It is well documented that many children report fear and become distressed with anesthesia induction by mask,1,2 but little is known about the way in which this distress is expressed. Previous research has demonstrated that children show a range of distress behaviors in response to medical procedures ranging from verbalizations of fear, to vocal protests, and attempts to escape.3 Understanding the different types of behavior children exhibit has important implications for identifying and managing children’s distress at anesthesia induction. For example, some children may display distress by crying but may be compliant with the procedure, whereas others may physically or verbally resist anesthesiologists’ attempts to place the mask. Although both of these types of distress behaviors are important, their implications and management may be very different. In addition to the topography of distress, little is known about children’s distress across time.
Despite the preponderance of studies that have examined children’s distress in some way, no study has examined children’s nondistress behavior during anesthesia induction. In a parallel body of literature on children’s procedural pain, attention has been devoted to both distress and nondistress behaviors. A range of children’s behaviors during procedural pain has been described including nonprocedural talk, using humor, engaging in distraction, asking medically related questions, and seeking emotional support. Based on theoretical and empirical data, some authors have hierarchically grouped children’s behaviors into “distress,” “coping,” and “neutral” behaviors.4 Behaviors such as nonprocedural talk, using humor, and engaging with distracting stimuli have often been termed “coping” behaviors because they are indicative of children’s attention being deployed away from the distressing event.4–6 Behaviors typically associated with the procedure, such as medically related talk, were termed “neutral” behaviors. Children’s behaviors termed “distress” included crying, verbal resistance, verbal fear, seeking information, and seeking emotional support.7
Current scoring methodologies of children’s distress have precluded an in-depth analysis of children’s perioperative behavior. The exact prevalences of different types of distress behaviors are unknown; there are no strong data on how many children resist anesthesiologists during induction or how many report fear. Furthermore, little is known about other behaviors exhibited by children in the perioperative process. What children do during nondistress time points (i.e., on the walk to the operating room [OR]), may help to identify those behaviors that would be adaptive if used to a greater extent during stressful timepoints.
The purpose of the current report is to use detailed behavior analytic methods to characterize children’s behavior from a large-scale project examining adult-child interactions in the perioperative environment. Second-to-second coding of discrete behaviors using computerized coding technology provided data on both the topography and the proportionality of children’s behavior. In this way, the current report examines the number of children displaying specific behaviors and the proportion of time children spend engaging in these behaviors across phases of anesthesia induction. A large age range of children allowed the examination of children’s behavior profiles across age.
Participants in this study were children with ASA physical status I or II who were a part of the National Institute of Health funded Behavioral Interactions-Perioperative (BIP) study. The BIP study is a large-scale multi-year project assessing the main effects and moderators of adult behaviors on children’s perioperative distress. Children recruited for the BIP study were aged 2–10 yr undergoing outpatient surgery with general anesthesia. Exclusion criteria included children with chronic illness, children with developmental delay, and children with parents who did not speak English. This current report includes 293 children who were part of the BIP study. Forty-eight percent of these children were female, and most were non-Hispanic Caucasian (85.7%). Thirty-five percent of children had previous experience with surgery. The most common surgical procedures were tonsillectomy and/or adenoidectomy (n = 96), followed by pressure-equalizing tube placement (n = 47), endoscopy (n = 40), urological procedures (n = 28), hernia repairs (n = 21), and dermatological procedures (n = 14).
Yale Preoperative Anxiety Scale (mYPAS) (Child)1
This observational measure of preoperative anxiety was developed and validated in previous investigations. The mYPAS consists of 27 items in five categories of behavior indicating anxiety in young children (Activity, Emotional expressivity, State of arousal, and Vocalization). Using kappa statistics, all mYPAS categories have good to excellent interobserver and intraobserver reliability (0.73–0.91), and when validated against other global behavioral measures of anxiety, the mYPAS had good validity (r = 0.64). The mYPAS score ranges from 22.5 to 100 with higher scores indicating greater anxiety. Since its development, this scale has been used in multiple investigations.8–11
Behavioral Coding System
Description of Coding System: Revised Perioperative Child-Adult Medical Procedure Interaction Scale (R-PCAMPIS)
The R-PCAMPIS is an observational behavioral coding system designed to capture children’s and adults’ behaviors in the perioperative setting. Based on the originally validated PCAMPIS,12 the R-PCAMPIS includes 44 operationally defined verbal and nonverbal behavioral codes. Modifications to the original PCAMPIS were made to facilitate the interface between the coding system and a behavioral collection computer system, Observer XT (Noldus, The Netherlands). Codes were subdivided into 1) state codes: those behaviors of which duration was a meaningful metric (e.g., cry) and 2) event codes: those in which frequency, not duration, was a meaningful metric (e.g., verbalizing negative emotion). Event codes are typically representative of verbalizations; in this way, whether a verbalization of “I’m Scared” takes 1 or 3 s to utter is of less interest than how many times an utterance like this is used. Although behavioral data on all individuals in the perioperative environment (i.e., physician, nurse, parent, and child) were collected for the BIP study, because of the extremely large amount of data, only child codes are described in this report. Child behaviors are listed in Table 1. Operational definitions and coding manual are available from the authors upon request.
Training of Raters
Two bachelor’s level and one master’s level researchers completed the behavioral coding. All coders underwent a 3-mo training protocol under the direction of the first author (JMC). This training process included two phases. First, coders were familiarized with the technological coding interface, Observer XT, via administration of a simplified set of behavioral codes. Second, coding of study-independent training videos was accomplished. Multiple raters coded each training video and met, at length, with the first author to discuss reliability statistics and disagreements. Raters were considered “trained” when they met a kappa criterion of 0.80 with the first author on training tapes.
Administration of the R-PCAMPIS was facilitated by using Observer® XT (Noldus), a behavior-analysis software package with the capabilities to code behaviors of one individual, or the interactions of many. This system allows for the linking of particular behaviors (e.g., nonverbal resistance) to the subject who initiated the behavior (e.g., child). In addition, the system allows each behavior coded to be linked to the subject to whom the behavior was directed (e.g., toward anesthesiologist). Data coding was accomplished in passes, with each behavior coded in a separate pass. Coding data in passes ensured that behaviors that were not mutually exclusive (i.e., cry and nonverbal resistance) were independently coded for duration. Real-time second- to-second data coding was used with onsets and offsets of state behaviors and onsets of event behaviors recorded. Although this methodology is time consuming, it ensures maximum reliability and validity of coding. Coding required approximately 4 h per participant.
Behaviors were coded into four phases: 1) behaviors occurring from the time the child left the holding room until they arrived at the OR door (walk to OR), 2) behaviors occurring from the time the child enters the OR to the time the mask is introduced (OR entry), 3) behaviors occurring from the time the child is notified of the mask to the time the mask is placed and remains in place (mask notification), and 4) behaviors occurring from the time the mask is placed to the time the child makes their last conscious movement (mask placement). Data obtained from the R-PCAMPIS include duration of state behaviors and frequency of event behaviors across phase.
Data Exporting and Compiling
Data were exported from Observer XT into text files and were then imported into a computer program, General Sequential Analysis Querier (GSEQ) for compilation. GSEQ was used to calculate summary data on the frequency, duration, and rate of behaviors across phases. Summary data were then imported into SPSS 17 for further analysis (for details on analyses conducted, see Statistical Analysis section).
Interrater reliability of individual behavioral codes was assessed by having two research assistants overlap on 10% of participants. Timed-event kappa coefficients were in the good-to-excellent range (range 0.77–1.0). Reliability assessment and discussion was a process repeated weekly throughout coding. One reliability subject was coded per week; once kappa values were calculated, coders met with the first author to discuss disagreements. Decisions on valid placement of behaviors into codes were incorporated into the final version of the observational records.
All procedures were approved by the Yale Human Investigation Committee (New Haven, CT). Participants were recruited by phone between 1 wk and 1 day before surgery or on the morning of surgery. Parents provided written informed consent, and children provided written assent as age appropriate (i.e., children older than 7 yr). After giving informed consent, parents completed a demographic questionnaire and measures relevant to the larger BIP study. All children were accompanied to the OR by one parent and no child received any sedative premedication. A trained research assistant rated anxiety using the mYPAS at the point of induction. Anesthesia was induced in a standardized manner; upon arrival in the OR, a Spo2 probe was placed on child’s finger and a scented anesthesia mask was presented to the child. O2/N2O was introduced in a ratio of 3/7 L flow for 2 min, and sevoflurane was started in a concentration of 0.5% then increased every three breaths to a maximum of 6%. All inductions were accomplished by pediatric anesthesiologists.
A trained research assistant using a hand-held digital video camera (Sony Handycam DCR-HC21) videotaped all children from the time the child left the holding area until anesthesia was induced. Digital video files were converted to .mpg files and imported into Observer XT software (Noldus) for coding, compiling, and analysis.
Statistical analyses were performed in a series of steps. First, descriptive data on the proportion of children displaying individual behaviors are reported. The varying length of observation was accounted for by dividing the duration of state behaviors by total observation time and multiplying by 100; therefore resulting in a statistic of the proportion of observation time during which a child displayed a particular state code. Frequency of event codes was divided by number of seconds in the observation and multiplied by 60 to obtain a rate of code display per minute. Descriptive data on the relevant statistic for each code are presented. Next, patterns of children’s behaviors across phases of induction were examined using repeated measures analyses of variance for each behavior. Bonferroni corrected P values were used to control for familywise error. Visual inspection was used to group behaviors that showed similar phase profiles, and a summary score for each profile was calculated for each child (sum of rate or proportion of codes in that profile). The relation between age and behavior profile was examined using one-way analysis of variance to compare mean profile scores across age group (2–3 yr, 4–6 yr, and 7–10 yr). Construct validity of behavior profiles was examined using correlations and hierarchical regression controlling for child age. Data compilation and summarization were accomplished using GSEQ for Windows (Bakeman and Quera, Atlanta, GA) and was analyzed using SPSS 17 (SPSS, Chicago, IL).
Overall Prevalence of Children’s Behaviors
The proportion of children displaying each behavior is shown in Table 1. The highest proportion of children displayed engagement in medical play which was usually indicative of children being involved in medical play with anesthesiologists (e.g., playing “astronaut” with the mask). The next most common behavior was nonverbally resisting the procedure. The proportion of observation time spent engaging in state behaviors and rate per minute of event behaviors are also shown in Table 1. Children cried and resisted up to 57.5% of the observation time. The most common event behavior was verbal resistance, and the least common event behavior was verbalizing positive affect about the procedure (e.g., “This is cool”). Notably, there were no differences in proportions of any displayed behavior between children who had previous experience with surgery and those who did not on any behavior(χ2’s range from 0.02 to 1.01, all P’s >0.05).
To better describe the prevalence of children’s distress, children’s display of the five clear distress behaviors (cry, scream, nonverbal resistance, verbal resistance, and negative verbal emotion) was examined further. Forty-two percent of all children displayed at least one of the five distress behaviors (Fig. 1) and 16.7% of all children displayed significant distress, characterized by at least three distress behaviors. Of children showing distress, the most common display was cry, verbal resistance, and nonverbal resistance together (23.8%), followed by nonverbal resistance alone (17.2%), and nonverbal resistance and cry together (16.4%). A significantly higher proportion of 2- and 3-yr-old children showed any distress behavior than 4–6-yr-old children, χ2 (1) = 21.4, P < 0.001, and 7–10-yr-old children, χ2 (1) = 22.4, P < 0.001. There was no significant difference in the proportions of 4–6- and 7–10-yr-old children showing distress, χ2 (1) = 0.36, P > 0.05.
Children’s Behavior by Phase of Induction
Figures 2a–d shows profiles of children’s behavior across phase of induction. To simplify presentation, codes demonstrating similar profiles across time are shown in separate graphs. The first profile included state behaviors that showed a pattern of steady increase across phase of procedure. This profile was termed Acute Distress and included cry, scream, and nonverbal resistance. The main effects of phase on cry and nonverbal resistance were significant (Fs = 94.3 and 72.7, respectively, P < 0.001). Although the F-value for phase on Scream was large, it was statistically nonsignificant, likely because of the small sample size F (1,2) = 15.4, P = 0.056. The second profile included event behaviors that showed a peak at mask notification and a sharp decrease at mask placement. These behaviors were in two conceptual groups: those that were termed “Procedural Engagement” including medical talk, informing on status, and positive affect about the procedure and those that were termed “Anticipatory Distress” including requesting support, negative verbal emotion, and verbal resistance. Medical talk, positive affect, and verbal resistance demonstrated significant quadratic patterns, Fquad = 6.46, 6.47, and 12.22, respectively, P < 0.05 for all behaviors. Quadratic trends across phase for informing status, requesting support, and negative verbal emotion were not statistically significant. The final profile of behaviors included those that peaked early in the phases, either on the walk to the OR or on OR entry. This profile was termed “Early Regulatory Behaviors” and included information seeking, nonprocedural talk, humor, and coping statements. Linear trends were significant for information seeking and nonprocedural talk, Fs = 8.73 and 3.88, respectively, P < 0.01 for all behaviors. Results were nonsignificant for humor and coping statements. Medical play showed a pattern that was between Early Regulatory and Procedural Engagement with peaks on the walk to the OR and at mask notification. Notably, the highest peak was on the walk, and the linear trend was significant, F (1,109) = 9.62, P < 0.001, so Medical play was deemed to be more similar to the Early Regulatory profile.
Behavior Profiles by Child Age
Profile scores were calculated for each child by summing the rates of codes that were found to have significant repeated measures findings within each profile: Early Behaviors (sum of rate of Nonprocedural Talk and Information Seeking), Anticipatory Distress (rate of Verbal Resistance), Engaging with Procedure (sum of Medical Talk and Positive Affect about Procedure), and Acute Distress (sum of proportion of Cry and Nonverbal Resistance). Results are shown in Figure 3. Older children displayed more Early Regulating Behaviors, F (2,281) = 7.8, P < 0.001 and Procedural Engagement, F (2,281) = 20.08, P < 0.001. Younger children displayed more Acute Distress, F (2,281) = 17.5, P < 0.001. There were no statistically significant differences across age in Anticipatory Distress, F (2,281) = 2.24, P > 0.05.
Construct Validity of Behavior Profiles
Preliminary construct validity of behavior profiles was examined by assessing the relations between profiles and mYPAS scores, a validated measure of children’s anxiety at induction. Correlations were in the expected directions. Scores on the Acute and Anticipatory Distress were both significantly positively correlated with mYPAS scores, r’s = 0.778 and 0.453, respectively, P’s < 0.001. Scores on the Early Behavior and Procedure Engagement profiles were significantly negatively correlated with mYPAS scores, r’s = −0.218 and −0.186, respectively, P’s < 0.01. Given that behavior profiles differed by age, a hierarchical regression was conducted to control for age. Child age was entered in Block 1 of the model, and the four behavior profiles were entered simultaneously in Block 2. The four profiles accounted for significant variance in mYPAS scores above and beyond child age, r2 change = 0.513, P < 0.001. The Acute and Anticipatory Distress and Early Regulatory Behaviors profiles had significant standardized coefficients at the P < 0.05 level. The standardized coefficient for the Procedural Engagement profile was nonsignificant.
To further explore the utility of these behavior profiles in identifying children at risk for anxiety, we examined the relations between behaviors exhibited on the walk to the OR and children’s distress at induction as assessed by the mYPAS. Similar to overall findings, there was a significant positive correlation between Acute and Anticipatory Distress behaviors on the walk to the OR and children’s anxiety at anesthesia induction, r’s = 0.146 and 0.197, respectively, P < 0.01. Children’s display of Early Regulatory Behaviors was significantly negatively correlated with children’s anxiety at induction, r = −0.123, P < 0.05. There was no significant association between Procedural Engagement on the walk and children’s anxiety at induction.
Results of this study indicated that more than 40% of children displayed distress during the process of anesthesia induction. Approximately 17% of children showed “significant distress” characterized by at least three of the following: attempts to escape the procedure, verbal protestations, crying, screaming, or verbally communicating fear or sadness. Although younger children were significantly more likely to display distress than older children, almost 30% of 7–10-yr-old children indicated distress in some way. The most common child behavior was nonverbal resistance (i.e., tried to push mask away), with children doing so, on average, 18.4% of the entire observation period and 42% of the time in which the mask was placed.
Two profiles of distress behaviors were identified in our data. One profile, termed Acute Distress, contained behaviors that increased over phases of induction, peaking at the point at which the mask was placed. These behaviors, especially nonverbal resistance, are those that are particularly problematic as they could interfere with the induction being accomplished in a smooth manner. Not surprisingly, younger children were higher on these types of behaviors. The second profile of distress behaviors termed “Anticipatory Distress” contained behaviors that peaked at mask notification including verbalizations of negative emotion and attempts to delay the procedure. It is notable that Anticipatory Distress may be artificially low at mask placement because the mask limits verbalizations on the part of the child. For this reason, attending to verbal distress behaviors earlier in the induction (i.e., at mask notification) could be a more accurate assessment of distress in older children than looking for crying or other Acute Distress. It is notable that there was a significant association between these behavior profiles on the walk to the OR and children’s distress at induction; not surprisingly children who show distress behavior early are more likely to be distressed later in the induction. This finding highlights the importance of addressing children’s distress early in the induction process. Important in and of themselves, the relations between these behaviors and clinically-relevant recovery outcomes should be examined.
A strength of this study is its attention to a wide range of behaviors rather than being limited to only the evaluation of distress. A focus on family-centered care mandates attention not only to distress but also to families’ overall experience in the health care setting.13 This study identified a profile of behaviors that are common early in the induction process (when children are less distressed), but sharply decrease at induction. Behaviors in this profile, including distracting nonprocedural talk, have been identified in the procedural pain literature as coping behaviors5 and may serve a similar function during induction. Supporting the importance of these behaviors, a negative association was evidenced between children’s use of Early Regulatory Behaviors and distress. Children who displayed more of these behaviors on the walk to the OR (as well as throughout the induction period) showed less distress at anesthesia induction. Similar to the findings of the Acute Distress profiles, this finding highlights the importance of early stages of induction. Similar to the pattern demonstrated here, previous research has demonstrated that children typically do not engage in coping behaviors at times of stress unless they are prompted by adults.14 Thus, future research will examine how to prompt these potentially regulating behaviors in children in later phases of the indication. Interestingly, information seeking has been included as a distress behavior in procedural pain contexts,5 but here it showed a similar profile to regulating behaviors. Information seeking may not be helpful in familiar environments (such as a pediatrician’s office), but may be helpful in unfamiliar perioperative environments.
The final profile of behaviors was indicative of children engaging with the medical procedure. This profile followed a similar pattern to Anticipatory Distress, but was indicative of children being involved in the induction by talking about medical topics and telling anesthesiologists how they were doing. This profile also contained indications that children enjoyed at least part of this process. Further research will examine what adults do to prompt these positive assertions from children.
Several methodological issues should be highlighted. First, few studies in the perioperative area have examined the nature of children’s distress, and those that do generally use one-time measures or composite measures of children’s distress behavior. This study examines individual child behavior and does so across phases of anesthesia induction. Furthermore, because this study used second-by-second data coding, we were able to gain data on how much of each behavior children displayed, rather than simple presence/absence. Finally, this article is the first of its kind to examine a range of child behaviors (rather than simply distress) in the perioperative setting. Identifying potential desirable behaviors provides additional targets for interventions that ameliorate children’s entire perioperative experience. In terms of limitations, it is notable that all children included in this study had their parents present at induction and none received sedative premedication, thus the degree to which these findings generalize to children without parents present or to those who have been premeditated is unknown.
We conclude that a clinically significant proportion of children display distress behaviors during anesthesia induction, with younger children displaying more Acute Distress. A profile of Early Regulatory Behaviors was identified that could serve as a coping mechanism if children were prompted to continue these behaviors throughout induction.
The authors thank Carrie Hammell, Megan Weinberg, and Cristina Novoa for their assistance in data coding.
1. Kain ZN, Mayes LC, Cicchetti DV, Bagnall AL, Finley JD, Hofstadter MB. The Yale Preoperative Anxiety Scale: how does it compare with a “gold standard?” Anesth Analg 1997;85:783–8
2. McCann M, Kain Z. Management of preoperative anxiety in children: an update. Anesth Analg 2001;93:98–105
3. MacLaren JE, Cohen LL, Cohen S. Children’s behavior during immunization injections: a principle components analysis. Child Health Care 2007;36:237–48
4. Blount R, Sturges J, Powers S. Analysis of child and adult behavioral variations by phase of medical procedure. Behav Ther 1990;21:33–48
5. Blount RL, Corbin S, Sturges J, Wolfe V, Prater J, James L. The relationship between adults’ behavior and child coping and distress during BMA/LP procedures: a sequential analysis. Behav Ther 1989;20:585–601
6. MacLaren JE, Cohen LL. A comparison of distraction strategies for venipuncture distress in children. J Pediatr Psychol 2005; 30:387–96
7. Jay SM, Ozolins M, Elliot CH, Caldwell S. Assessment of children’s distress during painful medical procedures. Health Psychol 1983;2:133–47
8. Finley GA, Stewart SH, Buffett-Jerrott S, Wright KD, Millington D. High levels of impulsivity may contraindicate midazolam premedication in children. Can J Anaesth 2006;53:73–8
9. Kain ZN, Mayes LC, Wang SM, Caramico LA, Krivutza DM, Hofstadter MB. Parental presence and a sedative premedicant for children undergoing surgery: a hierarchical study. Anesthesiology 2000;92:939–46
10. Vagnoli L, Caprilli S, Robiglio A, Messeri A. Clown doctors as a treatment for preoperative anxiety in children: a randomized, prospective study. Pediatrics 2005;116:e563–e567
11. Patel A, Schieble T, Davidson M, Tran MC, Schoenberg C, Delphin E, Bennett H. Distraction with a hand-held video game reduces pediatric preoperative anxiety. Paediatr Anaesth 2006;16:1019–27
12. Caldwell-Andrews A, Blount R, Mayes L, Kain Z. Assessing behavioral interactions in the perioperative environment: development of the P-CAMPIS. Anesthesiology 2005;103:1130–5
13. Institute of Medicine’s Committee on Quality of Health Care in the 21st Century. Crossing the quality chasm: a new health system for the 21st century. Washington, DC: National Acadamy Press, 2001
14. Cohen LL, Bernard RS, Greco LA, McClellan CB. A child-focused intervention for coping with procedural pain: are parent and nurse coaches necessary? J Pediatr Psychol 2002; 27:749–57