Discussion
The technique to treat pediatric UH, called adhesive strapping, compression, or binding, has regained popularity in Japan. There are many minor differences in this method (table 2), but the concept is the same: compressing the UH and maintaining it inward. As an example, we provide photos demonstrating how to conduct it at our site in online supplemental file 2.
To our knowledge, our meta-analysis is the first reported study to examine the evidence from cohort studies on the efficacy of AS on UH. Our meta-analysis analyzed information on a total of 684 patients with AS and 464 patients without AS. There was no significant difference in the efficacy of AS on the overall UH closure and that of large hernia compared with observation-only management. Furthermore, the subgroup analysis of recent studies did not show a significant difference.
We believe that the factors that affected our results were the sample sizes and the proportion of large hernias in the included studies. Regarding sample sizes, 80% of patients with UH are expected to heal spontaneously before attaining 1 year of age and 90% before 2 years of age. Therefore, we estimated the closure rates by AS and observation-only management as 95% and 90%, respectively. The necessary sample size is calculated as 475 cases each with α=0.05 and 1−β=0.80. We estimated these sample sizes as 85% and 80%, respectively, and the sample size needed rose to 946 each. The sample sizes of the included studies in our study were mostly less than 100 in each arm. Therefore, we could not exclude the possibility that insufficient sample sizes affected the results.
The proportion of large hernias in each arm should also be considered when interpreting the results. Some studies indicated that the size of the UH did not influence its closure.18 22–24 For instance, Heifetz et al23 reported that 72 out of 78 patients (92%) with UH >0.5 cm were cured spontaneously within 4 years. Meier et al24 revealed that spontaneous closure of UHs >10 mm in diameter with at least 5 mm protrusion seemed to occur by the age of 14. However, Walker25 indicated that the diameter of the internal fascial ring was the most relevant parameter to determine whether spontaneous UH closure would be accomplished and that the healing rate differed according to the UH size. In our study, the proportion of large hernias in the AS group was higher in four studies5 9 10 20 and was not mentioned in five studies. One study reported that the average size of UH in the AS group was smaller than that in the observation group.17 Although we did not show an advantage of AS for large hernias, there would be several variabilities in this subgroup analysis, such as insufficient patient numbers, differences in the proportion between the two groups, and the definition of “large hernias.” From these points of view, randomized controlled trials with strict definitions of large hernias measured by ultrasound are warranted to reach a conclusion.
Although our study did not show the significance of the efficacy of AS on the closure rate, it indicated some possible advantages through secondary outcomes. One of them was the efficacy of earlier UH closure.
As Cresson and Pilling26 described, there was already some agreement in the 1950s that AS has no advantage after the age of 6 months. While Cresson and Pilling also noted from their experience that AS rarely hastened UH healing,26 our result corresponded to a previous study that revealed that the treatment duration was shorter in the AS group than in the observation group.27 Once AS is applied at an earlier age, it accelerates the healing of UH and the duration of UH existence is expected to be shortened. Consequently, it reduces the burden and anxiety of parents. Therefore, it is rational that AS was mostly applied within 6 months in recent studies to close UH earlier.2 14 16 17 20 21 The earliest time of application was less than 1 month of age, 18 days old.20 However, no study has revealed how early AS could be safely applied and how effective AS would be when applied earlier. Regarding premature infants, it was also uncertain whether the corrected age should be adopted rather than the real age. Factors such as umbilical condition, gestational age, and body weight should likely be considered for AS application.
Another possible benefit of AS to babies with UH is that AS prevents umbilical protrusion with redundant skin. In most studies, UH closure was defined as the closure of the orifice regardless of the umbilical figure. Therefore, if an umbilical defect is closed spontaneously, patients with protruding umbilicus will undergo umbilicoplasty for cosmetic reasons. According to the included studies, the prevalence of the protruding umbilicus at the end of the study was 0%–9.4% and 10%–18.8% in the AS and observation groups, respectively. This result corresponded to a study that gathered data from multiple institutions through questionnaires; it reported that the prevalence rates of redundant skin among patients who were regarded as cured from UH were 107 of 908 (11.8%) in the AS group and 33 of 146 (22.8%) in the observation-only group.28 Regarding umbilicoplasty, Hayashida et al21 reported that the operation time of patients who tried AS was shorter than that of patients without AS. The authors also mentioned that the difficulty of umbilicoplasty depended on the existence of redundant skin. Therefore, if AS prevents redundant skin and disfigurement of the umbilicus, it may be beneficial to patients.
Prematurity is regarded as a factor for UH because there is a difference in UH prevalence between mature and premature infants.3 12 We did not find articles that compared the UH closure rates by maturity, but Kurobe et al16 reported that the closure rates among mature and premature infants who were treated with AS at the end of the study were 85.7% and 80%, respectively. Although they seemed different, it was unclear whether closure rates would change if the corrected age was used.
AS was implemented mainly in the mid-20th century and after the year 2000. Particularly in Japan, AS was reconsidered as a practical treatment for UH because many retrospective studies indicated its positive effects on UH closure. The discontinuation rate of AS was reported to be approximately 10%, and skin irritation due to plasters and tape is one of the determinants in deciding whether to continue AS. Since more skin-friendly plasters and tapes have been developed, we expected that the AS completion rate would rise and increase the successful UH closure rate with a lower complication rate. In our study, however, the comparison of the AS group with the observation group on UH closure among recent studies, all published in Japan, did not show a significant difference.
The most severe complications of UH are strangulation, incarceration, and evisceration, but their occurrence rates are considered low.13 29 Therefore, based on the prevention of these complications, AS has few advantages for patients. If AS is applied for earlier closure or the prevention of redundant skin, AS-related complications should be taken into consideration. The most common problem was skin irritation or dermatitis from a plaster, which was the most decisive factor regarding AS continuation. However, these skin problems were relatively prevented using skin-friendly plasters and films or by introducing a short refraining period. Other complications reported in Japanese literature were massive bleeding from the umbilical artery, strangulation due to compression, and UH perforation caused by skin ulceration, which were all reported as case reports.30 One study reported that AS caused delayed UH closure and increased severity, but other articles did not report similar events.12 Despite the low prevalence, doctors need to explain these possibilities and pay attention to the umbilical condition when changing AS.
This study had several limitations. First, this was a meta-analysis of cohort studies with somewhat different endpoints. While some studies provided the specific period of closure, others adopted whether the UH was closed at a specific age. Second, the sample sizes of each study were probably insufficient to reach a conclusion. Third, each meta-analysis, particularly that of the primary outcome and that of the efficacy at 6 months old, contained moderate heterogeneity. We presumed that this was due to the insufficient sample size of the included studies. Following the protocol, we did not conduct post-hoc subgroup analyses to determine other factors that affected the results. However, when related studies are accumulated, subgroup analyses should be conducted according to differences in follow-up periods, AS procedures such as with or without plugs to reduce UH, and the UH closure rate in the observation group. Next, there was a possibility that we may not have found some old studies. Finally, recent studies were only conducted in Japan; thus, language bias may exist. Therefore, further studies at higher evidence levels are necessary to reach a definitive conclusion.
In conclusion, our study did not clarify the significant difference in the overall efficacy of AS on UH closure. However, there might be advantages to accelerating the closure speed and preventing protruding umbilici with redundant skin. Due to the high heterogeneity of our study, further studies at higher evidence levels are warranted before reaching a definitive conclusion.
