Infant apnea detection after herniorrhaphy

ERWORTH EINEMANN

Infant Apnea Detection after Heiniorrhaphy Charlotte Bell, MD,* Rick Dubose, MD,? John Seashore, MD,$ Robert Touloukian, MD,+ Carol Rosen, MD,$ Tae H. Oh, MD;’ Cindy W. Hughes, MD,# Shawn Mooney, RN,** Theresa 2. O’Connor, PhD”ft Department of Anesthesiology, New Haven, CT.

Yale University

School of Medicine,

Study Objective: To elucidate risk factors for apnea in preterm infants discharged from the hospital and in full-term healthy infants. To determine the efficacy of real-time

*Associate Professor of Anesthesiology, Yale tAttending Anesthesiologist, Sunwest Anesthesia, El Paso, TX #Professor of Surgery, Section of Pediatrics, Yale BAssistant Professor of Pediatrics, Yale ilProfessor of Anesthesiology, Yale #Assistant Professor of Anesthesiology, Albany Medical Center, Albany, NY **Clinical Research Nurse, Department of Anesthesiology, Yale TtStatistician, Department of Anesthesiology, Yale Address reprint requests to Dr. Bell at the Department of Anesthesiology, Yale University School of Medicine, 333 Cedar Street, Box 3333, New Haven, CT 06510, USA. Supported in part by a grant from Corometrics Medical Systems, Inc., Wallingford, CT. Presented in part at the Annual Meeting of the American Society of Anesthesiologists, San Francisco, October 27-31, 1991. Received for publication April 22, 1994; revised manuscript accepted for publication June 27, 1994.

cardiopulmonary monitoring versus computerized storage and retrieval for infants at risk. Study Design: Prospective study. Setting: Operating roomsand pediatric patient care units of university medical center. Patients: 27 preterm infants and 20 full-term infants no more than 60 weeks’postconceptional age, who were admitted for elective herniorrhaphy. Interventions: Infants were monitored before and after herniorrhaphy with general anesthesia using an infant apnea impedance monitor, pulse oximetry, and nursing observation. Measurements and Main Results: Demographic information and medical history were correlated with postoperative apnea. The sensitivity and specificity of nursing observation and oximetry were compared with computerized apnea monitors. Five patients (11 %, four preterm, one full-term) were apneic postoperatively as recorded by computerized pneumocardiography. Previous apnea history, gestational age at birth, and postconceptional age at operation positively correlated with postoperative apnea. Nursing observationfailed to detect4 of 5 patients with documentedapnea (sensitivity 20%, positive predictive value 50%). Pulse oximetryfailed to detect3 of 5 patients with apnea (sensitivity 40%, positive predictive value 66%). Conclusions: Although it is easier to predict postoperative respiratory dysfunction in previously sick or very young infants, absolute predictability for all neonates remains elusive. Clinical monitors with both storage and retrieval capabilities and real-time monitoring increase our ability to detectsignificant events in children at risk for apnea after herniorrhaphy. Keywords:

Anesthesia, pediatric; complications: pneumocardiography, prematurity.

apnea,

Introduction Many authors have studied postoperative respiratory dysfunction after general anesthesia in neonates in an attempt to define risk factors present in the medical history. ‘~3 It is gener all y a g reed that younger, preterm infants comprise the group at greatest risk with an even higher incidence noted among

Journal of Clinical Anesthesia 7:219-223, 1995 0 1995 by Elsevier Science Inc. 655 Avenue of the Americas, New York, NY 10010

0952-8180/95/$10.00 SSDI 0952-8180(95)00001-X

Original Contributions

those children with preexisting respiratory dysfunction. However, because real predictability continues to elude neonatologists, surgeons, and anesthesiologists, most health care providers continue to use postconceptional or postnatal age as the final criteria for admission to the hospital instead of opting for discharge after herniorrhaphy. The recommended age at which susceptibility to postoperative apnea becomes sufficiently low that outpatient herniorrhaphy may be safely considered has varied from 40 to 60 weeks postconceptional age.1,3,6,’ How to effectively monitor infants admitted for postoperative cardiorespiratory observation is often as great a conundrum for caregivers as the decision to admit or discharge. This study was designed to answer three questions. First, can we characterize patterns of apnea among preterm infants discharged from the hospital and readmitted for inguinal herniorrhaphy? This population contrasts with a previous study detailing complications in infants having hernia repair prior to discharge from the Newborn Special Care Unit.5 Second, can we stratify risk or predict problems in the full-term population? And finally, what system of monitoring (nurse observation and real-time monitors vers’uscomputerized impedance pneumocardiography with data retrieval and storage) is most efficacious for detection of postoperative infant apnea?

diately after surgery until the next day [postoperative day 1 (POD l)]. A standardized data sheet was used to collect demographic data, medical history, and intraoperative data for correlation with postoperative events. All infants were anesthetized using a nitrous oxide/ oxygen mixture, halothane or isoflurane, and a nondepolarizing muscle relaxant (vecuronium). No patients received opioids. Prior to discharge, pneumocardiograms were retrieved from the impedance monitors and reviewed. Apnea was defined as cessation of breathing for greater than 15 seconds, bradycardia as heart rate (HR) less than 70, and desaturation as SpO, less than 90%. These limits were programmed into the impedance monitor and oximeters, with alarms sounding to alert the nurse-observer whenever these limits were reached. The sensitivity and specificity of apnea detection by nursing observation and oximetry were compared with computerized pneumocardiography with data storage and retrieval capabilities. Data were analyzed using chisquare, Student t-test, and Fisher’s Exact test. A p-value less than 0.05 was considered statistically significant.

Results Demographic data for all infants are listed in Table 1. Five patients (11%) sustained apnea and 42 patients did not. Periodic breathing was noted by monitor in 44 (94%) of 47 of patients. Four (80%) of five patients with apnea were formerly preterm. Three of the four patients were treated for respiratory events, two with increased ambient FIO, and one with reintubation and ventilatory assistance. The fifth apneic patient, who was full-term, had a brief episode of apnea in the recovery room 20 minutes after cessation of halothane and required no treatment. Apneic episodes occurred any time between 20 minutes and 72 hours postoperatively. Profiles of apneic patients are listed in Table 2. Factors that positively correlated with postoperative apnea included a previous history of apnea, gestational age at birth, and postconceptional age at operation (Table 3). A history of ventilator-y assistance (continuous positive airway pressure, Oz, intubation) correlated marginally with postoperative apnea (p = 0.059). Apnea was not recorded by monitor in any patient preoperatively. Other factors not associated with postoperative apnea are listed in Table 4.

Materials and Methods A total of 47 infants consecutively admitted for elective herniorrhaphy with postconceptional ages less than 60 weeks were prospectively studied for apnea or respiratory dysfunction after general anesthesia. Of these 47 infants, 27 were defined as preterm based on a gestational age of 37 weeks or less at birth; 20 infants were defined as full-term (gestational age greater than 37 weeks). All infants were studied using the Corometrics 500 E infant apnea impedance monitor with Event Link computerized data storage and retrieval system (Corometrics, Wallingford, CT), real time pulse oximetry (Nellcor N-200, Nellcor, Haywood, CA, or Ohmeda Biox 3700, Ohmeda, Louisville, CO), and nursing observation. Each child was monitored for at least four hours during the night prior to surgery (control period) and again imme-

Table 1. Demographic Data Full-term

Preterm

Gestational age at birth (wks) Postnatal age at operation (wks) Postconceptional age at operation (wks) Birth weight (g) Weight at operation (g) Note: Data are means f SD (range).

420

J.

Clin. Anesth., vol. 7, May 1995

32.1 10.3 42.4 1,740 3,864

+ + f f f -

3.34 3.26 2.45 612.6 942.9

(26-37)

(6-W (37-47) (820-3,180) (2,000-5,610)

39.5 k 1.57 6.7 46.2 3,192 4,541

f f + k

2.81 2.72 482.8 774.7

(38--14) (2-16) (42-54) (1,900-4,063) (3,000-5,800)

Infant apnea detection:Bell et al.

Table 2.

Profiles of Patients with Postoperative Apnea Patient No. 1

Size (g) Weight at operation Weight at birth Age (wks) PCA at operation PCA at birth Description of apnea Onset Diagnosed by

Last episode No. of events Treatment Cardiorespiratory history Preoperative apnea (xanthines) PDA Other respiratory/ neurologic RDSIBPD Hematocrit Maximum FIO, as neonate Intraoperative factors Maximum % halothane Muscle relaxation Total IV fluids/h

2

3

4

2,700 1,460

2,880 1,320

4,100 1,130

2,700 1,390

5,690 4,063

39 31

40 28

41 27

37 28

47 40

PACU pneumocardiogram

PACU nurse observation, pulse oximeter, and pneumocardiogram

PACU pneumocardiogram

<24 h >lO none

PACU Desaturation with feedings thought to be monitor malfunction and pneumocardiogram 36 h >lO 0, HB 35% x 24 h

25 h 2 O2 mask

several days >lO reintubation

PACU 1 none

+

+(+)

+

+(+)

-

asthma

+ -

+ IVH

-

-

28% 30% HB*

+

+

2+8%

30 room air

28 unknown

2%

2%

vecuronium 27 ml/kg

vecuronium 30 ml/kg

15 h Pneumocardiogram

100%

1.5%

2%

vecuronium 13 ml/kg

vecuronium 33 ml/kg

PACU = post anesthesia care unit; HB = headbox; PDA = patent ductus arteriosus; IVH = intraventricular respiratory distress syndrome/bronchopulmonary dysplasia.

Nursing observation failed to detect 4 of 5 patients with apnea documented by computerized pneumocardiography. Pulse oximetry failed to detect apnea in 3 of 5 patients. Both techniques indicated one false positive. In one case, a patient’s hospital stay was extended because of “monitor alarms” noted by the nurse observer without visualization of apnea or computer documentaTable 3.

5

36% room air

2% vecuronium 14 ml/kg hemorrhage; RDS/BPD =

tion. The second false positive was a “desaturation” display by pulse oximetry due to patient movement. Sensitivity and specificity for nursing assessment and pulse oximetry as compared with computerized apnea monitoring is listed in Table 5. Both pulse oximetry and nursing assessment were successful in diagnosing the single patient requiring treatment (reintubation and postoperative ventilatory support).

Factors Associated with Apnea ($I < 0.05)

History of preoperative apnea Mean gestational age (birth) (wks) Mean postconceptional age (wks) History of ventilatory assistance (CPAP, Oar intubation) f$ = 0.059 borderline)

Group A*

Group Bt

4 (80%) 30.8 f 5.4 40.8 2 3.8

11 (26%) 35.7 t 4.2 44.3 2 2.9

4 (80%)

14 (33%)

CPAP = continuous positive airway pressure. *Group A = patients with apnea, n = 5. tGroup B = patients without apnea, n = 42.

Discussion The results of this study are consistent with our previous report5 and that of others suggesting that younger, formerly preterm infants are at increased risk for postoperative apnea. These data also support previous reports that a history of apnea increases the risk of postoperative apnea in former preterm infants. ‘+’ In contrast to previous studies, we found no increased incidence of respiratory d sfunction relative to the weight at birth or operation, 2 percent of inhalation anesthetic,5 use of muscle relaxation,’ or percent hematocrit.8 J. Clin. Anesth., vol. 7, May 1995

221

Originul crmhibutio?ls Table 4.

Factors Not Associated with Apnea (p = NS)

Mean birth weight (g) Mean weight at operation (g) Mean hematocrit (%) ASA classification

History of RDS/BPD History of left to right shunt Apnea recorded by monitor preoperatively (night before surgery) Length of anesthesia (min) Mean volume of intraoperative intravenous fluids Mean % halothane Mean % isoflurane (n = 23) RDS/BPD = respiratory distress syndrome/bronchopulmonary

Table 5. Sensitivity, Specificity, and Positive Predictive Value for Nursing Assessment and Oximetry versus Computerized Pneumocardiography as the Gold Standard

242

Sensitivity

Specificity

Positive Predictive Value

20% 40%

98% 98%

50% 66%

J. Clin. Anesth., vol. 7, May 1995

Group B Patients without Apnea (II = 42)

1,873 -c 1,230 3,614 + 1,300 30 2 3.5 (3 of 5, <30%) I lof5 II 4of5

2,415 + 870 4,208 t 879 32 +- 3.9 (12 of 42, <30%) I 16of42 II 21 of 42 III 5of42 11 of42 6of42 Oof42 86k 17 20 f 7 1.8 k 0.6 1.4 k 0.6

3of5 2of5 Oof5 93 2 14 23 f 9 1.9 + 0.2 0.9 k 0.5

P 0.2 0.2 0.9 0.4 0.1 0.2 0.8 0.3 0.07 1.0

dysplasia.

The high incidence of periodic breathing noted in our study (94%) is consistent with previous descriptions of periodic breathing as a common respiratory pattern, particularly in preterm infants, without pathologic significance or correlation to central apnea.g It is also of interest that admission the night prior to surgery for monitoring failed to predict postoperative risk. Due to lack of benefit from preoperative monitoring and in support of insurance companies’ refusal to reimburse for preoperative admission, infants are now routinely admitted the same day of surgery regardless of risk stratification. It is not sufficient only to recognize and admit infants at risk for apnea. We must be confident of our ability to accurately diagnose apnea in the population we have targeted, as well as initiate clinically relevant treatment. Only 20% of patients with apnea confirmed by pneumocardiography were recognized by nurse observation. Muttit et al.” reported similar results with nursing detection of apnea occurring in 54% of episodes (compared with computerized pneumocardiography). In their report, nursing accuracy improved with increased duration of apnea or greater decrease in HR. This study confirms our hypothesis that nursing observation and real-time monitors cannot recognize events whose onset and resolution may occur within sec-

Nurse assessment Pulse oximetry

Group A Patients with Apnea (n = 5)

onds. The sound of apnea and/or pulse oximetry alarms may arouse the sleeping infant so that no clinical evidence of apnea is apparent to the observer responding to the alarm. Conversely, a squirming or crying infant will often cause a pulse oximeter to display 85% in error as a result of motion artifact.” Both of these situations may confound the observer’s ability to report an apneic event with accuracy. On the other hand, although computerized pneumocardiography has the ability to report apneic events with greater accuracy, it also confirms that most of these resolve without intervention and are only noted hours later on retrieval of stored information. These data are consistent with published reports that apnea itself is not necessarily associated with long-term sequelae,” and that more physiologic information is necessary to determine which events may be harmful.” The addition of nasal thermistors not used in this study would add information on apneic episodes since the incidence of obstructive apnea is reportedly 6% l3 and may be associated with a greater decrease in HR.14 It is noteworthy that only one full-term infant had monitor-diagnosed apnea. This brief episode occurred in the postanesthesia care unit 20 minutes after cessation of halothane. It was not recognized by direct nursing observation or pulse oximetry but was apparent on retrieval of computer data. No treatment was initiated and no further episodes occurred. This infant had no risk factors and the brief episode of apnea had no significant effect on an otherwise unremarkable recovery. In this situation-a full-term infant with no risk factorscomputerized pneumocardiography may well be “overmonitoring,” offering information that does not affect clinical outcome. Although our one case of apnea in a full-term infant was self-resolving, several cases of apnea requiring treatment in full-term infants have been reported. On careful retrospective examination, each of these reports of

Infant apnea detection: Bell et al.

respiratory dysfunction in full-term infants yields a notable risk factor. It is possible that although age is a helpful determinate of risk for preterm infants, more subtle risk factors are at lay in the full-term child. In 1988, Tetzlaff et al. p5reported clinically significant postoperative apnea in a 3.2 kg full-term infant who was 42 weeks postconceptional age at the time of surgery for congenital cataracts. A murmur consistent with a atent B ductus arteriosus was noted. Previous reports5*’ have noted that presence of patent ductus arteriosus at any time in the neonatal history increases the risk of apnea as much as the presence of respiratory distress syndrome. A similar report of postoperative apnea in a full-term infant by Noseworthy et al. ’ indicated that gentamicin was given along with pancuronium, possibly prolonging the effects of neuromuscular blockade. Cote and Kelly described a full-term infant with postoperative apneaI found retrospectively to have a markedly abnormal sleep study. Finally, Karayan et al.” reported significant postoperative apnea in a full-term infant “without known risk factors” who received a combination of general and regional anesthesia. In the preterm population, the combination of parenteral sedation and spinal anesthesia are known to increase the risk of apnea.*‘,*’ It is our contention that information regarding apnea after surgery with general anesthesia in full-term infants is limited. It seems likely from the above-mentioned reports and our data that apnea occurs very rarely in fullterm infants and is most likely associated with risk factors not yet well defined. A large multicenter prospective study will be necessary to clearly delineate a population at risk. Until that time, very young infants (less than two months of age) are generally not considered candidates for outpatient surgery. We should appreciate that age in and of itself may not be the most important risk factor, but rather a method by which other risk factors may be avoided. Regardless of how we catalogue risk factors for postoperative respiratory dysfunction, it is imperative that we have the ability to detect clinically significant events, In this study, only one patient had serious postoperative respiratory dysfunction necessitating reintubation, mechanical support, and intensive care monitoring. His past medical history contained many of the markers that should alert clinicians to the probability of this occurring, including prematurity (28 wks gestational age), previous apnea requiring xanthines, and respiratory distress syndrome. His operation occurred at less than 40 weeks postconceptional age with an operative weight of less than 3.0 kg and finally, postoperative respiratory dysfunction was obvious to caregivers by clinical observation and confirmed by all monitoring devices. Although these severe cases of respiratory dysfunction are relatively easy to predict and clinically diagnose, this study indicates that the majority of apneic or hypoxic events are not obvious. It is difficult to predict the clinical relevance of these events, which often resolve without treatment, usually requiring only routine care (such as postoperative 0,). Until we can determine unequivocally which children are at risk for clinicalIy sig-

nificant events, it remains prudent to monitor all infants with significant risk factors using both real-time monitors and those with storage and retrieval capabilities.

References 1. Liu LMP, Cot6 CJ, Goudsouzian NG, et al: Life-threatening apnea in infants recovering from anesthesia. Anesthesiology 1983;59:506-10. 2. Mayhew JF, Bourke DL, Guinee WS: Evaluation of the premature infant at risk for postoperative complications. Can J Anuesth 1987;34:627-3 1. 3. Kurth CD, Spitzer AR, Broennle AM, Downes JJ: Postoperative apnea in preterm infants. Anesthesiology 1987;66:483-8. 4. Welborn LG, Ramirez N, Oh TH, et al: Postanesthetic apnea and periodic breathing in infants. Anesthesiology 1986;65:65861. 5. Gollin G, Bell C, Dubose R, et al: Predictors of postoperative respiratory complications in premature infants after inguinal herniorrhaphy. J Pediatr Surg 1993;28:244-7. 6. Warner LO, Teitelbaum DH, Caniano DA, Vanik PE, Martin0 JD, Servick JD: Inguinal herniorrhaphy in young infants: perianesthetic complications and associated preanesthetic risk factors. J Clin Anesth 1992;4:455-61. 7. Malviya S, Swartz J, Lerman J: Are all preterm infants younger than 60 weeks postconceptual age at risk for postanesthetic apnea? Anesthesiology 1993;78: 1076-8 1. 8. Welborn LG, Hannallah RS, Luban NL, Fink R, Ruttimann UE: Anemia and postoperative apnea in former preterm infants. Anesthesiology 1991;74:1003-6. 9. Glotzbach SF, Baldwin RB, Lederer NE, Tansey PA, Ariagno RL: Periodic breathing in preterm infants: incidence and characteristics. Pediatrics 1989;84:785-92. 10. Muttitt SC, Finer NN, Tierney AJ, Rossmann J: Neonatal apnea: diagnosis by nurse versus computer. Pediatrics 1988;82: 713-20. 11. Polage JA: Pulse oximetry: technical aspects of machine design. Int Anesthesiol Clin 198’7;25: 137-53. 12. (Anonymous) National Institutes of Health consensus development conference on infantile apnea and home monitoring. Pediatrics 1987;79:292-9. 13. Kurth CD, LeBard SE: Association of postoperative apnea, airway obstruction, and hypoxemia in former premature infants. Anesthesiology 1991;75:22-6. 14. Henderson-Smart DJ, Butcher-Puech MC, Edwards DA: Incidence and mechanism of bradycardia during apnoea in preterm infants. Arch Dis Child 1986;61:227-32. 15. Tetzlaff JE, Annand DW, Pudimat MA, Nicodemus HF: Postoperative apnea in a full-term infant. Anesthesiology 1988;69: 426-8. 16. Uejima T: Postoperative apnea [Letter]. Anesthesiology 1989;70: 721. 17. Noseworthy J, Duran C, Khine HH: Postoperative apnea in a full-term infant [Letter]. Anesthesiology 1989;70:879-80. 18. Cot& CJ, Kelly DH: Postoperative apnea in a full-term infant with a demonstrable respiratory pattern abnormality. Anesthesiology 1990;72:559-61. 19. Karayan J, LaCoste L, Fusciardi J: Postoperative apnea in a full-term infant [Letter]. Anesthesiology 1991;75:375. 20. Cox RG, Goresky GV: Life-threatening apnea following spinal anesthesia in former premature infants. Anesthesiology 1990;73: 345-7. 21. Welborn LG, Rice LJ, Hannallah RS, Broadman LM, Ruttimann UE, Fink R: Postoperative apnea in former preterm infants: prospective comparison of spinal and general anesthesia. Anesthesiology 1990;72:838-42.

J. Clin. Anesth., vol. 7, May 1995

223