The Length of Gomco Clamp Timing and Its Effect on Bleeding

For this institutional review board–exempt, prospective study (HUM00107811), credentialed attending physicians from the Division of Pediatric Hospital Medicine were assigned to one of the following 2 groups for the duration of the study: those who would perform the procedure with a clamp time of at least 5 minutes (group 1) and those who would perform the procedure with no minimum clamp time (group 2). No minimum clamp time was chosen versus a set time to allow providers to remove the clamp as soon as they removed the patient’s foreskin, because the time required to remove foreskin may vary on the basis of provider. Physicians were split into 1 of the 2 groups on the basis of their level of experience to allow for providers of similar years’ experience be represented equally in both groups. In addition, 2 providers (M, S) were placed in group 1 on the basis of their personal preference. Any newborn male circumcised by a member of the Division of Pediatric Hospital Medicine was included in the study. These included patients on the pediatric newborn service or general pediatric team (transferred from the NICU or pediatric newborn service) who met the criteria for an elective circumcision: no anatomic contraindication, received intramuscular vitamin K, at least 10 hours old, normal vital signs, informed consent obtained from a parent for the procedure, and had no known bleeding disorders. The pediatric newborn service includes newborns of at least 35 week’s gestation. Circumcisions performed on patients on the newborn service typically are performed during the first 2 days of life whereas circumcisions performed on patients on the general pediatric team are typically performed a few days before discharge. When a circumcision procedure note was completed, a notification was sent to the study team via the electronic medical record. Study dates were September 28, 2015, through October 9, 2016.

Providers in both groups 1 and 2 followed the same procedural steps for the Gomco circumcision, including placing the bell and base plate in opposition, tightening the screw nut (Supplemental Video 1; Supplemental Fig 2) and removing the foreskin with a scalpel in a circumferential manner at the level of the base plate. After removing the foreskin, providers in group 1 waited for the remaining 5 minutes before unscrewing the nut and removing the bell from the base plate. Providers in group 2 unscrewed the nut and removed the bell from the base plate after excising the foreskin; a minimum clamp time was not required. Clamp time was defined as the time from when the screw nut was securely tightened until it was unscrewed. Thus, the clamp time included the time it takes to remove the foreskin. The providers in group 2 would document the clamp time in minutes and seconds. At the completion of the procedure, physicians in both groups wrote their procedure note, including whether there was any bleeding or other complications occurring during or immediately after the procedure. Per hospital protocol, nursing staff examined the circumcision site at intervals of 15 and 30 minutes’ postprocedure and documented any bleeding in the nursing notes section of the medical record. If more than minimal bleeding is noted, pressure is typically applied first and microfibrillar collagen hemostat (MCH) second if there is continued bleeding. Analgesia was provided by using 1% lidocaine without epinephrine (given as a nerve block), oral sucrose (pre- and intraprocedure), and acetaminophen (as needed postprocedure). Petroleum jelly was applied to the penis after the procedure to keep the diaper from sticking to the wound¹⁰  and for adhesion prevention.

Data points pertinent to this outcome study were agreed on by study members and included date of birth, date of circumcision, length of clamp time for the procedure (minutes and seconds), comments made by physician or nursing staff, whether pressure and/or a MCH was applied, and whether bleeding was severe enough to warrant operative hemostasis. MCH application need was determined by the nurse and/or physician and is typically based on the amount of bleeding. Weight, gestational age, Apgar scores, and demographic data were not included because researchers in a previous study did not find those data points to be relevant to bleeding rates in circumcisions performed by the Gomco method.¹¹  Data points were extracted from the circumcision procedure note, the nursing newborn flow sheets, and individual physician and nursing notes in the electronic medical record for both groups. Nursing and relevant physician notes were reviewed from time of circumcision until discharge to include any delayed bleeding events.

This study was intended to be a randomized control trial; however, 2 physicians preferred to be in group 1, which was accommodated. Because of this, we followed an observational design in which the predictor was clamp timing and outcome was pressure and/or MCH application. Analysis was adjusted for physician’s years of experience.

The sample size calculations were based on the number of physicians needed per each clamp time group. A random sample of 10 physicians per each group with an average number of patients of 55 per physicians for a total 20 physicians and 1100 patients will provide 80% power to detect a difference of 5% between 2 groups with a type I error α = 0.05. Our final sample of 10 physicians and 605 patients in the no minimum clamp time group and 13 physicians and 647 patients in the 5-minute clamp time group was higher.

Statistical analysis was done by using R version 4.0.3.¹²  Generalized linear mixed effect models (glmer in lme4 package) with logit link were used to model the odds of having an adverse event for each of the outcomes (pressure applied and MCH applied). We used a random intercept at the level of the provider to account for clustering of patients within physician and a fixed effect for clamp time group. Years of experience of physicians was included in the model to account for the range of experience among physicians across both groups. We ran 2 models for each outcome: (1) the primary model which includes all physicians and (2) a secondary (sensitivity) model, which excludes 2 physicians (1 for each clamp time group) who both had a much higher rate of adverse outcome compared with others and zero years of experience. Significance was set at P < .05.

The final analysis sample included 23 physicians and 1252 patients: 13 physicians and 647 patients in group 1 and 10 physicians and 605 patients in group 2 for the primary analysis. Nine patients were removed. See Fig 1 for study flow diagram. Thirty-eight patients had documented clamp times of 5 minutes (the default template note time) but had circumcisions performed by providers in group 2. These cases likely all had clamp times of less than 5 minutes, with the provider forgetting to modify the default time in their note. These cases remained in their original group for analysis but were excluded from the clamp time length calculations because the error in time recording. There was 1 case in which a group 1 provider did not report a clamp time. This case remained in group 1.

The majority of circumcisions were performed on day of life (DOL) 0 to 2 (n = 1229), with the remaining occurring on DOL 3 to 46 (n = 23). The clamp time range in group 2 was 1 minute to 4 minutes and 30 seconds, with a mean of 2 minutes and 5 seconds and a median of 2 minutes. See Table 1 for individual provider clamp times, years’ experience, pediatric hospital medicine full-time equivalent before the study start date, and bleeding rates requiring pressure and/or MCH application.

The percentage of circumcisions with MCH application because of bleeding were similar between the 2 groups. The estimated percentages from the generalized mixed effects model, adjusted for provider’s years of experience, were 5.1% (95% confidence interval [CI]: 3.1–8.1) for group 1 and 5.6% (95% CI: 3.5–8.8) for group 2. The percentage of circumcisions with pressure applied because of bleeding was much lower than MCH application. Circumcisions that had pressure applied were estimated to be 0.6% (95% CI: 0.2–1.7) for group 1 and 1.3% (95% CI 0.5–3.1) for group 2. The difference between groups, in both categories, was not statistically significant (P = .794 and 0.215, respectively) (Table 2).

In a subanalysis of the DOL 3 to 46 group, there were no cases of pressure application and only 1 case of MCH application, which was in group 1. There were no cases that had both MCH and pressure application documented in this group. We ran a generalized linear mixed model to examine the differences of DOL <3 vs DOL 3 to 46 groups, adjusting for provider experience. The estimated MCH application rate for DOL 3 to 46 was 3.87% (95% CI: 0.5–23.3) and for DOL <3 was 5.34% (95% CI: 3.8–7.5). The difference between the DOL <3 and DOL 3 to 46 groups was not statistically significant (P = .743). Operative hemostasis was not required for patients in any of the above groups.

There were 2 outlying providers who had much higher rates of bleeding compared with the others. These 2 providers (A, K) performed 56 and 34 circumcisions, respectively, and were both in their first year of performing circumcisions. A sensitivity analysis after removing these providers yielded estimated bleeding rates for each group as follows: 0.9% (95% CI: 0.4–2.1) in group 1 and 0.9% (95% CI: 0.4–2.1) in group 2 for pressure application and 4.3% (95% CI: 2.7–6.8) in group 1 and 5.0% (95% CI: 3.1–7.9) in group 2 for MCH application. The difference between these groups was not statistically significant (Table 2).

This prospective study evaluating clamp time length and its effect on postprocedural bleeding did not reveal a statistically significant difference in bleeding outcomes measured by pressure or MCH application. In addition, we found bleeding rates, regardless of clamp length, within the range to those in the literature, 0.6% to 9.6%.^11,13–16  This study is important for several reasons. First, it is well understood that standardizing procedures decreases errors and improves patient outcomes,¹⁷  and there is no consensus in the literature regarding appropriate clamp time recommendations for circumcisions performed with the Gomco clamp. As previously mentioned, the initial article suggesting a 5-minute clamp time gives no evidence for such a recommendation. Subsequent reports, including the Academy of Pediatrics 2012 Technical Report, have carried on the recommendation without citing any scientific evidence. In addition, several other important sources do not include a recommended clamp time.^3,18–21  With this study, we help provide new information about this common procedure that may potentially save time and improve newborn care.

It is not surprising that some sources do not comment on clamp times because, anecdotally, providers have been unscrewing the Gomco clamp immediately after the foreskin is removed for years without bleeding complications. The anecdotal evidence is understandable because the average adult can exert 40 to 100 lb of force turning their dominant wrist in a pronation-supination action to drive the screw nut.²²  The Gomco clamp converts this force into 8000 to 20 000 lb of hemostatic force against the prepuce.²³  This amount of force delivers a nearly immediate crush injury. It should be mentioned that there are likely provider differences in how tightly the nut is turned, noting that “tightening it well” is the recommendation.¹⁰  At our institution, providers are taught to turn the nut as tightly as they can and to use gauze (placed over the nut) or pliers to unscrew the nut if needed.

Because it has been demonstrated that standardization of surgical procedures reduces the risk of poor outcomes,¹⁷  the need for appropriate clamp times to be known and available to providers is essential. Additionally, it has been a quality initiative of hospitals to incorporate lean processes with safe, efficient patient flow.²⁴  This process incorporates the standardization of procedures and has since been adopted by all members of our Division of Pediatric Hospital Medicine. With an average clamp time of 2 minutes and 5 seconds in group 2, if a provider is performing 8 circumcisions, the provider saves almost 25 minutes. This “extra” time can be spent providing other patient care responsibilities or teaching residents and students and allows more time for those assisting in the procedure (ie technicians or nurses) to do other tasks as well.

In addition, less time restrained on the circumcision board may lead to a decrease in patient discomfort and newborn stress because the time spent immobilized will be decreased. Newborns may experience less fussiness and sucrose administration. There is also opportunity for a decrease in cough frequency or a respiratory event. Because normal newborns may experience cough or emesis in the postpartum period, owing to the introduction of feedings and transient lower esophageal sphincter relaxation,²⁵  there is potential for a decrease in these events. Some providers recommend a nothing per os time period before circumcision to reduce the risk of aspiration or regurgitation.^5,26  On occasion, these episodes require suctioning and/or tilting of the patient on the circumcision board to protect the airway. This is concerning for several reasons: positioning on the circumcision board does not always allow for the newborns head to be raised appropriately, and moving the patient risks a breach in the sterile field, can be a distraction to the provider, and interrupts the procedure. A shorter procedure increases the likelihood that providers may experience this interruption less frequently.

Lastly, although this was not an aim of our study, we found that circumcision experience and bleeding rates were significantly higher in 2 providers (A, K) who had no experience with circumcision’s before our study. This is consistent with the observation of Feinberg et al that current operator experience is a more important factor than long-term experience or patient-related variables. Similar findings linking experience to outcomes have been found in surgical studies.^27–31  However, it is possible that newer providers may have been more likely to apply pressure and/or MCH. It is also possible that nursing may have been biased toward applying pressure and/or MCH to the newer provider’s patients.

This study had some limitations. First, each physician has his or her own skill level and experience when performing circumcisions with the Gomco clamp. However, all providers are credentialed attending physicians with similar training. In addition, 2 physicians were placed in group 1 because of personal preference. However, varied levels of experience were evenly spread between the 2 groups (Table 1). Second, pressure and/or MCH application depended on nurse and/or physician judgment and, for that reason, may represent more a measure of cautiousness than actual need for therapy. However, the same variation in nursing staff was equally present for both groups. In addition, for patients who had MCH application but pressure application was not noted, it is plausible that pressure application did occur but may not have been documented. As mentioned above, there were some challenges with documentation of circumcision clamp times. Lastly, circumcision bleeding complications are uncommon, particularly those requiring MCH application. By documenting a large number of circumcisions per physician, we were able to obtain precise estimates of the rates of bleeding complications for each physician in this study. However, a larger sample of physicians would be needed to obtain more precise estimates of the rates of bleeding complications overall and the physician-level factors that are associated with these uncommon events.

In this prospective study, we evaluated clamp time duration and bleeding outcomes for neonatal circumcisions. Clamp time was not associated with bleeding risk. However, provider experience revealed that novice providers had more bleeding.

University of Michigan Department of Pediatrics Charles Woodson Clinical Research Fund, John Schmidt MD, David Stewart MD, Brittany Allen MD, Nicholas Helmstetter MD, Timothy Capecchi MD.