Does timing of inguinal hernia repair in preterm infants influence serious adverse effects?
Charles Coté MD, Priti Dalal MD, and Myron Yaster MD
Inguinal hernia repair is one of the most common surgeries performed in preterm infants. Whether the inguinal hernia should be repaired prior to or after discharge from the neonatal intensive care unit is controversial: either alternative can result in adverse events.1 An early repair can result in anesthetic induced apnea, bradycardia, desaturation, the need for prolonged mechanical ventilation, and delayed discharge from the NICU. A late or delayed repair after discharge from the NICU may increase the risks of hernia related complications like incarceration.
In a randomized controlled, multi-institutional study, today’s PAAD by the HIP trial investigators1 sought to finally and definitively answer this question. Interestingly (or maddeningly), this well thought out and funded study included pediatric surgeons and neonatologists but no anesthesiologists, despite the fact that most of the complications of early surgery are anesthesia related. Once again, I have to go back to the famous line in the immortal movie Casblanca: “I’m shocked, just shocked that gambling is going on here”.
I’ve asked 2 anesthesiologists who are thought leaders in this area, Drs. Charles (Charlie) Coté and Priti Dalal, to assist in writing today’s PAAD. As you will see, Charlie and Priti dissect the findings of this article in great detail, resulting in a PAAD that is much longer than our target of 5-6 minute reads. On the other hand, the stated conclusions are so important, you will almost certainly be faced with making decisions in your daily practice based on this article. Myron Yaster MD
Original article
HIP Trial Investigators; Blakely ML, Krzyzaniak A, Dassinger MS, Pedroza C, Weitkamp JH, Gosain A, Cotten M, Hintz SR, Rice H, Courtney SE, Lally KP, Ambalavanan N, Bendel CM, Bui KCT, Calkins C, Chandler NM, Dasgupta R, Davis JM, Deans K, DeUgarte DA, Gander J, Jackson CA, Keszler M, Kling K, Fenton SJ, Fisher KA, Hartman T, Huang EY, Islam S, Koch F, Lainwala S, Lesher A, Lopez M, Misra M, Overbey J, Poindexter B, Russell R, Stylianos S, Tamura DY, Yoder BA, Lucas D, Shaul D, Ham PB 3rd, Fitzpatrick C, Calkins K, Garrison A, de la Cruz D, Abdessalam S, Kvasnovsky C, Segura BJ, Shilyansky J, Smith LM, Tyson JE. Effect of Early vs Late Inguinal Hernia Repair on Serious Adverse Event Rates in Preterm Infants: A Randomized Clinical Trial. JAMA. 2024 Mar 26;331(12):1035-1044. doi: 10.1001/jama.2024.2302. PMID: 38530261; PMCID: PMC10966421.
Before discussing the findings let’s review some basics. Apnea of prematurity is commonly defined as a “cessation of breathing for > 20 seconds or less if accompanied by bradycardia (HR 100 beats/min or hypoxemia, cyanosis, or pallor.2,3 Apnea is extremely common in premature infants (gestational age<37 weeks at birth) with an incidence that increases with decreasing gestational age. Anemia is also a well known major contributing factor but was not included in the study design.2 In today’s PAAD, premature infants were randomly assigned to early inguinal hernia repair (before discharge from NICU) vs late outpatient surgical repair (after discharge from NICU and > 55 weeks PMA).1 Post menstrual age (PMA) was used rather than post conceptual age (PCA), which is often used in other studies and often results in confusing and conflicting results.4 PMA is ~ 2 weeks or longer than PCA. Thus, the 55 week PMA cut off used as a criteria for classifying the infants into early versus late repair group, in this study, is equivalent to the 53 weeks PCA.
A deeper dive:
The intent of the current paper was to compare the incidence of serious adverse events during a 10-month follow-up period with a secondary outcome of total hospital days following repair. Study subjects were enrolled approximately 2 weeks prior to estimated discharge. All of the adverse events were collected daily by post hoc chart review for inpatients and weekly for outpatients. Families were contacted by phone at 30 days after hernia repair and 10 months after enrollment. They stratified enrolled patients by GA (<28 weeks or ≥ 28 weeks). They reported that 137/159 in the early surgery cohort and 122/149 in the late cohort had apnea but they described “serious apnea” as the need for increased oxygen, additional respiratory support such as bag mask ventilation, increased respiratory support or initiation of mechanical ventilation. However, based on Table 3 in the article, some infants had increases in mechanical ventilation or prolonged intubation as intervention for apnea which makes comparison between the two groups difficult since sedation/analgesics may be confounding factors. Bradycardia was defined by the need for pharmacologic intervention (agent not specified); no threshold definition was presented. There are no data regarding what monitoring was used to assess these events (expired CO2, pulse oximetry, EKG, etc.), how long they were monitored in the PACU, where the apnea event occurred (PACU, PICU/NICU, ward, at home) or how long after anesthesia and whether the post-operative destination was also NICU in the late repair group.
338 infants were enrolled with 172 in the early repair cohort and 166 in the late repair cohort, of whom 320 underwent surgical repair. Complete data were available for 308 infants (159 early vs 149 late repair).152/159 in the early cohort underwent repair prior to discharge primarily under general anesthesia (99%) with a median postmenstrual age of 41 weeks. In the late repair cohort 129/149 underwent their repair at a median postmenstrual age of 57 weeks but 39 underwent repair <55 weeks PMA. 28% (N = 44) of the early repair cohort experienced at least 1 serious adverse event (including 1 local anesthetic overdose) vs 18% (N = 27) in the late repair cohort, a risk difference estimated as -7.9% but a total of 61 serious adverse events were recorded in the early repair cohort and 40 in the late repair cohort. The risk reduction for the primary outcome (occurrence of any of the identified serious adverse events during the 10 month observation period) was particularly evident in those <28 weeks GA and those with bronchopulmonary dysplasia (favoring late surgery). They also found a greater number of hospital days in the early cohort (19 vs 16 days) but the range of hospital days was similar in both cohorts. The authors concluded that their findings support late repair after discharge from NICU. However, is this conclusion supported by the data presented in the article? We aren’t quite so sure.
This study is quite confusing as to how the data are presented. For example, data regarding the timing of surgery are presented in interquartile ranges rather than the exact frequency ranges plotted against PMA or PCA, which confounds comparison of these data with published studies. Anesthesia techniques/drugs/ analgesics were not standardized and there is no report of anesthesia data which may have impacted the cardio-respiratory adverse events in the early phase of post-operative recovery. Table 1 presents summary data for 320 infants (163 early and 157 late); these summary statistics are misleading, since in reality, only 152 (not 163) were included in the final early cohort and 129 (not 157) actually had their repair late (IQR 52-61 weeks PMA). The data regarding PMA (presumably at time of surgery) in Table 1 is also confusing because the authors state a PMA of 39 weeks (IQR 37-41 weeks) in the late hernia group, contrary to information in the text, which states PMA 57 weeks (IQR 52-61 weeks. There are no data regarding the distribution of small for gestation age, appropriate for gestational age or large for gestational age, nor are there data regarding the number of NICU days prior to the procedure. Twenty-eight did not have a late repair due to: resolution of hernia (N = 17), refusal by parent (N = 4), unrelated death (N=2), and other (N = 5). In addition, 39 patients in the late cohort actually underwent repair earlier than 55-weeks PMA, including 11 where there was “concern for incarceration”; there are no data describing if those infants in whom there was a concern for incarceration were elective, urgent or emergent.
If we dig down a bit further into the data presented in table 3, there were 4 episodes of cardiopulmonary resuscitation in each cohort but 3 cardiac arrests in the late cohort and none in the early cohort. There are no data regarding how many patients required intubation, mechanical ventilation, simple stimulation, bag mask ventilation or just supplemental oxygen for treatment of apnea. Likewise, bradycardia without the need for pharmacologic intervention is common in former preterm infants after anesthesia but such events were not reported, only those considered serious and requiring pharmacologic intervention were included. This would seem to be a rather late response in our opinion, and no data are provided as to how low the heart rates went during such events or what drugs were used to “pharmacologically” treat the bradycardic events. Bradycardia in most published studies is generally defined as a heart rate < 80; the “need for pharmacologic intervention” was their definition of a serious event but it is unusual for this type of response. The fact that they reported 9 such interventions (5 early and 4 late) is suggestive of much more serious contemporaneous events such as bradycardia induced by severe hypoxemia. Since their definition of apnea depended upon chart review of interventions, they likely missed many brief apneas and brief bradycardia that self-resolved. Indeed, in a microanalysis of my (CC) previous study, we found a nearly 50% greater incidence of apnea and bradycardia events when continuous data collection was used rather than nursing observation.2
Additionally, details regarding the three cardiac arrests in the late repair cohort are absent and these very serious events are not even discussed as to causation or outcomes. It is also unclear how many infants were managed with spinal anesthesia which would have reduced early but not late apnea,5 although surgery was done under general anesthesia in 99% of the early group and 98% of the late group.
We would certainly expect a greater incidence of apnea and bradycardia in the early cohort as the increased incidence with lower GA and PCA (PMA) at the time of surgery has been well established and was the basis of the current paper and study design. It is our interpretation that their outcomes should have focused upon the anesthesia related events (apnea, bradycardia, cardiopulmonary resuscitation) but these were mentioned in passing with no discussion. Prolonged intubation and pneumonia certainly would be expected in the younger patients (N = 8), and perhaps should not be counted as serious adverse events in the early group as they were not yet NICU graduates. There were 37 severe apneas and bradycardic events in the early cohort but obviously some patients had more than one such event and there were 17 such events in the late cohort. The greater number of serious apneic events in the younger cohort is certainly consistent with previously published data. The manner in which the overall data are presented is somewhat confusing since only 129 not 149 of the late cohort infants actually had a late repair and 39 were repaired when younger than 55 weeks PMA, indicating significant overlap in PMA between the two groups, with the differentiating characteristic being inpatient (NICU) or outpatient status). We do not have a frequency vs PMA distribution for patients with such events, in particular for these 39 patients. Of far greater concern is their report of 3 cardiac arrests in the late cohort vs none in the early cohort, and 6 incarcerated hernias in the late cohort vs 2 in the early cohort (not counting the 11 operated upon < 55 weeks PMA due to “concerns for inguinal hernia incarceration”). If we examine the incidence of incarcerated hernia (2 early vs. 6 late) then my (CC) power analysis suggests that 1246 patients would need to be studied for an alpha of 0.05 with 0.8 power, however if we throw in the 11 cases where there was “concern for incarceration” (now 17 in the late cohort) then the number of cases drops to 88 patients per cohort. This analysis would have strengthened the discussion section of their report; avoiding an urgent or emergent procedure would seem preferable both from surgical and anesthesia perspectives.
In conclusion, this paper unfortunately had a great potential for providing improved insight into the risks for these vulnerable infants, but the data collection and interpretation fall far short of providing much in the way of meaningful guidance beyond what we already know. They have confirmed that the later the PMA and the older the GA at the time of surgery, the lower the incidence of post operative apnea. Had they provided the timing of such events after anesthesia and performed a frequency analysis of apnea plotted against weeks PMA, they would have shed much more light on this issue. Unfortunately, despite all the money and time spent on collecting these data, nothing new is provided. The good news is that, as with many studies, two new rocks were turned over. 1) A threefold greater incidence of incarceration in the late cohort (6 vs 2) (plus 11 others in the late group operated for fear of incarceration) is actually suggestive that early repair is indicated rather than waiting for it to become urgent or even emergent (no data provided). We would interpret this to mean that early surgery is indicated rather than late surgery; 2) of even greater concern is why did those 3 cardiac arrests occur and what could have been done differently to prevent them? As they always say with most papers “further study is needed”!
Send your thoughts and comments to Myron who will post in a Friday reader response.
References
1. Investigators HT. Effect of Early vs Late Inguinal Hernia Repair on Serious Adverse Event Rates in Preterm Infants: A Randomized Clinical Trial. Jama 2024;331(12):1035-1044. DOI: 10.1001/jama.2024.2302.
2. Coté CJ, Zaslavsky A, Downes JJ, et al. Postoperative apnea in former preterm infants after inguinal herniorrhaphy. A combined analysis. Anesthesiology 1995;82(4):809-822.
3. Eichenwald EC. Apnea of Prematurity. Pediatrics 2016;137(1) (In eng). DOI: 10.1542/peds.2015-3757.
4. Engle WA. Age terminology during the perinatal period. Pediatrics 2004;114(5):1362-4. (In eng). DOI: 10.1542/peds.2004-1915.
5. Davidson AJ, Morton NS, Arnup SJ, et al. Apnea after Awake Regional and General Anesthesia in Infants: The General Anesthesia Compared to Spinal Anesthesia Study--Comparing Apnea and Neurodevelopmental Outcomes, a Randomized Controlled Trial. Anesthesiology 2015;123(1):38-54. (In eng). DOI: 10.1097/aln.0000000000000709.