| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Although former preterm birth infants are at risk for postoperative apnea after surgery, it is unclear whether the same is true of full-term birth infants. We evaluated the incidence of apnea in 60 full-term neonates and infants undergoing pyloromyotomy both before and after anesthesia. All subjects were randomized to a remifentanil- or halothane-based anesthetic. Apnea was defined by the presence of prolonged apnea (>15 s) or frequent brief apnea, as observed on the pneumocardiogram. Apnea occurred before surgery in 27% of subjects and after surgery in 16% of subjects, with no significant difference between subjects randomized to remifentanil or halothane anesthesia. This apnea was primarily central in origin, occurred throughout the recording epochs, and was associated with severe desaturation in some instances. Of the subjects with normal preoperative pneumocardiograms, new onset postoperative apnea occurred in 3 (23%) of 13 subjects who received halothane-based anesthetics versus 0 (0%) of 22 subjects who received remifentanil-based anesthetics (P = 0.04). Thus, postoperative apnea can follow anesthesia in otherwise healthy full-term infants after pyloromyotomy and is occasionally severe with desaturation. New-onset postoperative apnea was not seen with a remifentanil-based anesthetic. IMPLICATIONS: Abnormal breathing patterns can follow anesthesia in infants after surgical repair of pyloric stenosis. Occasionally, these patterns can be associated with desaturation. New-onset postoperative apnea was not seen with a remifentanil-based anesthetic.
Infants born prematurely experience apnea after anesthesia and surgery (14). In contrast, disturbances in respiratory control in full-term infants are less well documented. Although postoperative apnea has been noted in case reports for otherwise healthy full-term infants (58), respiratory monitoring in 361 full-term neonates and infants undergoing hernia repairs has demonstrated no postoperative apnea (1,912). In only one study (13) was postoperative apnea in a young infant noted (6 wk old, full term). Chipps et al. (14) looked at pneumocardiograms in full-term infants undergoing pyloric stenosis repair and found decreased apnea in the postoperative period and no new onset of postoperative apnea. Given the concerns about postoperative apnea and opioids, there is controversy regarding anesthetic choice for young infants. In the past, anesthesiologists were reluctant to give opioids to neonates and young infants because of slower drug metabolism and associated respiratory depression. However, the analgesic properties of opioids make their use desirable for surgical procedures. Remifentanil is an ultra-short-acting opioid metabolized by nonspecific esterases. This rapid on/rapid off effect has led some anesthesiologists to use remifentanil in infants to minimize post-opioid related side effects (15,16). The purpose of this part of the study was to determine perioperative respiratory patterns of young infants undergoing pyloric stenosis repair and to compare an inhaled anesthetic with an opioid-based anesthetic on postoperative breathing patterns.
This multicenter study was conducted at seven sites (Childrens Hospital of Philadelphia, PA; Childrens Hospital of Pittsburgh, PA; The Childrens Hospital, Boston, MA; The Johns Hopkins Hospital, Baltimore, MD; Childrens Hospital and Regional Medical Center, Seattle, WA; Hermann Hospital, Houston, TX; and Lucille Packard Childrens Hospital at Stanford, Palo Alto, CA). Each local IRB approved this randomized, open-label, single-blinded study. All subjects were full-term infants (gestational age 37 wk, gestational weight 2500 g). Enrollment criteria included ASA patient class I or II, <8 wk postnatal age, a diagnosis of pyloric stenosis, and scheduling for pyloromyotomy. Exclusion criteria included an ASA patient class III or greater, family history of malignant hyperthermia, documented history of preoperative apnea and bradycardia, family history of sudden infant death syndrome, liver failure, Downs syndrome, hypersensitivity to or recent (within 12 h) administration of opioids, previous participation in this study, or psychiatric illness of parent or guardian. After written informed consent was obtained from parent or guardian, the subjects were entered into the study. Blood chemistries were recorded from the chart before surgery (chloride, bicarbonate, and hematocrit). All subjects underwent a thermistor pneumocardiogram pulse oximetry recording (Edentrace E000052 recorder with Edentec retrieval software; Edentrace Archival software, version 5, and Edentrace Analysis software, version 1.2; Nellcor Puritan Bennett, Plea-santon, CA) for a minimum of 2 h immediately before surgery and for 12 h beginning in the immediate postoperative period. A thermistor sensor taped below the nares monitored airflow, chest wall impedance monitored breathing rate, electrocardiogram monitored heart rate, and pulse oximetry monitored arterial oxygenation. Oxygen saturation was multiplexed with heart rate to help identify artifacts. After the preoperative pneumocardiogram was completed and the subjects fluid and electrolyte imbalances were corrected, the subject entered the operating room for surgical repair of the pyloric stenosis. Anesthesia was induced with propofol, and the trachea was intubated after atropine and succinylcholine administration. Subjects were randomized to receive nitrous oxide and oxygen along with either remifentanil or halothane for the maintenance part of their anesthesia. Randomization was in a 2:1 ratio of remifentanil:halothane. Analgesia was achieved with acetaminophen and local infiltration of the wound by the surgeons during surgery for both groups. Details of the anesthetic administration are presented in Part I (17). After the end of surgery and after the trachea had been extubated, the subject entered the postoperative care unit, where the thermistor pneumocardiogram pulse oximetry recorder was restarted immediately upon entering and continued for 12 h. The subjects were subsequently discharged home after meeting discharge criteria set forth by the surgeons at their respective institutions. One investigator blinded to group assignment read the pnuemocardiograms (CDK). Oxygen saturation values were considered accurate when the heart rate by pulse oximetry matched the heart rate by electrocardiogram. The following variables were analyzed:
Active time was identified by the presence of motion artifact on the thermistor, impedance, or pulse oximetry recording. Quiet time was the difference between total time and active time. The apnea index is the total number of seconds of apnea divided by the total number of seconds of quiet time multiplied by 100. An abnormal thermistor pneumocardiogram study was defined as an apnea index >5 for the entire study or >10 for any time epoch or the presence of any prolonged apnea. Because subjects are most susceptible to apneic type events during quiet time (sleep), the following variables were analyzed after normalization to quiet time:
Apneic episodes were categorized as central, obstructive, or mixed events. Central apnea was identified by the cessation of both airflow and chest wall movement, obstructive apnea by the cessation of airflow with continued chest wall movement, and mixed apnea by the cessation of airflow with and without chest wall movement in the same apneic episode (2).
Brief apnea was defined as the cessation of nasal air flow for >5 and <15 s (2,18,19). Prolonged apnea was defined as cessation of airflow for Data were analyzed in time epochs as follows: 059 min, 60119 min, 120179 min, 180359 min, 360539 min, and 540720 min. The short duration of early epochs was designed to capture respiratory impairment related to residual anesthetics. Summary statistics are reported as mean, SD, and range for normally distributed variables and median, 25th and 75th percentiles for nonnormally distributed variables. Statistical significance was defined as P < 0.05. Treatment comparisons of dichotomous data were analyzed with Fishers exact test. Comparisons of pre- versus postoperative dichotomous variables were analyzed with an exact McNemars test. Comparisons of pre- versus postoperative continuous variables were analyzed by Wilcoxons signed rank test on the differences for nonnormally distributed variables and by a paired Students t-test for normally distributed variables. Logistic regression, adjusting for side effects, was used to determine the effect of a set of continuous variables on a dichotomous variable.
Demographic data for the 60 subjects entered into the study appear in Table 1. A total of 85% (51 of 60) of the subjects had preoperative recordings, and 93% (56 of 60) had postoperative recordings longer than 2 h. Eleven subjects did not have both pre- and postoperative recordings. Before surgery, the median (25%75%) total recording time and quiet time for the Remifentanil group were 175 min (113545 min) and 119 min (43484 min), and for the Halothane group, 178 min (124259 min) and 129 min (85209 min), respectively. After surgery, median total recording time and quiet time for the Remifentanil group were 705 min (614712 min) and 537 min (450572 min), and for the Halothane group these values were 707 min (429711 min) and 507 (354590 min), respectively. The total recording time and quiet time did not differ significantly between treatment groups.
Table 2 summarizes the incidence of abnormal pneumocardiograms in subjects with both pre- and postoperative recordings. Overall there were no statistically significant differences in the incidence of abnormal preoperative studies between remifentanil and halothane (9 of 32 vs 5 of 19, P = 0.58) or of abnormal postoperative studies between remifentanil and halothane (4 of 36 vs 5 of 20, P = 0.16). Of the subjects with both pre- and postoperative pneumocardiograms (both treatment groups pooled), 9 (64%) of the 14 subjects with abnormal preoperative pneumocardiograms had normal postoperative pneumocardiograms, and 3 (9%) of the 35 subjects with normal preoperative pneumocardiograms had abnormal pneumocardiograms after surgery. However, the tendency to change from abnormal to normal was not statistically different from the tendency to change from normal to abnormal (P = 0.146 by the exact McNemars test).
Of the five subjects whose results remained abnormal in the postoperative period, the ages ranged from 2 to 6 wk, and all had very abnormal postoperative studies (apnea indexes >10 and multiple prolonged apneic episodes). Three of these subjects were recalled after discharge (pneumocardiograms were read 1 to 2 wk after discharge) for a follow-up pneumocardiogram because the apneic episodes continued through the entire postoperative recording (720 min). These three follow-up studies at several weeks were normal. In the other two subjects, the pneumocardiograms normalized in the last few hours after surgery, and follow-up studies were unnecessary. Of the three subjects who were normal before surgery and abnormal after surgery, two were abnormal on the basis of a single prolonged apneic episode, one in the 0- to 59-min epoch and the other in the 540- to 720-min epoch. The third subject had, at the 0- to 59-min epoch, three prolonged apneic episodes and an abnormal apnea index (14.6), subsequently followed by a normal study to 720 min. Seven subjects (five remifentanil subjects versus two halothane subjects) received single-dose morphine 0.1 mg/kg IV in the postoperative period; none of these subjects had abnormal postoperative studies. Five subjects had surgical complications (three subjects required reoperation for rebleeding or perforation, one subject had a peritoneal leak, and one subject had intraoperative serosal tears requiring a prolonged closure); none of these subjects had an abnormal postoperative study. Table 3 summarizes the pneumocardiogram data for all subjects by study group. There were no statistically significant differences between remifentanil and halothane with respect to preoperative versus postoperative changes in any of the variables studied. Apnea index does not change over time in either the pre- or postoperative setting (Fig. 1). However, combining both study groups, the apnea index, frequency of apnea, frequency of central apnea, and frequency of apnea with desaturation overall were all less during postoperative studies compared with preoperative study (P < 0.01). Before surgery, apnea indexes more than 5 were noted in 14 subjects and after surgery in 7 subjects. Before surgery, 17 epochs in eight subjects had an apnea index more than 10 (maximum index, 28.2); after surgery this occurred in 4 epochs in three subjects (maximum index, 20.3). The maximum frequency of apneic events was 90 events per hour before surgery and 46 events per hour after surgery. Before surgery, seven subjects (14%) had prolonged apnea (number of events, 114); after surgery, eight subjects (14%) had this (number of events, 114). Central apnea accounted for the majority of apneic events in both the pre- and postoperative period, with no difference between drug study groups.
Ninety percent of subjects had at least one episode of desaturation (SpO2 <90%) before surgery, and 43% had this occur after surgery. On occasion, the desaturation was severe (SpO2 <70%). Desaturation without apnea was an infrequent event before surgery (11 events in four subjects) and after surgery (4 events in one subject). There was a significantly lower mean respiratory rate (P = 0.03) before versus after surgery. None of the following variables were predictive of desaturation (SpO2 <90%) before or after surgery: serum chloride, serum bicarbonate, hematocrit, or age. None of the subjects with emesis had abnormal postoperative studies.
This study evaluated the breathing patterns of term infants undergoing pyloromyotomy after either halothane- or remifentanil-based anesthesia. Of note was the presence of apnea both before and after surgery and the finding that in both time periods, apnea was primarily central in origin. Although the incidence of pre- and postoperative apnea was the same for the two groups and the n of this study was small, the onset of new postoperative apnea (i.e., subjects with a normal preoperative pneumogram and an abnormal postoperative pneumogram) was observed only in subjects anesthetized with halothane. The role of opioids in the management of neonates and infants is controversial. In a study comparing the effects of age on postoperative apnea in subjects receiving a fentanyl-based anesthetic, Hertzka et al. (13) determined that infants three months of age and older were no more susceptible to apnea than older children and adults. However, in this study they did note one six-week-old, full-term infant who developed apnea. At present, no one has evaluated the postoperative respiratory depressant effects of opioids in children younger than two months of age in whom tracheal extubation after surgery was anticipated. Apneic events can follow surgery in young infants who were born preterm (14) and at term (68). Apnea has also been described as a postoperative complication of pyloromyotomy. Andropoulos et al. (5) described four infants who developed apnea after pyloromyotomy. Although apnea was determined by clinical observation and no details of the anesthetic were noted, one of the four infants sustained a life-threatening apneic episode requiring tracheal intubation and cardiopulmonary resuscitation. In a prospective evaluation of 30 infants undergoing pyloromyotomy who were monitored with pre- and postoperative pneumograms, Chipps et al. (14) noted that postoperative apnea indices, respiratory distress index (RDI = the sum of central and obstructive apneas >10 seconds per hour), and arterial saturation were improved compared with the preoperative value. In this study, they concluded that pyloric stenosis, pyloromyotomy, and a standardized anesthetic (thiopental, nitrous oxide, and halothane) did not result in apnea. Although Chipps et al. demonstrated no new apnea, they reported an RDI at the upper end of normal (4.2; normal RDI, <5.0), suggesting that some subjects had abnormal apnea indices. Several mechanisms, which are not mutually exclusive of one another, may account for apnea before and after pyloromyotomy. The repeated vomiting associated with pyloric stenosis leads to a hypochloremic metabolic alkalosis and increased cerebral spinal fluid pH. The cerebral spinal fluid pH, a factor influencing respiratory drive, may remain increased for some time after correction of the serum electrolyte disturbance (5). Although we did not observe a relationship between desaturation and serum bicarbonate before surgery, this may be because of a discrepancy between brain and blood bicarbonate concentrations. Resolution of apnea from the preoperative to postoperative period may be coincident with the resolution of central alkalosis over time. Residual anesthetic drugs and postoperative discomfort may also play a role in the resolution of apnea after pyloromyotomy. In a small study of 20 infants, Wolf et al. (20) noted that clinically observed postoperative apnea occurred in 3 of 11 infants anesthetized with isoflurane but in none of the 9 infants anesthetized with desflurane. Subanesthetic concentrations of residual inhaled anesthetics are known to depress the hypoxic ventilatory drive. The difference between the Halothane and Remifentanil groups in postoperative apnea may result from subanesthetic concentrations of halothane depressing breathing. Opioids have also been shown to influence respiratory drive. Goldberg et al. (21) have shown in adults with an alfentanil-based anesthetic that arterial desaturation and respiratory drive depression occurred even when patients were easily arousable by verbal stimulation. However, remifentanil with its ultrashort pharmacokinetic profile and a lack of apnea in the postoperative period may be a suitable and predictable anesthetic for infants undergoing pyloromyotomy. Our study demonstrated not only that apnea did not increase after the use of intraoperative remifentanil, but also that it did not occur in subjects (n = 7) from either study group after the administration of single doses of IV morphine for operative analgesia. A possible criticism of our study is the clinical significance of the end point, apnea as detected by pneumograms. Although pneumograms can objectively quantitate breathing patterns, the relationship between breathing patterns and postoperative outcomes remains uncertain. Thus pneumogram abnormalities may be a surrogate end point. In our study all apnea episodes resolved spontaneously and no interventions were required to treat them. Although apneic episodes were associated with arterial desaturation, the clinical relevance of transient desaturations occurring over a few hours is unknown. It is also unclear what the normal background incidence of apnea is for infants of this age and whether the incidence of apnea noted for subjects in this study is truly abnormal. In fact, monitors for apnea, heart rate, and saturation are seldom used in any of the centers participating in this study, either before surgery or until discharge after surgery. Despite many clinical studies, the clinical implication of apnea in full-term infants and its relationship to sudden infant death syndrome remains uncertain. However, in former preterm infants, the clinical implication of apnea is more certain, because death and long-term disability have been reported (9,18,19) and have led to the recommendation to monitor former preterm infants in both hospital and home settings. The significance of postoperative apnea in term infants is unknown. In our study, the three subjects with grossly abnormal pneumograms at the time of discharge were normal in their follow-up evaluations three months later. Our findings may also be considered limited by the use of halothane as a maintenance anesthetic. Future studies will need to address the differences between remifentanil and newer anesthetics such as sevoflurane, desflurane, and propofol on postoperative apnea. In conclusion, pre- and postoperative apnea occurs in relation to general anesthesia in infants undergoing pyloromyotomy. Rarely, this apnea is severe. The incidence of apnea decreases from the pre- to postoperative period. A remifentanil-based anesthetic provides a reasonable alternative to an inhaled anesthetic technique, and despite remifentanils being an opioid, it does not increase the incidence of apnea.
Supported by Glaxo Wellcome, Inc., Research Triangle Park, NC.
Presented at the meetings of the Society for Pediatric Anesthesia, Ft. Meyers, FL, February 26, 2000, the American Society of Anesthesiologists, San Francisco, CA, October 17, 2000, and the American College of Clinical Pharmacy, Tampa, FL, October 22, 2001.
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|