A national survey of ambulatory surgery done in 2006 reported 3.3 million ambulatory procedures in children younger than 15 years in the United States out of 53.3 million ambulatory procedures for all ages.1 The scope of pediatric ambulatory surgery continues to expand, and appropriate patient selection is crucial for ensuring safe outcomes. A major area of concern for pediatric ambulatory anesthesia is reducing the incidence of perioperative respiratory adverse events (PRAEs), but tools to identify the patients at an increased risk are lacking.2
PRAEs are among the most common critical incidents that can occur in children who have undergone anesthesia.3 The incidence of PRAE varies depending on the perioperative adverse events studied3,4 and remains higher in children even when there is no active respiratory infection.5–7 PRAEs in pediatric anesthesia are associated with higher odds (2.5 times) of increased hospital stay after outpatient surgery, 2 times longer hospitalization, 30% higher excess hospital costs, and 58% higher indirect costs among outpatients.8 Laryngospasm, hypoxemia, and bronchospasm were the most common adverse respiratory events reported as anesthetic-related critical events. Although PRAE can be easily detected and treated, they can progress to life-threatening complications.9 Previous studies have identified the risk factors of PRAE: younger age (with an 8% reduction in the incidence in PRAE for every additional year of age), type of surgery, use of desflurane, sleep-disordered breathing, emergency surgery, and anesthetic care without a pediatric anesthesiologist3,6,7,9–11; however, the risk factors identified in these studies are not consistent with modern pediatric ambulatory anesthesia practice in many centers.7,10 For a child undergoing ambulatory anesthesia, identifying the odds of PRAE occurring is an important concern for the anesthesiologist. Individual patient risk stratification is crucial in deciding whether to postpone the procedures requiring anesthesia or to perform an intervention before administering anesthesia.2 The risk assessment tool should be easy to use and practical because of the pace and the limited time available for preoperative evaluation and assessment of ambulatory anesthesia pediatric patients.9
The primary aim of the study was to develop and validate a risk prediction tool for the occurrence of PRAE from the onset of anesthesia induction until discharge from the postanesthesia care unit (PACU) in children younger than 18 years undergoing elective ambulatory anesthesia for surgery and radiology. This risk prediction tool was validated with an independent internal cohort of patients from the study population. The incidence of PRAE was studied.
IRB approval was obtained for this study. This was a retrospective cohort analysis of a prospectively collected quality improvement (QI) database. Data on patients older than 18 years were excluded from this study. Anesthetic management was at the discretion of the attending anesthesiologist. Data entered into the QI database were derived from the electronic medical records and from the observational information of events in the PACU from 2007 to 2012. The results presented in this article are derived from the analysis of the aforementioned QI database.
The QI project database was designed as a prospective, single-center registry initiated in 2007 by an improvement team to document the incidence of PRAE in children undergoing noncardiac ambulatory anesthesia procedures (surgery or radiology). The QI team consisted of quality and safety leaders in the perioperative area, risk management representatives, project managers, decision analyst, and QI consultants.
Data collection for the QI project did not require informed written patient consent because there was no direct contact with patients or families, data collection involved observation of routine care, and the data were already available in the patient’s medical records.12 The QI team acquired data prospectively on children of all ages presenting for ambulatory anesthesia for surgery or radiology for the years 2007 to 2012 (each quarter for 2 weeks in the surgical population [total of 8 weeks of data per year] and for 4 weeks in the radiology population [total of 16 weeks of data per year]). This method of data collection ensured a sample of at least 10% that was representative of our patient population receiving ambulatory anesthesia for surgical or radiological procedures. The data collection sheet was standardized and approved by the hospital’s perioperative safety committee. A pilot study using the datasheet was conducted with a small group of trained perioperative registered nurses and certified registered nurse anesthetists who collected the data. Certified registered nurse anesthetists reviewed these sheets after completion by the nurses. Subsequently, a clinical research coordinator entered the data into an existing QI database. A limited number of predictor data were collected.
Anesthetic management included induction with sevoflurane with oxygen and nitrous oxide either in an induction room or in the operating/procedure room. An IV line was placed when the child was deeply anesthetized. Further administration of medications was at the discretion of the anesthesiologist. A case with active upper respiratory infection was cancelled if the child was febrile, chest auscultation revealed adventitious lung sounds, or the room air oxygen saturation was <94%. Data were collected on the candidate variables that were determined to be clinically significant predictors of PRAE: age, sex, ASA physical status, morbid obesity, preexisting pulmonary disorder, preexisting neurologic disorder, and location of ambulatory anesthesia (surgery versus radiology). A wide range of related comorbidities (symptoms and diagnoses) were classified into similar categories: preexisting pulmonary disorder (Appendix 1) and preexisting neurologic disorder (Appendix 2). Morbid obesity was defined based on the Centers for Disease Control’s defined body mass index for age growth charts for girls and boys. The body mass index (weight in kilograms/height in square meter) was derived from the collected height and weight variables from the database.
The primary aim of the study was to develop and validate a risk prediction tool for the occurrence of PRAE from the onset of anesthesia induction until discharge from the PACU in children younger than 18 years undergoing elective ambulatory anesthesia for surgery and radiology. This was measured as a composite score of occurrence of any 1 or a combination of the following intraoperative events: laryngospasm or bronchospasm; or postoperative events: apnea/hypopnea, bronchospasm, laryngospasm, or prolonged oxygen requirement. Apnea/hypopnea was defined by the need for bag mask ventilation; bronchospasm by the use of albuterol; laryngospasm by the requirement for positive pressure ventilation of >20 cm H2O or administration of succinylcholine; and oxygen requirement by a continued oxygen need 2 hours postoperatively to maintain Spo2 > 92%. The occurrences of adverse events were calculated only once (irrespective of the number of adverse event an individual patient had) while deriving the composite PRAE for each patient. The above variables (bag/mask, positive pressure ventilation, and succinylcholine) were documented only when they were used to manage adverse events and not for providing routine care (e.g., bag mask ventilation during induction, succinylcholine for securing airway for surgery or radiology).
Data were collected on 20,514 patients during the study period from the aforementioned QI project database (Fig. 1). The present analysis is restricted to children younger than 18 years (n = 19,059). To reliably build and validate the prediction tool, we used a split-sample analytical method for obtaining the derivation and the validation samples.13 Based on this method, study participants were assigned to either derivation cohort or validation cohort based on the year when the child underwent ambulatory anesthesia, i.e., for surgery or radiology. The study patients were assigned to a derivation cohort (n = 8904) if they had ambulatory anesthesia between the years 2007 and 2009 or to a validation cohort (n = 10,155) if they had ambulatory anesthesia between the years 2010 and 2012. Demographic and clinical characteristics of the study participants in both cohorts were compared using t tests for continuous variables and χ2 test for discrete variables as appropriate.
Derivation of Risk Prediction Tool
The risk prediction tool was developed based on data from the derivation cohort. All covariates, associated with the composite event at an a priori P < 0.10 in bivariate analysis were included in the multivariable logistic regression model. A backward stepwise elimination procedure was used to identify variables that were significantly associated (P < 0.05) with an increased risk of a composite event in the multivariable tool. Biologically relevant 2-way interaction terms were explored in the final model, and no significant interactions were found. A risk score for each covariate was given based on the β regression coefficients of the significant covariates from the final multivariable model. By using these risk scores, we subsequently developed a composite risk score by summing the individual risk scores from each covariate.
Validation of Risk Prediction Tool
The performance of the risk prediction tool developed from the derivation cohort was assessed separately in an independent validation cohort. The predictive accuracy of the derivation and validation tool was assessed using tests for discrimination and calibration. Model discrimination performance was evaluated using the standard measures of sensitivity, specificity, and predictive values based on a cutoff point for the risk score. The risk score was derived for the composite adverse events. In addition, overall model discrimination was assessed using the C-statistic representing a model’s ability to correctly distinguish the subjects having the event from the subjects not having the event. Model calibration was assessed by Hosmer-Lemeshow statistic that predicts how well the model fits the events in the data based on the agreement between the actual observed and the predicted probabilities of events. Statistical analysis was conducted using SAS version 9.4 (SAS Institute Inc., Cary, NC).
A total of 19,059 patients split into 2 cohorts (47% in the derivation cohort and 53% in the validation cohort) composed our study population (Fig. 1). The overall incidence of composite PRAE in children younger than 18 years was 2.8% (n = 520). The incidence of PRAE was 3.6% in neonates (≤28 days), 3.9% in infants (>28 days to ≤1 year), 3.2% in toddlers (>1 year to ≤3 years), 2.3% in children (>3 years to ≤13 years), and 2.7% in teenagers (>13 years to ≤18 years). The demographics and other clinical data on all patients are provided in Table 1. The occurrence of PRAE in the overall population and the derivation and validation cohorts are provided in Table 2. Age was entered into the regression model as a binary variable with a cutoff of 3 years. This age cutoff was derived from the receiver operating characteristic (ROC) curve (Supplemental Digital Contents 1 and 2, Supplemental Figs. 1 and 2, http://links.lww.com/AA/B374 and http://links.lww.com/AA/B375, respectively) and from a clinical importance standpoint from previous studies.
Development of Risk Tool
In the derivation cohort, 3.9% (n = 335) of the patients developed a composite PRAE. The results from the univariate analysis are provided in Table 3. From multivariable analysis (Table 4), the candidate variables significantly associated with composite PRAE were age ≤ 3 years (versus >3 years), ASA physical status II and III (versus ASA I), morbid obesity, preexisting pulmonary disorder, and anesthesia for surgery (versus radiology). We derived the risk scores based on the β regression coefficients of the covariates in the final multivariable model. We assigned a score of 0.5 or 1 or 1.5 to each covariate depending on the proximity of their β regression coefficients to the nearest whole number. Based on this, a risk score of 0.5 was assigned to age ≤ 3 years (β = 0.55), 0.5 to ASA physical status II (β = 0.50), 1 to ASA physical status III (β = 0.79), 1 to preexisting pulmonary disorder (β = 1.01), 1 to obesity (β = 0.95), and 1.5 to type of procedure (β = 1.37). We then multiplied each score by 2 to make the risk score an integer, which will help for easy interpretation by clinicians. Thus, age ≤ 3 years got a score of 1 (0.5 × 2); ASA physical status II got a score of 1 (0.5 × 2); ASA physical status III, preexisting pulmonary disease, and obesity each got a score of 2 (1 × 2); and finally type of procedure got a score of 3 (1.5 × 2). The individual covariate scores were summed to obtain an individual risk score ranging from 0 to 11. We plotted the risk scores using ROC curve and obtained an optimal cutoff point of 4 based on the minimum distance between sensitivity and specificity of the risk scores in the ROC curve (Fig. 2). Accordingly we chose a risk score of ≥4/11 to define high-risk categories, 1 to 3/11 to define intermediate-risk categories, and 0/11 to define low-risk categories.
Accuracy and Validation of Risk Tool
In the derivation cohort, 1.8% (n = 185) of patients developed a composite PRAE. The predictive ability of the risk tool was tested in the independent validation cohort. Based on the cutoff risk score of ≥4, the discriminant performance (±SE) of the risk tool had a negative predictive value of 98.2 ± 0.002, positive predictive value of 5.8 ± 0.003, sensitivity of 77.6 ± 0.020, and specificity of 49.2 ± 0.005 in the derivation cohort. Likewise, the discriminant performance of the tool in the validation cohort had 99.1 ± 0.001 negative predictive value, 2.8 ± 0.002 positive predictive value, 76.2 ± 0.030 sensitivity, and 49.7 ± 0.005 specificity (Supplemental Digital Contents 3 and 4, Supplemental Tables 1 and 2 and Supplemental Fig. 3, http://links.lww.com/AA/B376 and http://links.lww.com/AA/B377, respectively). The model C-statistic and the corresponding SE for the derivation and validation cohorts was 0.64 ± 0.01 and 0.63 ± 0.02, respectively. The Hosmer-Lemeshow goodness-of-fit tests for derivation cohort (P = 0.99) and validation cohort (P = 1.0) were nonsignificant, indicating that the 2 models had a good fit.
Pediatric ambulatory surgery was first described in 1845.14 Improvements in techniques markedly increased the range of ambulatory anesthesia in both volume and complexity in and out of the operating room.15 In children, PRAEs are the most common adverse events in patients undergoing anesthesia.3,4,16,17 Compared with adult closed claims, pediatric closed claims had a larger prevalence of PRAE, and postoperative respiratory events were a major cause of deaths, severe injuries, and successful monetary claims.18–20
The incidence of PRAE in previous studies was 13% to 15%.7,10 In our study, it was 2.8% perhaps reflecting the change in complications with current anesthetic practice and inclusion of patients having only ambulatory anesthesia, although the definition of PRAE must be reviewed before making a head-to-head comparison of each study.7,21,22 The differences are related to the definitions, medications, or airway devices used to manage adverse events. Various risk factors have been associated with PRAE.7,9,11,16,23 However, our study provides a comprehensive risk score assessment for each patient to assess the risk of PRAE. To prevent and minimize PRAE, preventive strategies have to be implemented early on.22 A previous study used a multivariable model to identify the risk factors for PRAE.10 This study included subjects whose procedures might have been cancelled in other institutions, and it was unclear whether there was preoperative optimization of asthma and other respiratory diseases and whether inclusion of certain unmeasured comorbidities increased the risk of the population as a whole because of confounding.21 In our study, we described a comprehensive risk prediction tool that is easy to use in the limited time frame available for preoperative assessment for ambulatory anesthesia patients. The risk prediction tool in our study is specifically applicable to patients undergoing ambulatory anesthesia for elective noncardiac procedures both in and out of the operating room. The prediction tool is also validated with a different set of patients enrolled in our hospital in different years. When the risk score is <4, the chance of having a complication is <1%. Combining preoperative and intraoperative factors to predict postoperative outcome may facilitate the development of a prospective tool to direct and triage care at the time of scheduling and could allow a predischarge stop and check. There is potential for reduced resource utilization when patients who are candidates for a short PACU stay followed by discharge can be identified a priori.
We studied composite PRAEs because they are patient-oriented global outcomes. Composite PRAEs include both intraoperative and postoperative respiratory events. From a practical clinical viewpoint, we believe that predicting intraoperative serious adverse respiratory events is at least as important as predicting additional postoperative events. Serious intraoperative adverse respiratory events may necessitate additional anesthetic precautions, premature case termination, and may interfere with the surgical operative procedure. They may also lead to poor postoperative outcomes/complications and the necessity of prolonged postoperative monitoring and possibly overnight admission. PRAEs are associated with longer hospitalization, increased hospital costs, and increased indirect costs among outpatients.8 Although a score of ≥4/11 may appear low for the high-risk category, this score is derived from ROC analysis and will identify the children at risk for PRAE.13,24
Preoperative predictor variables were chosen based on the individual risk factors described in the literature and from our clinical experience.6,7,10 Common preexisting airway problems encountered in practice are asthma and upper respiratory infection. The presence of airway sensitivity, eczema, and family history of airway disorders are risk factors for PRAE.10 In our study, we included patients with a wide range of preexisting pulmonary disorders to enable easy risk prediction for clinicians; this variable was an independent predictor for increased risk for composite PRAE. Most pediatric anesthesiologists would cancel and reschedule a surgery if a patient had a temperature of >38°C, wheezing, and/or preoperative hypoxemia, because the evidence shows that these children are at higher risk for postoperative complications.21,25 In the presence of active upper respiratory infection, this cancellation policy is consistent with our practice.
Previous studies in children have shown the incidence of PRAE increases with decreasing age.6,7,10,16 Our study results are consistent with this and show an increase in PRAE in children aged 3 years or younger compared with children older than 3 years. ASA score has been shown to be associated with postoperative complications in adults and children.26–28 Variation is reported in the correct classification of ASA scores.29 In spite of this limitation, the ASA score had moderate interrater reliability and validity based on its correlation to postoperative outcomes.30 Our multivariable risk prediction tool demonstrated that a higher ASA physical status (II or III vs I) was strongly associated with higher odds of PRAE in children. Obesity increases the incidence of perioperative adverse events. Both the number and the severity of the comorbidities (i.e., asthma, sleep apnea) that contribute to respiratory adverse events are increased in obese patients.31,32 The prevalence of morbid obesity in our study cohort was 1.6%. Results from our study are consistent with previous studies showing that obesity is a predictor for apnea/hypopnea, a prolonged oxygen requirement, and composite PRAE.23 Surgery had more PRAEs compared with radiology, possibly because of differences in requirement of opioids and airway management between both groups.
Patients with neuromuscular disorders have a higher risk of PRAE particularly when opioids are administered.22,33 Patients with neuromuscular diseases are dependent on a wakeful state to maintain adequate ventilation.34 In our study, preexisting neuromuscular disease/hypotonia was not an independent risk factor for PRAE. This is possibly related to our particular style of practice that incorporates preoperative optimization, awake extubation, and early awakening in the PACU. Sex was not an independent risk factor in a previous study,7 and our results are in concordance with those findings.
Our statistical methodology deserves discussion. We used a split-sample technique where the data were split into 2 samples based on the year of ambulatory anesthesia.13 This provided a different cohort for validation than the cohort used for tool development. Our large derivation cohort of 8904 patients ensured adequate model precision and goodness of fit. With an overall PRAE of 2.8% and 7 candidate predictor variables, allocating 15 cases for each predictor variable, as has been suggested to be best practice, would have required 3752 patients.35 Nonetheless, our data are derived from a single-center registry and lack external validation.
A limitation of our study was lack of available data on surgical procedures and other intraoperative variables. Airway surgeries and longer duration of anesthesia are known risk factors for PRAE.7,36 Hence, this study can only partially define the predictive factors, and the risk score has to be interpreted with this background. Second, follow-up beyond the PACU was not performed, and the data on serious perioperative morbidity and/or mortality were not available. Even so, PRAE can be a key driver for the occurrence of respiratory complication after discharge from the PACU. Therefore, risk assessment is the first step in the mitigation of adverse events. Third, a subgroup of patients with pulmonary disorders and morbid obesity may have also had sleep apnea. A previous study has identified sleep-disordered breathing as a risk factor for PRAE.9 Fourth, the model developed has high negative predictive values, but the sensitivity and specificity values are modest. Finally, there is no universally accepted definition of PRAE; hence, the generalizability of PRAE should be interpreted within the context of variables used in the study.
The risk tool developed in this study is easy to use and can help identify children at risk for PRAE. The risk prediction tool may be adapted clinically and used to design targeted intervention22 in children at high risk for PRAE.
Appendix 1. Preexisting Pulmonary Disease
Appendix 2. Preexisting Neurologic Disease
Name: Rajeev Subramanyam, MBBS, DNB, MNAMS, MD, MS.
Contribution: This author helped with study design, data analysis, and manuscript preparation.
Attestation: Rajeev Subramanyam approved the final manuscript, attests to the integrity of the original data and the analysis reported in this manuscript, and is the archival author.
Name: Samrat Yeramaneni, MBBS, PhD.
Contribution: This author helped with data analysis and manuscript preparation.
Attestation: Samrat Yeramaneni approved the final manuscript and attests to the integrity of the original data and the analysis reported in this manuscript.
Name: Mohamed Monir Hossain, PhD.
Contribution: This author helped with data analysis and manuscript preparation.
Attestation: Mohamed Monir Hossain approved the final manuscript and attests to the integrity of the original data and the analysis reported in this manuscript.
Name: Amy M. Anneken, MS.
Contribution: This author helped with data collection and manuscript preparation.
Attestation: Amy M. Anneken approved the final manuscript and attests to the integrity of the original data and the analysis reported in this manuscript.
Name: Anna M. Varughese, MD, MPH.
Contribution: This author helped with conduct of the study, data collection, and manuscript preparation.
Attestation: Anna M. Varughese approved the final manuscript and attests to the integrity of the original data and the analysis reported in this manuscript.
This manuscript was handled by: James A. DiNardo, MD.
1. Cullen KA, Hall MJ, Golosinskiy A. Ambulatory surgery in the United States, 2006. Natl Health Stat Report. 2009;11:1–25.
2. August DA, Everett LL. Pediatric ambulatory anesthesia. Anesthesiol Clin. 2014;32:411–29.
3. de Graaff JC, Sarfo MC, van Wolfswinkel L, van der Werff DB, Schouten AN. Anesthesia-related critical incidents in the perioperative period in children; a proposal for an anesthesia-related reporting system for critical incidents in children. Paediatr Anaesth. 2015;25:621–9.
4. Kurth CD, Tyler D, Heitmiller E, Tosone SR, Martin L, Deshpande JK. National pediatric anesthesia safety quality improvement program in the United States. Anesth Analg. 2014;119:112–21.
5. Bhananker SM, Ramamoorthy C, Geiduschek JM, Posner KL, Domino KB, Haberkern CM, Campos JS, Morray JP. Anesthesia-related cardiac arrest in children: update from the Pediatric Perioperative Cardiac Arrest Registry. Anesth Analg. 2007;105:344–50.
6. Murat I, Constant I, Maud’huy H. Perioperative anaesthetic morbidity in children: a database of 24,165 anaesthetics over a 30-month period. Paediatr Anaesth. 2004;14:158–66.
7. Mamie C, Habre W, Delhumeau C, Argiroffo CB, Morabia A. Incidence and risk factors of perioperative respiratory adverse events in children undergoing elective surgery. Paediatr Anaesth. 2004;14:218–24.
8. Oofuvong M, Geater AF, Chongsuvivatwong V, Chanchayanon T, Sriyanaluk B, Saefung B, Nuanjun K. Excess costs and length of hospital stay attributable to perioperative respiratory events in children. Anesth Analg. 2015;120:411–9.
9. Tait AR, Voepel-Lewis T, Christensen R, O’Brien LM. The STBUR questionnaire for predicting perioperative respiratory adverse events in children at risk for sleep-disordered breathing. Paediatr Anaesth. 2013;23:510–6.
10. von Ungern-Sternberg BS, Boda K, Chambers NA, Rebmann C, Johnson C, Sly PD, Habre W. Risk assessment for respiratory complications in paediatric anaesthesia: a prospective cohort study. Lancet. 2010;376:773–83.
11. Oofuvong M, Geater AF, Chongsuvivatwong V, Pattaravit N, Nuanjun K. Risk over time and risk factors of intraoperative respiratory events: a historical cohort study of 14,153 children. BMC Anesthesiol. 2014;14:13.
12. Varughese AM, Hagerman N, Townsend ME. Using quality improvement methods to optimize resources and maximize productivity in an anesthesia screening and consultation clinic. Paediatr Anaesth. 2013;23:597–606.
13. Khorana AA, Kuderer NM, Culakova E, Lyman GH, Francis CW. Development and validation of a predictive model for chemotherapy-associated thrombosis. Blood. 2008;111:4902–7.
14. Steward DJ. The origins and development of pediatric outpatient surgery. Paediatr Anaesth. 2014;24:632–4.
15. Coté CJ, Cravero J. Brave new world: do we need it, do we want it, can we afford it? Paediatr Anaesth. 2011;21:919–23.
16. Tay CL, Tan GM, Ng SB. Critical incidents in paediatric anaesthesia: an audit of 10 000 anaesthetics in Singapore. Paediatr Anaesth. 2001;11:711–8.
17. Lightdale JR, Mahoney LB, Fredette ME, Valim C, Wong S, DiNardo JA. Nurse reports of adverse events during sedation procedures at a pediatric hospital. J Perianesth Nurs. 2009;24:300–6.
18. Morray JP, Geiduschek JM, Caplan RA, Posner KL, Gild WM, Cheney FW. A comparison of pediatric and adult anesthesia closed malpractice claims. Anesthesiology. 1993;78:461–7.
19. Morris LG, Lieberman SM, Reitzen SD, Edelstein DR, Ziff DJ, Katz A, Komisar A. Characteristics and outcomes of malpractice claims after tonsillectomy. Otolaryngol Head Neck Surg. 2008;138:315–20.
20. Subramanyam R, Chidambaran V, Ding L, Myer CM III, Sadhasivam S. Anesthesia- and opioids-related malpractice claims following tonsillectomy in USA: LexisNexis claims database 1984-2012. Paediatr Anaesth. 2014;24:412–20.
21. Lerman J. Perioperative respiratory complications in children. Lancet. 2010;376:745–6.
22. von Ungern-Sternberg BS. Respiratory complications in the pediatric postanesthesia care unit. Anesthesiol Clin. 2014;32:45–61.
23. El-Metainy S, Ghoneim T, Aridae E, Abdel Wahab M. Incidence of perioperative adverse events in obese children undergoing elective general surgery. Br J Anaesth. 2011;106:359–63.
24. Hasbun R, Bijlsma M, Brouwer MC, Khoury N, Hadi CM, van der Ende A, Wootton SH, Salazar L, Hossain MM, Beilke M, van de Beek D. Risk score for identifying adults with CSF pleocytosis and negative CSF Gram stain at low risk for an urgent treatable cause. J Infect. 2013;67:102–10.
25. Tait AR, Malviya S. Anesthesia for the child with an upper respiratory tract infection: still a dilemma? Anesth Analg. 2005;100:59–65.
26. Wolters U, Wolf T, Stützer H, Schröder T. ASA classification and perioperative variables as predictors of postoperative outcome. Br J Anaesth. 1996;77:217–22.
27. Tiret L, Hatton F, Desmonts JM, Vourc’h G. Prediction of outcome of anaesthesia in patients over 40 years: a multifactorial risk index. Stat Med. 1988;7:947–54.
28. Malviya S, Voepel-Lewis T, Chiravuri SD, Gibbons K, Chimbira WT, Nafiu OO, Reynolds PI, Tait AR. Does an objective system-based approach improve assessment of perioperative risk in children? A preliminary evaluation of the ‘NARCO.’ Br J Anaesth. 2011;106:352–8.
29. Owens WD, Felts JA, Spitznagel EL Jr. ASA physical status classifications: a study of consistency of ratings. Anesthesiology. 1978;49:239–43.
30. Sankar A, Johnson SR, Beattie WS, Tait G, Wijeysundera DN. Reliability of the American Society of Anesthesiologists physical status scale in clinical practice. Br J Anaesth. 2014;113:424–32.
31. Dixon AE, Holguin F, Sood A, Salome CM, Pratley RE, Beuther DA, Celedón JC, Shore SA; American Thoracic Society Ad Hoc Subcommittee on Obesity and Lung Disease. An official American Thoracic Society Workshop report: obesity and asthma. Proc Am Thorac Soc. 2010;7:325–35.
32. Arens R, Muzumdar H. Childhood obesity and obstructive sleep apnea syndrome. J Appl Physiol (1985). 2010;108:436–44.
33. Racca F, Mongini T, Wolfler A, Vianello A, Cutrera R, Del Sorbo L, Capello EC, Gregoretti C, Massa R, De Luca D, Conti G, Tegazzin V, Toscano A, Ranieri VM. Recommendations for anesthesia and perioperative management of patients with neuromuscular disorders. Minerva Anestesiol. 2013;79:419–33.
34. Al-Ruzzeh KVRD, Hines RL, Marschall KE. Stoelting’s Anesthesia and Co-existing Disease. 20126th ed.. Philadelphia, PA: Saunders. 202.
35. Harrell F Jr. Regression Modeling Strategies: With Applications to Linear Models, Logistic Regression, and Survival Analysis. 2001. New York, NY: Springer-Verlag.
36. McAlister FA, Bertsch K, Man J, Bradley J, Jacka M. Incidence of and risk factors for pulmonary complications after nonthoracic surgery. Am J Respir Crit Care Med. 2005;171:514–7.