration groups were returned to their bassinets at the end of the trial.

Drug Administration. - Newborns in the topical groups had their assigned drug applied by the assistant at the conclusion of the baseline A trial. Complete monitoring continued during application of the topical and for at least 4 minutes following the application. Afterward, the leads were disconnected, and the newborns were returned to their bassinets for a minimum of 90 minutes to await surgery.

Baseline B. - Approximately 10 minutes before the circumcision, physiological monitoring resumed. The newborn was moved to a padded, molded circumcision apparatus (Circumstraint, Olympic Medical, Seattle, Wash) and allowed to settle. Then a 2-minute recording was obtained from the quiet newborn.

Preparation and Drug Infiltration. - this trial involved preparation of the newborn for the surgical procedure. Newborns in the topical groups had the dressing and drug removed, and their legs but not their arms were restrained. They were again allowed to settle for 8 minutes before surgery began. Newborns in the 2 infiltration groups had their legs restrained, and a ring block or dorsal penile nerve block was administered by the physician, followed by an 8-minute rest. Once this phase or trial began, restraints remained on until the end of the circumcision, since our pilot had indicated that newborns become more distressed by the removal and reapplication of restraints than by having them remain on.

Circumcision. - An experienced urologist, obstetrician, or pediatrician performed the surgery. The Gomco clamp procedure was used for circumcision in all cases.

Postcircumcision. - At the end of the circumcision, the newborn was released from the restraints and monitored for at least 8 minutes. Then a gauze and petroleum jelly dressing were applied to the penis. The newborn was returned to his parents. Six hours after circumcision blood was collected from all newborns so that metHB levels could be compared across groups.

Study Blinds

The research assistant was unaware of group assignment until the drug application phase began. She remained blind to the type of topical (placebo or EMLA) but became aware of the general group: infiltration or topical. The physician became aware of some details of group assignment immediately before the circumcision. Like the assistant, he remained blind about the type of topical administered.

Since type of infiltration procedure could be detected by a cursory look at the injection site, we did not attempt to blind physicians or assistants. In order that the study blind not be broken, no sensory testing was done to determine the adequacy of the block.

Data Preparation

Computer records of continuous heart rate data and videos of newborn behavior were prepared for analysis by matching physiological data and corresponding video segments for trials using event markers inserted during data collection. Heart rate graphs were constructed for each newborn over all trials. The data were examined for completeness.

Videotapes were viewed independently by 2 assistants who were unaware of group assignments. They recorded the periods that the newborns spent crying. As agreement among assistants was required, discrepant ratings of periods of newborn crying were resolved by having the raters and a third evaluator re-rate the tape. Periods of newborn crying were used to compute the proportion of time newborns spent crying (for any trial, the proportion of crying equals time spent crying divided by total trial duration).

When assistants viewed videotapes, they documented events (including the surgical procedures) and time of occurrence. On the basis of these observations of events, 4 stages of the circumcision were identified. Codes were inserted in the data files to demarcate stages of circumcision for analysis of heart rate and crying. The stages consisted of cleansing (cleansing and draping); separation of foreskin from glans (probing, release of foreskin from glans, incision of foreskin); application of the clamp (bell on, clamp on, foreskin excised); and clamp removal (clamp off, bell off, examination, and wiping with swab).

Statistical Analyses

Since opinions vary about analysis of repeated measures, heart rate data were analyzed in several ways. We undertook a time-series analysis but could not identify a single model. Hence, we turned to analysis of variance (ANOVA) to examine heart rate. Data were handled with a number of ANOVA methods that provided consistent results. We present one approach in this article: repeated measures ANOVA of change in heart rate. Each measure of heart rate was subtracted from the baseline B mean heart rate, which produced the variable heart rate change B (ie, the rise or fall in heart rate). This approach was preferable to analysis of covariance (using baseline B as the covariate), since between-subject variance exceeded within-subject variance, and heart rate was a reliable measure.23 Repeated measures reported in this article have been referred to as trial, stage, or time.

Other variables were assessed with nonparametric analyses or ANOVA. Repeated measures ANOVA were used to assess proportion of cry. When appropriate, the Tukey honestly significant difference (HSD) was used for post hoc comparisons. For all statistical tests reported, including post hoc comparisons, the significance level was P=.05.

We estimated sample size from previous studies about neonatal circumcision,24 using a heart rate difference of 26 beats per minute and an SD of 26. We planned to recruit 27 newborns per group, based on alpha = .05 and desired power = 0.90. 25

                          Eligible Patients
                        Not Randomized (N=23)
           n=20 Newborns Were Unable to Remain in Hospital
        n=3 Parents Believed There Was No Need for Anesthetic
                          Randomized (N=54)
     |                   |                 |              |
     |                   |                 |              |
Placebo (n=12)   Eutectic Mixture of   Ring Block  Dorsal Penile
(1 Withdrawn)    Local Anesthetics     (n=13)          Nerve Block
                      (n=15)  (n=14)

Figure 2. - Profile of the study population.


Part way through the trial, physicians and members of the research team remarked that there were obvious differences in behavior of newborns in either infiltration group compared with those in both topical groups. This led us to reexamine our estimation of effect size as well as the ethical matters related to sample size.26 We, therefore, ended enrollment with the following sample sizes: ring block, n=12; dorsal penile nerve block, n=14; EMLA, n=15; and placebo, n=11. The total number of completed cases was 52. Our projected sample estimated power of 0.90 and a smaller effect than was observed. Our computed power for the sample data exceeded 0.95 for main research questions.

We enrolled 54 newborns; 2 were withdrawn by parents soon after data collection commenced-one because the parents were unable to remain in hospital until completion of the study and the other so that the newborn could receive phototherapy. Parents of another 23 newborns were approached to participate but declined; their primary reason was that they were unable to remain in hospital for the duration of the study (Figure 2).

Characteristics of Newborns

Birth weight of 50 of the newborns ranged from 2500 to 4805 g. The other 2 newborns were fraternal twins of 38 weeks' gestation, and they weighed 2230 and 2305 g. Of the newborns, 75% were delivered vaginally and 25% by cesarean section. Newborns in this study were 1 to 3 days of age when circumcised (92% were 2 days of age or younger). The groups were not significantly different for the following characteristics: weight, mean, 3594 g; newborn age, mean, 1.5 days; and 5-minute Apgar scores, mean, 9. Feedings were not withheld prior to circumcision, and time of last feeding was not significantly different for the groups (mean, 116 minutes).

Effect of Physicians and Assistants

The 4 treatment groups were compared to determine if physicians and assistants had any differential effects on groups, since they could not be blinded to type of infiltration vs the topical. There were no significant differences among physicians (n=4) or assistants (n=3) in the time taken to complete each stage of circumcision, as well as the heart rate of the newborns and their crying during these stages (ANOVA).

Treatment Effects

The main outcome measures used to assess treatment efficacy were proportion of time spent crying and heart rate, and the key intervals of time were the circumcision and postcircumcision periods. Differences among groups during the various trials or stages of circumcision were of primary interest (stage or trial by treatment interactions), as were overall group differences (main treatment effect). Table 1 presents the groups' mean proportion of time crying and heart rate.

Hypothesis 1.- We tested the first hypothesis using 2-way ANOVA (group and stage). Heart rate change B in the placebo group was significantly greater than in any other group at all stages of circumcision (Tukey HSD), thus confirming hypothesis 1.

Tukey post hoc analysis also indicated that the 3 anesthetics led to significantly different outcomes during 2 stages of the circumcision. During foreskin separation, all groups were significantly different from one another. After the placebo group, heart rate change B was greatest for the EMLA group, followed by dorsal penile nerve block and ring block. During the clamp-off stage, the EMLA group's heart rate change B was smallest, and the placebo group's was the greatest. The 3 anesthetic groups did not differ significantly during the cleansing or clamp-on stages. Figure 3 illustrates mean heart rate change for the 4 groups during the stages of circumcision.

We assessed group differences in heart rate change B during the immediate postcircumcision trial using 2-way ANOVA (group and time). We compared the groups at four 30-second intervals immediately following the last stage of circumcision. The placebo group's heart rate change B was greatest compared with the other groups (and this difference occurred at each time interval). These results also confirm hypothesis 1.

Two-way ANOVA was used to assess cry proportions for the 4 groups during the 4 stages of circumcision and during the four 30-second intervals immediately after surgery. The placebo group cried a significantly greater overall proportion of time than all other groups, an average of 85% of the time during circumcision and 63% of the first 2 minutes following circumcision (a significant group effect). The groups did not differ significantly at any trial or interval (interaction of group and time).

Hypothesis 2.- The groups were also compared to assess distress caused by the drug administration procedure. The topical groups were combined for comparison with ring block and dorsal penile nerve block groups. The proportion of time that newborns cried during their drug administration trial was not significantly different for the groups (1-way ANOVA). Heart rate change B increased for all groups during drug administration, but the increase was greatest for the ring block, followed by dorsal penile nerve block. Within 15 seconds following the injections, the heart rates of newborns in the ring block and dorsal penile nerve block groups were not significantly different (t test). On average, newborns in the infiltration groups stopped crying 92 seconds after the block was administered, whereas newborns in the topical groups stopped crying 63 seconds after drug application.

Hypothesis 3. - A common belief among clinicians and other health care workers is that the injection of a block is as painful or more painful than circumcision without anesthetic. We tested hypothesis 3: that performing a circumcision without anesthetic causes more distress than providing dorsal penile nerve or ring blocks (1-way ANOVA). To conduct this analysis, it was necessary to compare different groups at 2 different points in time (infiltration groups during the block, compared with placebo group during circumcision).

Heart rate change B data for the dorsal penile nerve and ring block groups during drug administration were compared with heart rate change B data for the placebo group during circumcision. The placebo group's heart rate change B during circumcision was significantly greater than the heart rate change B of the infiltration groups during drug administration. The average length of time taken to perform a ring block was 48 seconds, whereas dorsal penile nerve block was completed on average in 25 seconds. These times can be compared with the average time taken to perform circumcision on the placebo group (3.5 minutes). Thus, heart rate change was greater and lasted longer with an unanesthetized circumcision compared with the effect of and time for infiltration of a block.

Adverse Outcomes

Analysis of metllB levels was conducted with nonparametric statistics. Of metHB levels in the EMLA group, 60% were above the median for the entire sample, which was significantly higher than in the other groups (using the Kniskal-Wallis ANOVA test). Two newborns in the EMLA group bad margin-ally elevated metHB levels that did not require treatment (2.4% and 4.5%). The mean metHB levels for the placebo, EMLA, dorsal penile nerve block, and ring block groups were 0.6%, 1.3%, 0.7%, and 0.4%, respectively.

Injection sites were observed by the assistants 4 hours after circumcision. They recorded the occurrence of minor bruising at injection sites for half of the newborns in both infiltration groups. No newborn required treatment.

A serious postsurgery incident was observed in 1 newborn in the placebo group. There was nothing remarkable about this newborn's history prior to the circumcision. His 1- and 5-minute Apgar scores were 9 and 10. He was last fed 3 hours before circumcision. During and following circumcision, the newborn reacted much the same as others who received a placebo (continuously elevated heart rate and high-pitched cry). About 2.5 minutes after the conclusion of surgery, the new-born had an episode that included abnormal posture (lack of tone in limbs), several periods of apnea (one lasting more than 25 seconds), and projectile vomiting. No physiological data were recorded during the episode, since the computer had lost contact with the monitor immediately prior to its onset. The newborn recovered following the episode.

Another newborn in the placebo group had a choking episode with apnea after surgery. This began about 3.5 minutes after the circumcision and lasted less than 30 seconds.

Table 1. - Outcomes Measures Used to Assess Treatment Efficacy

Outcome Measures     Placebo     EMLA*        DPNB  Ring Block
------------------------------------------------------------ ---------
Trial or stage
duration, mean (SD)

Drug application     69.8(17.0)  70.4(13.5)   24.6(6.8)  48.2(16.5)
Cleansing            19.7(10.8)  28.9(22.3)   24.4(12.1)  23.4(16.4)
Separation           52.5(7.6)   65.6(20.8)   57.8(11.0)  54.7(17.0)
Clamp on            125.6(74.1) 131.5(78.1)  137.1(72.2) 124.7(75.9)
Clamp off            16.5(5.0)   21.9(17.9)   21.4(6.7)  20.0(7.2)
------------------------------------------------------------ ---------
Heart rate change B
(from baseline B,
mean (SD)

Drug application    19.0(29.7)  19.0(13.5)    30.4(23.3)  35.8(28.6)
Cleansing           22.0(21.9)  13.0(19.2)    17.4(23.6)  20.6(33.6)
Separation          53.0(46.2)  41.4(31.4)    30.9(32.6)  20.2(32.1)
Clamp on            43.2(41.2)  16.6(31.5)    19.0(31.8)  18.1(26.3)
Clamp off           43.9(35.4)   3.8(23.7)    15.7(31.3)  14.0(26.0)
30 s after          38.1(37.8)  11.9(29.3)    12.6(26.1)  22.1(27.2)
60 s after          36.1(41.4)   2.7(27.2)     4.5(17.3)  20.5(27.9)
90 s after          32.3(38.8)   3.0(27.4)    -1.2(15.4)  9.5(26.4)
120 s after         19.1(28.5)   6.5(24.4)    -1.2(14.1)  12.3(27.0)
------------------------------------------------------------ ---------
Proportion of time crying
mean (SD)
Drug Application    0.63(0.36) 0.60(0.34)    0.83(0.23)  0.82(0.21)
Cleansing           0.78(0.29) 0.55(0.34)    0.58(0.38)  0.52(0.43)
Separation          0.98(0.05) 0.82(0.17)    0.65(0.33)  0.54(0.38)
Clamp on            0.85(0.21) 0.55(0.32)    0.41(0.35)  0.37(0.24)
Clamp off           0.81(0.25) 0.35(0.37)    0.36(0.42)  0.32(0.40)
30 s after          0.56(0.34) 0.41(0.42)    0.22(0.30)  0.51(0.34)
60 s after          0.66(0.41) 0.25(0.38)    0.15(0.29)  0.37(0.38)
90 s after          0.62(0.39) 0.35(0.44)    0.13(0.31)  0.34(0.44)
120 s after         0.67(0.45) 0.31(0.41)    0.10(0.27)  0.38(0.46)

*EMLA indicates eutectic mixture of local anesthetics.

Heart               X placebo, * EMLA; O DPNB; @ Ring Block





40                                X              X

30                 O

20      @          @

        O                         O              O
        *                         @              @
10                                               *
        |          |              |              |
     Cleanse   Separation   Bell and Clamp   Clamp Off

Fig. 3 - Heart rate change from baseline B through the stages of the
[Note:  This electronic presentation is only an approximate
represention of the original graph.]


Several beliefs explain the practice of performing circumcision without anesthesia.7 One is that newborns experience little or no pain from circumcision. Without exception, newborns in this study who did not receive an anesthetic suffered great distress during and following the circumcision, and they were exposed to unnecessary risk (from choking or apnea). Therefore, we have concluded that circumcision should be performed with anesthetic.

It is also not true that the injection from either ring block or dorsal penile nerve block causes as much pain as the circumcision. Circumcision without anesthesia causes greater distress than the pain of the infiltration, and this distress occurs for a longer period, and recovery takes longer.

A third view is that superior surgical technique eliminates newborn distress from circumcision, thus removing the need for anesthetic. Although our physicians were highly experienced in performing circumcision and had excellent surgical technique, every newborn in the placebo group exhibited extreme distress during and following circumcision.

Another justification for not providing anesthetic is the occurrence of infantile amnesia. It is true that as adults these newborns will not be able to retrieve the memory of their surgery and distress. This fact, however, cannot justify the practice of performing surgery without anesthetic.

Of the 3 anesthetics considered in this investigation, ring block is clearly superior. It provides satisfactory anesthesia for all stages of the circumcision. While newborns fussed periodically over the restraints or being handled, they typically did not react to the most nociceptive elements of the circumcision (such as foreskin separation and incision). On average, their heart rate rose about 20 beats per minute from baseline as preparation for the surgery began with placement of the drapes.

While the ring block takes about twice as much time to administer as the dorsal penile nerve block, newborns recover from both at the same rate. The ring block is easily accomplished, and the risks appear to be minimal.16

Dorsal penile nerve block is the second choice. Physiological and behavioral data confirmed that the dorsal penile nerve block was not fully effective during the foreskin separation stage of the circumcision. Newborns cried and their heart rates rose at the start of this stage of circumcision. The clinical experience of one of us with adults (J.M.) and that of others27 supports the findings of this study. Adult males have complained about incomplete anesthesia with dorsal penile nerve block.

As to ease of administration, the dorsal penile nerve block may be technically more difficult to achieve than the ring block. In assessing safety, a low rate of complications has been found for dorsal penile nerve block, including a 1.2% rate of bruising or hematoma based on retrospective chart review,12 and 11% based on survey.13 These rates may have been higher had examinations been made and recorded for all cases. Although assistants in our study recorded that minor bruising occurred in half of the newborns who had ring block or dorsal penile nerve block, their observations were not sufficiently detailed to evaluate extent of bruising. None of the newborns required treatment for bruising.

Eutectic mixture of local anesthetics was also not fully effective during the stage of surgery where the foreskin is separated from the glans and incised. Eutectic mixture of local anesthetics was less effective than the dorsal penile nerve block judging by the newborns' responses, which were more intense and prolonged. All newborns in the EMLA group became rapidly distressed at the outset of the foreskin separation stage. These newborns typically settled by the clamp-off stage of circumcision. Overall, these outcomes suggest that EMLA does not penetrate to the depth required for circumcision and is not a prudent choice.

The technique for administering EMLA poses a number of problems. Considerable technical skill is needed to apply the adhesive, semipermeable dressing that covers the topical. Regular monitoring is required to make sure that the drug has not shifted off the target area. The amount of time required between drug application and circumcision may not be practical in many nurseries.

Some clinicians have suggested that EMLA may be more effective if placed both over and under the foreskin. That approach would likely cause the newborn distress, since insertion of a small probe during circumcision causes distress for newborns who do not have a block. The theoretical risk of methemoglobinemia remains, although it was not observed in this study.

Lack of sensory testing to determine adequacy of infiltration block, lack of systematic and detailed examination of injection sites, and length of postoperative follow-up of newborns are limitations of our study. It would have been useful to follow newborns for several days to assess postoperative discomfort, but this was not possible with the current practice of short-term hospital stays following delivery. Sensory testing was not done to ensure the consistency of the handling of newborns. Inadequate performance of blocks seems unlikely, as there were no differences in outcomes across physicians.

The results of this study suggest that an anesthetic should be administered to newborns prior to undergoing circumcision. The ring block was the most effective.

This research was supported by grant 66O9-1824-60A from the National Health Research Development Programs, Health Canada Ottawa, Ontario.


  1. Wiswell T, Roscelli J. Corroborative evidence for the decreased incidence of urinary tract infections in circumcised male infants. Pediatrics 1986; 78:96-99.
  2. Moses S, Bradley J, Nagelkerke N, Ronald A, Ndinya-Achola J, Plummer F. Geographical patterns of male circumcision practices in Africa: association with HIV seroprevalence. Int J Epidemiol 1990;19:693-697.
  3. Maden C, Sherman K, Beckman A, et al. History of circumcision, medical conditions, and sexual activity and risk of penile cancer. J Natl Cancer Inst 1993;85:19-24.
  4. Poland R. The question of routine neonatal circumcision. N Engl J Med. 1990; 322:1312-1315.
  5. Wallerstein E. Circumcision, the uniquely American medical enigma. Urol Clin North Am 1985;12:123-132.
  6. Rabinowitz R, Hulbert W. Newborn circumcision should not be performed at birth without anesthesia. Birth 1995;22:45-46.
  7. Wellington, N, Rieder M. Attitudes and practices regarding analgesia for newborn circumcision. Pediatrics 1993;92:541-543.
  8. Toffler, W. Sinclair A, White K. Dorsal penile nerve block during newborn circumcision: underutilization of a proven technique? J Am Board Fam Prac 1990;3:171-174.
  9. Kirya, C, Werthman M. Neonatal circumcision and penile dorsal nerve block: a painless procedure. J Pediatr. 1978;92:998-1000.
  10. Holve RL, Bromberger PJ, Groveman HD. Klauber MR, Dixon SD, Snyder JB. Regional anesthesia during newborn circumcision. Clin Pediatr. 1983; 22:813-818.
  11. Stang H, Gunnar M, Snellman L, Condon L, Kestenbaum R. Local anesthesia for neonatal circumcision: effects on distress and cortisol response. JAMA 1988;259:1507-1511. [Abstract]
  12. Fontaine P, Dittberner, D, Scheltema K. The safety of dorsal penile nerve block for neonatal circumcision. J Fam Pract. 1994; 39:243-248.
  13. Snellman L, Stang H. Prospective evaluation of complications of dorsal penile nerve block for neonatal circumcision. Pediatrics. 1995;95:705-708.
  14. Benini F, Johnston C, Faucher D, Aranda J. Topical anesthesia during circumcision in newborn infants. JAMA. 1993, 270:850-853.
  15. Taddio A, Stevens B. Craig K, et. al. Efficacy and safety of lidocaine-prilocaine cream for pain during circumcision. N Engl J Med. 1997;336:1197-1201.
  16. Broadman L, Hannallah R, Belman B, Elder P, Ruttimann U, Epstein B. Post-circumcision analgesia: a prospective evaluation of subcutaneous ring block of the penis. Anesthesiology. 1987;67:1197-1201.
  17. Harper v. Rutter N. Development of emotional sweating in the newborn infant. Arch Dis Child 1982;57:691-695. [Abstract]
  18. Owens M, Todt E. Pain in infancy: neonatal reaction to heel lance. Pain. 1984;20:77-86.
  19. Rawlings D, Miller P, Engel R. The effect of circumcision on transcutaneous PO2 in term infants. AJDC. 1980; 134:676-678.
  20. Grunau R, Johnson C, Craig K. Neonatal facial and cry responses to invasive and non-invasive procedures. Pain. 1990; 42:295-305.
  21. Gunnar M, Malone S, Vance, G, Fisch R. Coping with aversive stimulation in the neonatal period: quiet sleep and plasma cortisol levels during recovery from circumcision. Child Dev. 1985; 56:824-834.
  22. Lander J. Hodgins M, Nazarali S, McTavish J. Ouellette J, Frieson E. Determinants of success and failure of EMLA. Pain. 1996; 64:89-97.
  23. Streiner D, Health Measurement Scales. 2nd ed. Oxford, England: Oxford University Press; 1995.
  24. Williamson P, Williamson M. Physiologic stress reduction by a local anesthetic during newborn circumcision. Pediatrics. 1983;71:36-40.
  25. Neter, J. Wasserman W, Kutner M. Applied Linear Regression Models. 3rd ed. Homewood, Ill: Irwin Inc; 1983.
  26. Freedman B. Equipoise and the ethics of clinical research. N Engl J Med. 1987; 317:141-145.
  27. Stav A, Gur L, Gorelik U, Ovadia L, Isaakovich B, Sternberg A. Modification of the penile block. World J Urol. 1995;19:251-253.
  28. Schoen E, Fischell A. Pain in neonatal circumcision. Clin Pediatr. 1991;30:429-432.

From the Faculties of Medicine and Nursing, University of Alberta (Drs. Lander, Metcalfe and Muttit), and the Division of Research Administration and Pharmacy, University of Alberta Hospitals (Ms. Brady-Freyer and Ms. Nazarili), Edmonton, Alberta
Reprints: Janice Lander, PhD, Clinical Sciences, University of Alberta, Edmonton, Alberta, Canada T6G 2G3 (e-mail:

(File revised 16 October 2006)

Return to CIRP library