Perioperative Hypoxemia and Postoperative Respiratory Events in Infants with Hypertrophic Pyloric Stenosis

Background Normalization of metabolic alkalosis is an important pillar in the treatment of infantile hypertrophic pyloric stenosis (IHPS) because uncorrected metabolic alkalosis may lead to perioperative respiratory events. However, the evidence on the incidence of respiratory events is limited. We aimed to study the incidence of peroperative hypoxemia and postoperative respiratory events in infants undergoing pyloromyotomy and the potential role of metabolic alkalosis. Materials and Methods We retrospectively reviewed all patients undergoing pylo-romyotomy between 2007 and 2017. All infants received intravenous ﬂ uids preoperatively to correct metabolic abnormalities close to normal. We assessed the incidence of perioperative hypoxemia (de ﬁ ned as oxygen saturation [SpO 2 ] < 90% for > 1min) and postoperative respiratory events. Additionally, the incidence of dif ﬁ cult


Introduction
Pyloromyotomy is the standard treatment for infantile hypertrophic pyloric stenosis (IHPS) and a frequently performed procedure in infants.IHPS occurs predominantly in males, aged from 2 to 12 weeks and causes nonbilious projectile vomiting. 1,2][5] It is assumed that metabolic alkalosis in infants with IHPS may affect the central control of ventilation and the respiratory drive, leading to perioperative apnea and extubation difficulties. 6In a recent literature review, we showed an incidence of postoperative respiratory events in infants with IHPS of 0.2 to 16%. 7However, this review was hampered by a small number of included patients and the lack of studies that primarily investigated the occurrence of apnea in infants with IHPS.In addition, the included studies used different definitions of apnea that led to heterogeneous results regarding the occurrence of postoperative respiratory events.Furthermore, the association between postoperative apnea and metabolic alkalosis was unclear, because there was a lack of data regarding the extent of metabolic alkalosis prior to surgery and potential bias by other factors contributing to apneic episodes.Yet, despite the limited evidence normalization of the biochemical abnormalities remains an important, internationally accepted pillar in the treatment of IHPS to prevent potential perioperative respiratory events.][10][11] In a recent Delphi analysis, we obtained consensus among an international expert panel that recommended, among others, pH less than 7.45, bicarbonate less than 26 mmol/L, and chloride more than or equal to 100 mmol/L as cutoff values before safe pyloromyotomy. 12wo recent studies showed a positive association between preoperative serum bicarbonate and preoperative respiratory events and increased time to extubation post-surgery respectively. 13,14Both studies suggested that a lower cutoff value of bicarbonate than used nowadays should be considered.However, more evidence is required to assess the risk of the development of peri-and postoperative respiratory events in infants with IHPS and the influence of metabolic alkalosis.
The aim of this retrospective study is to evaluate the incidence of perioperative hypoxemia and postoperative respiratory events in infants with IHPS undergoing pyloromyotomy and whether there is a correlation with the extent of preoperative metabolic alkalosis.

In-and Exclusion Criteria
Patients with IHPS who were treated with open or laparoscopic pyloromyotomy at our two pediatric surgical centers (Amsterdam UMC, location AMC and VUmc, the Netherlands) between 2007 and 2017 were included in this retrospective study.Infants were selected by using diagnostic and treat-ment codes specific for IHPS and pyloromyotomy in accordance with the National Health Insurance system.Infants with (congenital) abnormalities of the respiratory tract, respiratory tract infections, or other conditions potentially influencing the breathing regulation were excluded from the analysis.The infants who underwent exploratory surgery or pyloromyotomy in combination with any other surgical procedure or whose medical files or anesthesia records were incomplete were excluded as well.All perioperative data were retrieved from an automatic electronic data management system, although this varied over time and location.The Medical Ethics Committee and the Clinical Research Board reviewed the study protocol and confirmed that the Medical Research Involving Human Subject Act ("WMO") did not apply (2018.210)and waived written informed consent.This manuscript adheres to the applicable CONSORT guidelines.

Data Extraction
Hospital files were reviewed by one of the authors (FvdB) regarding sex, age, medical history, clinical course, and laboratory results (capillary or venous blood gas analysis [pH, base excess, bicarbonate, and partial pressure of carbon dioxide], sodium [Na], potassium [K], chloride [Cl], glucose, and hemoglobin).In the Netherlands, infants with IHPS are usually referred from regional hospitals to a pediatric surgical center for operative treatment.Therefore, we also screened the letter of referral in regard to the resuscitation policy, laboratory values, and the clinical course.In addition, we reviewed surgical technique, anesthetic regimen, dose and type of anesthetic agents, information regarding intubation, timing and duration of the procedure, and intra-and postoperative (respiratory) events.

Perioperative Care
The diagnosis of pyloric stenosis was made by ultrasound and/or by test feeding with visualization of peristaltic waves.After admission to the hospital, infants generally received a nasogastric tube and intravenous fluid rehydration.During the study period, various combinations of sodium chloride and dextrose were used for resuscitation.The most commonly used resuscitation policies were intravenous sodium chloride 0.45% with dextrose 5% or sodium chloride 0.3% with glucose 3.3% with standard addition of 10 to 20 mEq/L potassium in case of hypokalemia.This was modified depending on the specific needs of the infant.Ongoing losses via the nasogastric tubes were compensated with normal saline.If hypochloremic metabolic alkalosis was normalized close to normal (no exact cutoff values were applied), open or laparoscopic pyloromyotomy was performed.The choice for the open or laparoscopic approach was up to the surgeons' preference based on their experience with these procedures.At the theater, the nasogastric tubes were suctioned before induction frequently while turning the infant into a side and prone position.Prophylactic antibiotics were given infrequently at the request of the attending surgeon.Postoperatively infants were transported to the postanesthesia care unit (PACU) or-if indicated-to the pediatric intensive care unit (PICU).

Classification of Respiratory Events
No standard classification tool for respiratory events is available.Therefore, we classified the respiratory events dichotomously as present or absent independent of the severity and duration.Infants were classified as having hypoxemia if oxygen saturation levels (SpO 2 ) dropped below 90% during induction of anesthesia, intraoperatively, or during emergence.Additionally, we scored deep desaturations (SpO 2 > 80%) and prolonged desaturations (SpO 2 < 90% for 2 minutes or 5 minutes, respectively).Arterial oxygen saturation was monitored continuously by a peripheral pulse oximeter and documented automatically by our anesthesia information management system (AIMS).The monitoring system averaged the saturation over 10 seconds and all data were stored once per minute.All AIMS records were manually inspected to decipher and disregard measurement artifacts.In the postoperative period, infants were classified as having respiratory events if one or more of the following items was noted in the hospital files during recovery or postoperative course until 2 days after surgery: hypopnea (respiratory rate < 20/min), apnea, desaturation (SpO 2 80% or SpO 2 < 90% for 2 minutes), or stridor.Only in neonates and ex-premature infants younger than 60 postconceptional weeks, respiratory monitoring on the ward was continued up to the following day.
Induction was classified as the difference in time between start induction and incision; intraoperative or surgical time as the difference in time between incision and discontinuation of sevoflurane or continuous intravenous propofol and emergence as the difference between discontinuation of sevoflurane and extubation as indicated by drop or stop of end-tidal carbon dioxide measurement.We assessed all laboratory values collected during the first 3 days of clinical admission and used the most abnormal value as admission laboratory values.Preoperative values were defined as the last obtained blood sample before pyloromyotomy.

Statistical Analysis
We used SPSS version 26 for all statistical analyses.Normality assumption was assessed using the Shapiro-Wilk test.Continuous data were summarized as median with interquartile range (IQR) and compared using the independent T-test or Mann-Whitney U test.Categorical data were summarized as frequencies with percentages and analyzed using Pearson's chi-squared test or Fisher's exact test, respectively.Significance levels were set at p-value of 0.05 (uncorrected).Multivariate logistic regression analysis was performed to analyze the association between pH, bicarbonate, chloride, and respiratory events while correcting for prematurity and birth weight.The oral morphine equivalent was calculated for intravenous fentanyl, sufentanil, and intravenous morphine.

Patient Characteristics
During the study period, 478 infants underwent open or laparoscopic pyloromyotomy.Twenty-six infants were ex-cluded due to (congenital) abnormalities of the cardiorespiratory tract (n ¼ 4), respiratory tract infection (n ¼ 4), other conditions potentially influencing the central breathing regulation (n ¼ 10), or (combination of) different surgical procedures (n ¼ 8).We excluded another 46 infants from further analysis because medical files or anesthesia records were incomplete.Of 406 included infants, the majority were male (n ¼ 345, 85.0%).In total, 213 infants underwent laparoscopic pyloromyotomy (52.5%) compared to 193 infants who underwent open pyloromyotomy (47.5%).Slightly more infants were operated on at location 1 than at location 2 (n ¼ 252 vs. n ¼ 154).Patient characteristics are shown in ►Table 1.
All infants received general anesthesia (►Table 2).Inhalation induction solely was used in 57 infants (14.0%); all other infants underwent intravenous induction.Rapid sequence induction (RSI) or modified RSI (mRSI) was used in 71 infants (17.5%).Of 333 infants, 25 (7.5%) were classified as difficult by the attending pediatric anesthetist.Furthermore, 17 infants required three or more attempts for successful endotracheal intubation.However, the number of attempts was only noted in 73 infants.Ten infants received dexamethasone and ten infants received naloxone during the emergence or postoperative period.

Incidence of Perioperative Hypoxemia and Postoperative Respiratory Events
More than half of the patients (n ¼ 208; 51.2%) developed one or more desaturations (SpO 2 90% for > 1 min) during the perioperative period, 130/406 (32.0%) during induction; 43/406 (10.6%) during the intraoperative period and 112/406 (27.6%) during emergence.The incidence and severity of perioperative respiratory events are shown in ►Table 3.In total 86/406 infants (21.2%) developed respiratory events during the postoperative course.The incidence of postoperative respiratory events is further specified in ►Table 4. Twelve infants (3.0%) developed apnea, three infants developed respiratory insufficiency necessitating prolonged (non)invasive ventilation in the PICU, and three infants experienced a not further specified incident.Most of the infants with postoperative respiratory events developed one or more episodes of hypoxemia (n ¼ 51; 59.3%).The incidence of hypoxemia differed largely between both participating centers.This was most likely due to differences in the documentation of peripheral pulse oximeter data on the PACU, presumably leading to an underestimation.No significant differences were found between infants with and without hypoxemia or respiratory events, except for birth weight and type of procedure (►Table 5).Infants with postoperative respiratory events had a significantly lower birth weight compared to infants without postoperative respiratory events (3,200 vs. 3,550 g, p < 0.001).Open pyloromyotomy was associated with more perioperative hypoxemia (74.4% vs. 25.6%,p < 0.001) and fewer postoperative respiratory events (32.6% vs. 67.4%,p ¼ 0.002) compared to a laparoscopic procedure.Furthermore, infants classified as difficult to intubate had more perioperative desaturations compared to infants without difficult intubation.However, difficult intubation did not lead to differences in respiratory events during the continued clinical course.

The Extent of Metabolic Derangement
Median serum levels (IQR) of pH, bicarbonate, and chloride, as measured during the first days of admission, were 7.45 (0.10), 30.0 (6.3) mmol/L, and 100 (8.0) mmol/L respectively.After fluid and electrolyte administration, laboratory values were corrected to about normal (pH 7.41 [0.05], bicarbonate 24.8 [3.8] mmol/ L and chloride 106 [5.0] mmol/L).Multivariate analysis of admission laboratory values of pH, bicarbonate or chloride and the presence of perioperative hypoxemia or postoperative respiratory events, adjusted for prematurity and birth weight, showed no significant association except for pH serum values and incidence of hypoxemia during the emergence of anesthesia that was statistically significant but was considered as clinically irrelevant (►Table 6).Furthermore, we found a significant association between higher preoperative pH serum values and incidence of hypoxemia during induction and emergence, but this effect is rather small (►Table 6).The same applied to preoperative laboratory values of chloride and induction.

Discussion
In this retrospective study, we evaluated the incidence of perioperative hypoxemia and postoperative respiratory events in infants with IHPS undergoing pyloromyotomy and found an incidence of 51% of at least one episode of  hypoxemia during the perioperative period and of 22% of postoperative respiratory events in 406 consecutive infants.Furthermore, we evaluated whether there was a correlation between the occurrence of respiratory events and the extent of preoperative metabolic alkalosis and did not find a clinically relevant association between admission or preoperative laboratory values reflecting metabolic alkalosis and peri-or postoperative respiratory events.There is a lack of a uniform general definition of pediatric hypoxemia, which makes it difficult to directly compare our results with others.6][17][18] In a large prospective Australian study of 9,297 children who had general anesthesia for surgical or medical interventions, elective or urgent procedures, 919 (10%) of the children experienced oxygen desaturation (SpO 2 < 95% for unspecified duration). 16The findings of this study cannot be fully extrapolated to our study population because of age differences.It is well known that infants have an increased risk to develop hypoxemia. 17de Graaff et al evaluated the incidence of hypoxemia in relation to age in pediatric noncardiac surgery patients in a retrospective cohort study. 18They found that 135 of 589 infants (aged between 0 and 3 months) developed mild hypoxemia (SpO 2 < 90% for at least 1 minute) and 127 of 589 infants developed severe hypoxemia (SpO 2 < 80% for at least 1 minute), leading to a total incidence of hypoxemia of 44% in this age group.The incidence of hypoxemia was higher in infants younger than 1 month compared to infants aged 1 to 3 months.
When we compare the incidence of hypoxemia during the different perioperative periods, we found that most desaturations occurred during induction.Almost a third of the study population developed a desaturation of SpO 2 less than 90% for at least one minute, while 16.7% developed a severe desaturation of SpO 2 less than 80% (►Table 3).Most infants received intravenous induction or a combination of inhalation induction with intravenous induction.RSI or mRSI was documented in 17.5%.The choice for the optimal anesthetic induction technique in patients with IHPS is still controversial.0][21] There are also varying opinions on the risks of hypoxemia when using RSI and mRSI in infants with IHPS. 22,23The incidence of difficult intubation was 7.5% in our cohort.This is higher compared to the results of the NECTARINE (NEonate-Children sTudy of Anaesthesia pRactice IN Europe) study, which prospectively investigated the incidence of difficult intubation in infants under the age of 60 postconceptional weeks and found an incidence of 5.8% of difficult intubations. 24However, they defined difficult intubation as 2 or more intubation attempts, while in our study infants were classified as having difficult intubation by the attending pediatric anesthetist regardless of the number of intubations attempts.
In this study, we found an incidence of postoperative apnea of 3.0%, which is in line with the results of 0.2 to 16% in a recent systematic review. 7The total incidence of postoperative respiratory events, consisting of desaturation, stridor, hypopnea, and apnea was 21.2%.We found that infants who underwent open pyloromyotomy had more often a period of perioperative hypoxemia and less often postoperative respiratory events compared to infants who underwent a laparoscopic procedure.The increased incidence of perioperative hypoxemia during open pyloromyotomy may be explained by the invasiveness of the open approach.Conversely, the increased incidence of hypoxemia in the postoperative period in the laparoscopic group may be explained by pneumoperitoneum or hypercarbia.However, these effects might be mitigated by concurrent changes in anesthetic or surgical management over the 10-year period.
We did not find a clinically meaningful association between admission or preoperative laboratory values reflecting metabolic alkalosis and respiratory events.Although preoperative serum chloride was significantly associated with hypoxemia during induction, serum chloride was not associated with hypoxemia during the other time intervals.We did not correct for multiple testing; thus, this single association may be caused by a multiple testing problem and should be interpreted with caution.Furthermore, contrary to our hypothesis that metabolic alkalosis would lead to desaturation, increased serum pH was negatively associated with desaturation.However, these are just very small differences, especially when compared to other risk factors like recent or current upper airway infections, which can respectively double or quadruple the risk of desaturations during sedation. 25Our results are contrary to the results of Gilbertson et al who found a rather weak correlation between preoperative serum bicarbonate and emergence time. 14Remarkably, we found a correlation of bicarbonate at admission as a marker for alkalosis and preoperative respiratory events in our earlier analysis. 13Thus, within a certain range of acidbase and electrolyte correction, minor deviations are not associated with an increased risk of perioperative respiratory events.The most important limitation of this study is the retrospective design.Furthermore, not all infants did receive standard respiratory monitoring postoperatively at the ward and documentation was different between both centers.This, the incidence of postoperative respiratory events, including apnea, may have been underestimated.Nevertheless, we found a high incidence of respiratory events throughout the whole perioperative period.In conclusion, the results of this research show that infants with IHPS have a risk to develop perioperative hypoxemia and postoperative respiratory events.Although these events are mostly very short and easily treated, they are a potential threat to the patient, especially in an institution with relatively low volumes of this patient group.Further research is required to prospectively identify the safe range of acid-base and electrolyte correction and other adjustable factors such as the anesthetic technique that might reduce the relatively high rates of peri-and postoperative respiratory events.

Table 1
Patient characteristicsPerioperative Hypoxemia and Postoperative Respiratory Events in IHPS van denBunder et al. 487This document was downloaded for personal use only.Unauthorized distribution is strictly prohibited.

Table 2
Perioperative and anesthetic details

Table 3
Incidence of perioperative hypoxemia split by severity and duration Abbreviation: SpO 2 , oxygen saturation level.Note: Values are number (percentage).

Table 4
Incidence of postoperative respiratory eventsPerioperative Hypoxemia and Postoperative Respiratory Events in IHPS van denBunder et al. 489This document was downloaded for personal use only.Unauthorized distribution is strictly prohibited.

Table 5
Patient demographics and anesthetic details for infants with hypoxemia versus infants without hypoxemia during the different per-and postoperative time intervals Abbreviations: IQR, interquartile range; mRSI, modified rapid sequence induction; SpO 2 , oxygen saturation level.Note: Values are median[IQR]or number (percentage).Hypoxemia means SpO 2 90% during at least 1 minute.Respiratory events were classified when more of the following was present: hypopnea (respiratory rate < 20/min), apnea, hypoxemia (SpO 2 90% for > 1 minute) or stridor.

Table 6a
Correlation between admission laboratory values and perioperative and postoperative respiratory events : Values are odds ratio (OR) with 95% confidence interval (95% CI), corrected for prematurity and birthweight.a Due to the narrow interval pH is shown as per 0.01 increase. Note

Table 6b
Correlation between preoperative laboratory values and perioperative and postoperative respiratory eventsPerioperative Hypoxemia and Postoperative Respiratory Events in IHPS van denBunder et al. 491This document was downloaded for personal use only.Unauthorized distribution is strictly prohibited.
Note: Values are odds ratio (OR) with 95%-confidence interval (95% CI), corrected for prematurity and birthweight.aDue to the narrow interval pH is shown as per 0.01 increase.European Journal of Pediatric Surgery Vol.33 No. 6/2023 © 2023.Thieme.All rights reserved.