Journal of Pediatric Psychology, Vol. 17, No. 6, 1992, pp. 757-773
Perceived Control as a Predictor of Distress in Children Undergoing Invasive Medical Procedures1 Paul J. Carpenter2 University of Rochester Medical Center
Children's anticipatory attributional assessment of the source(s) of perceived control is hypothesized to play an important role in the etiology of distress in children undergoing invasive medical procedures. Four perceived control types, based on learned helplessness theory, are specified by a conceptual model that guides this research: Mutual, Powerful Other, Personal, and Unknown. Among 73 children between the ages of 4 and 18 having their blood drawn, it was predicted that children with an attributional analysis of unknown perceived source of control prior to the impending medical procedures would experience a heightened level of anxiety (procedure-related distress). As predicted, children with an anticipatory attributional assessment of unknown perceived control interfered with or extended the medical procedure significantly more (41%) than children who could attribute some perceived source of control (13%). They were also rated by themselves, the parents, and a trained clinical observer as manifesting significantly greater (p < .05) anticipatory procedural distress using both cognitive (subjective) and behavioral (objective) assessment perspectives. These findings were independent of children's age. This paper supports the need for additional theory-driven research and the importance of investigating the role of attributional variables in the etiology of procedure-related distress in children. KEY WORDS: medical procedures; distress; perceived control; learned helplessness. 'The author thanks Axel Kaires and his staff in the Outpatient Phlebotomy Department at the University of Rochester Medical Center for their support and help with this research. The author also thanks Yvette Auger for her valuable contribution in the collection of study data, Barbara Cichetti for the preparation of the manuscript, and Gary R. Morrow for his helpful critique of an earlier draft of this paper. Portions of this paper were presented at the annual meeting of the Society of Behavioral Medicine, Chicago, April 20, 1990. This research and the preparation of the manuscript were supported in part by grants from the American Cancer Society (IN-18-31; PBR52). 2 AU correspondence should be addressed to Paul J. Carpenter, Box 704, Behavioral Medicine Unit, University of Rochester Cancer Center, 601 Elmwood Avenue, Rochester, New York 14642. 757 0I46-8693/92/I200-O757M6.5O/0 © 1992 Plenum Publishing Coiporalion
Downloaded from http://jpepsy.oxfordjournals.org/ at University of Winnipeg on August 19, 2015
Received September 7, 1991; accepted March 25, 1992
758
Carpenter
Downloaded from http://jpepsy.oxfordjournals.org/ at University of Winnipeg on August 19, 2015
The safe and efficient administration of invasive medical procedures poses one of the greatest challenges to the pediatric health care delivery system (Carpenter, 1991; Peterson, 1989). As a result, a fair amount of research has been focused on this investigative area. The need for investigating the efficacy of behavioral interventions for addressing this clinical problem is paramount. However, before additional behavioral intervention programs are evaluated and recommended for clinical use, more research focused on the etiology of procedure-related distress is needed. As we (Carpenter, 1991) and others (Peterson, 1989) have underscored, scientific progress in this has been slow due to a host of theoretical, conceptual, and methodological issues. Clearly there is a need for conceptual consistency and a greater emphasis on theory-derived investigations. In an effort to help stimulate progress in this investigative area, I made a distinction between strategy-based versus model-based research intervention approaches (Carpenter, 1991). To summarize, the strategy-based research approach asks the empirical question: What behavioral strategy or combination of strategies is most effective in helping children to cope successfully with invasive medical procedures? The accumulation of research using this empirical approach indicates that any behavioral strategy is better than none in helping to ameliorate the noxious physical and psychological effects of invasive medical procedures (e.g. Blount, Davis, Powers & Roberts, 1991; Ludwick-Rosenthal & Neufeld, 1988). Despite these positive effects, the results also suggest that many children do not benefit from the universal application of specific behavioral strategies. Furthermore, within the strategy-based research approach, investigations assessing the efficacy of matching specific coping styles of children (e.g., repressor, sensitizer) with specific behavioral strategies thought consistent with a particular coping trait have yielded results that are inconsistent and sometimes opposite of that predicted (cf. Blount et al., 1991). The absence of conceptual models for the development and empirical evaluation of many of these investigations have limited their potential clinical usefulness. In contrast, the model-based research approach asks the empirical question: What specific factors are helpful in facilitating the efficacy of specific behavioral strategies in helping specific children cope with medical procedures? Within this approach, the identification of specific intervention strategies is not the primary objective. Rather, the effective management of procedure-related distress involves a complex interaction between the process by which a specific behavioral strategy is facilitated and the specific factors that are predictive of children who manifest procedure-related distress. Although several reviews of the literature pertaining to procedure-related distress in children (e.g., Blount et al., 1991; Jay, 1988; Ludwick-Rosenthal & Neufeld, 1988; Peterson, 1989; Zeltzer, Jay, & Fisher, 1989) and the report of a subcommittee of a conference focused on the management of pain associated with procedures in children with cancer (Zeltzer
Perceived Control
759
et al., 1990) have consistently underscored the need for a greater emphasis on model-based research investigations and clinical trials, this approach continues to be rare. The literature in experimental psychology and other basic sciences provides a rich source of empirically tested models that may apply to the investigation of this clinical problem, however, there has been minimal transfer of this technology to this investigative area. The integration of basic and applied research in this investigative area is an important goal.
APPLICATION OF COGNITIVE/ATTRIBUTION THEORY TO INVESTIGATING THE ETIOLOGY OF PROCEDURE-RELATED DISTRESS IN CHILDREN Downloaded from http://jpepsy.oxfordjournals.org/ at University of Winnipeg on August 19, 2015
Although the construct of "control" has been discussed repeatedly in the literature (e.g., Fosterling, 1988; Steptoe, 1989), the importance of this variable in conceptualizing the etiology of procedure-related distress in children has not received much systematic investigation (Steptoe, 1989). Because recent research has revealed that "giving control" to children during a medical procedure was associated with greater distress (Blount et al., 1991), conceptual clarity between the behavioral and attributional definition of control is warranted. The behavioral definition of control refers to a child's ability to make a decision or choice regarding how or when a medical procedure is performed. For example, a parent or the clinician performing the medical procedure may transfer decisional control to a child regarding when the procedure might begin or which arm he/she prefers to use. The cognitive/attributional perspective, on the other hand, defines control as a child's perception that he/she can successfully cope with the physical and psychological stressors of an impending medical procedure. This perception involves the positive integration and trust in the following ingredients: (a) an understanding of what will happen, what will be experienced, and when it will happen; (b) access to a perceived effective coping strategy; (c) if needed, having a trusted significant other (e.g., parent) to provide active emotional support and coping guidance; and (d) cognitive-behavioral rehearsal to work through anticipatory anxiety. Thus, we hypothesize that children's attributional assessment of perceived control to an impending aversive event, such as an invasive medical procedure, plays a significant role in the etiology of distress behaviors elicited during these stressful events. Attributional theory associated with the model of learned helplessness provides a fundamental structure for conceptualizing individual differences in childrens' distress responses to invasive medical procedures, and for potentially developing interventions to ameliorate such distress among pediatric patients. According to the original theoretical formulation of learned helplessness
760
Carpenter
Downloaded from http://jpepsy.oxfordjournals.org/ at University of Winnipeg on August 19, 2015
(Seligman & Maier, 1967), which has been summarized by Nolen-Hoeksema, Seligman, and Girgus (1986), experience with uncontrollable stressors can lead to the expectation that no responses in one's repertoire of responses will control future outcomes. The model hypothesizes that this expectation or attributional assessment results in specific behavioral (decreased response initiation and persistence), cognitive (inability to perceive existing alternatives to control outcomes) and, in humans, emotional (depression and lowered self-competence) deficits or symptoms. Four primary theoretical inadequacies of the original model of learned helplessness have been raised: (a) It does not explain when helplessness deficits would be stable in time and when they would be unstable; (b) it does not explain why helplessness deficits would generalize to multiple domains of outcomes and when they would be specific to one domain; (c) it does not explain why people would lose self-competence when they perceived they were helpless; and (d) it does not account for individual differences in humans' susceptibility to helplessness. In response to these theoretical issues, Abramson and her colleagues (Abramson, Seligman, & Tiesdale, 1978; Abramson, Garber, & Seligman, 1980) proposed a reformulation. According to this reformulation, the attributional assessment that people formulate for good and bad outcomes influence their expectations about future outcomes, and thereby influence their reactions. The reformulation incorporates three dimensions that are hypothesized to influence the symptoms of learned helplessness that individuals experience following an event: (a) Stability—causes can be stable in time or they can be unstable. If a person explains a bad event by a cause that is stable rather than unstable in time, he or she will expect bad events to reoccur in the future and helplessness deficits will be chronic; (b) Generality—causes can have effects in many areas of an individual's life, or they can effect only one area. If an individual explains a bad event by a cause that has global effects instead of by a cause that influences only that specific event, he or she will expect bad events to occur in multiple domains and helplessness deficits will generalize across domains; and (c) Locus of control—causes can either be internal or external to the individual. If a person explains a bad event by a cause internal to him/herself rather than external, he or she will be more likely to show lowered self-competence. To explain individual differences in vulnerability to learned helplessness, Abramson et al. (1978, 1980) hypothesized that people who habitually attribute bad events to internal, stable, and global causes will be more likely to experience the general and longlasting symptoms of helplessness than people with the opposite attributional style. We have drawn from learned helplessness theory to conceptualize and empirically investigate the role of children's cognitive appraisals or attributional assessments in predicting individual differences in the magnitude of distress observed during the course of an invasive medical procedure. Our previous
761
Perceived Control
Table I. Role of Perceived Source of Control in Conceptualizing the Etiology of Distress in Children Undergoing Invasive Medical Procedures Role of Child
Role of Powerful Other
Perceived Control By Child (Child perceives that effective coping is dependent on some strategy used by him/her)
Perceived Control From Powerful Other (Child perceives that effective coping is dependent on some strategy used by a powerful other—e.g., parent)
No Perceived Control From Powerful Other (Child perceives that effective coping is independent of any strategy used by a powerful other)
I Mutual Control (Effective coping is in both my control and the control of a powerful other)
Personal Control (Effective coping is only in my control)
"Table adapted from Abramson et al. (1978, 1980).
No Perceived Control By Child (Child perceives that effective coping is independent of any strategy used by him/her)
Control From Powerful Other (Effective coping is only in the control of a powerful other)
Unknown Control/Helplessness (Effective coping is not in anyone's control)
Downloaded from http://jpepsy.oxfordjournals.org/ at University of Winnipeg on August 19, 2015
research has demonstrated the potential importance of children's subjective selfreport of anticipatory fear and pain as potent predictors of clinician rated procedure-related distress (Carpenter, 1990). As an initial step in applying the multifaceted, theoretical dimensions associated with learned helplessness theory for evaluating individual differences in children's emotional/behavioral responses to invasive medical procedures, a conceptual model that focuses on the dimension of control versus external sources is formulated. Future investigations will examine the additional influences specific to the dimensions of stability and generality. Table I presents a conceptual typology of the construct of children's perceived source(s) of control in predicting the severity of distress experienced by children during the course of actual invasive medical procedures. Based on this conceptual typology, we predict that children who attribute an unknown source of perceived control to an impending medical procedure will experience an acute state of helplessness that triggers capacitating anxiety resulting in excessive avoidance behavior that interferes with or extends the safe and efficient administration of a medical procedure. The literature pertaining to learned helplessness theory had devoted limited attention to the construct of anxiety and the symptoms associated with this
762
Carpenter
Downloaded from http://jpepsy.oxfordjournals.org/ at University of Winnipeg on August 19, 2015
affective state. Therefore, literature pertaining to individual's response patterns to the experience of acute versus chronic pain is helpful for clarifying the importance of this construct. This literature also supports the rationale for the hypothesis of increased anxiety and avoidance behavior proposed in the present investigation. In fact, the learned helplessness literature has been repeatedly criticized for an overemphasis on depressive symptomology as the primary outcome of learned helplessness (Fosterling, 1988). The pain literature suggests that the response patterns of individuals exposed to an acute as opposed to a chronic stressor and who also experience a perceived loss of control or helplessness are different (Amtz & Kelly, 1989). In addition, this literature also suggests that other mediating constructs, such as loss of control/lack of control and helplessness/hopelessness, might also be important conceptual components of the learned helplessness syndrome (Mineka & Kelly, 1989). More specifically, knowledge of a specific impending acute painful event (e.g., having blood drawn) may trigger an attributional assessment in which an individual perceives no source of control in successfully coping. This attributional assessment leads to a state of acute helplessness that can result in a response pattern during the painful event characterized by increased anxiety and avoidance behavior. On the other hand, chronic pain may trigger an attributional assessment in which the individual perceives a generalized inability to ever cope with the pain that can lead to a state of both chronic helplessness and hopelessness resulting in a response pattern characteristic of depressive symtomatology (Arntz & Kelly, 1989). With respect to children, perceived source(s) of control becomes even more conceptually important. The helplessness syndrome in children can result from the absence of a parent, overstimulation of a parent, and/or nonresponsive parenting. If parents do not give the child adequate cues or warning signals about impending stressful or aversive events, the child becomes extremely anxious at inappropriate times. The child cannot predict what will happen and therefore experiences helplessness. A child's excessive fear is frequently the result of inadequate communication or signals from parents or significant others. Children can cope with unpleasant events (e.g., invasive medical procedures) better in the long run if they are given a warning and an honest explanation. The child whose parent is nonresponsive is deprived of control over emotional stimulation. On the other hand, the parent who is overresponsive to his/her child and does not balance the intensity or frequency of his/her own responses to those emitted by his/her child is also out of synchrony. Thus, lack of synchronized responsiveness leads to helplessness. The key to a child's coping mastery, then, is a sensitive, synchronized response of a parent or significant other. Thus, the importance of perceived mutual control is stressed by its inclusion within the conceptual model presented in Table I.
Perceived Control
763
EMPIRICAL VALIDATION OF THE CONCEPTUAL MODEL To empirically test the hypothesis that children who attribute unknown perceived source(s) of control to an impending medical procedure are at greatest risk for experiencing procedure-related distress, we systematically examined the relationship between children's anticipatory cognitive appraisal of perceived source(s) of control with respect to a specific medical procedure (phlebotomy) and the magnitude of avoidance behaviors manifested during the actual procedure.
METHODS
A total of 73 children who were having their blood drawn in the pediatric section of the Outpatient Phlebotomy Department of the University of Rochester Medical Center were studied. The children ranged in age from 4 to 18 with an average age of 10.0 (SD = 3.8). Thirty percent (« = 72) were between the ages of 4 and 7, 19% (n = 14) between the ages of 8 and 9, 25% (n = 18) between the ages of 10 and 12, and 26% (n = 19) between the ages of 13 and 18. Fifty-two percent (« = 38) of the children were male and 77% (n = 56) Caucasian. Fifteen percent (n = 11) of the children were having their blood drawn for the first time, 54% (n = 39) have their blood drawn on a regular basis (7% annually, 34% monthly, and 13% weekly) and 31% (n = 23) have had their blood drawn on an irregular basis. Eighty-three percent (n = 60) of the children were accompanied by their mothers, 9% (n = 7) were accompanied by their fathers, and 8% (n = 6) by both their mothers and fathers, and 77% (n = 56) were told in advance that they would be having their blood drawn.
Measures Children's Perceived Control In reviewing the literature specific to the assessment of children's perceived control, we were particularly drawn to the methodology developed by Connell (1985) becauses of its conceptual and scientific rigor. The Multidimensional Measure of Children's Perceptions of Control emerged from this work. Furthermore, in comparison to other existing measures of children's perceived control (e.g., Children's Multidimensional Health Locus of Control Scale) which pri-
Downloaded from http://jpepsy.oxfordjournals.org/ at University of Winnipeg on August 19, 2015
Subjects
764
Carpenter
Downloaded from http://jpepsy.oxfordjournals.org/ at University of Winnipeg on August 19, 2015
marily assesses generalized expectancy of perceived control, Connell's methodology was designed to assess children's domain-specific perceptions of control. The primary focus of Connell's research using this assessment methodology has been on the investigation of school-aged children's perceptions of perceived control specific to domains within the school environment (i.e., cognitive, social, and physical; Connell, 1985). Connell encouraged investigators to modify his items to reflect the specific domain being assessed. Thus, the Multidimensional Measure of Children's Perceptions of Control represents a structured methodology rather than a fixed scale. More specifically, this methodology standardizes the assessment of three sources of perceived control (Internal, Powerful Other, and Unknown) for any specified domain of interest. Two items are used to measure each perceived source of control. The items are worded identically across each specific domain of interest and are different only in reference to the specific domain being sampled. A response set ranging from 1 to 4 (very true, sort of true, not very true, not at all true) is used in scoring all items. A set of flexible, yet standardized, guidelines for the administration of the assessment methodology is recommended to ensure maximum rapport. Connell's experience in administering this scale indicates that the most valid results emerge when a flexible approach is used. A modification of the Multidimensional Measure of Children's Perceptions of Control consisting of 22 items, using the methodology described above, was used to assess children's source(s) of perceived control to the impending medical procedure. The modification involved a rewording of each of the two items within each source of perceived control (Personal, Powerful Other, and Unknown) specific to the target medical procedure being assessed (e.g., when my mother or father are in the room when I get my blood drawn, I can stay still). Consistent with the way we operationalize procedure-related distress (see discussion below for more detail) all perceived control items target "staying still" as the primary coping behavior. In addition, the 10-item Rosenberg Self-Esteem Scale (Rosenberg, 1979) is embedded into the questionnaire to assess children's perceived self-competence. Assessment of self-competency was included because of the importance of this construct within learned helplessness theory (Abramson et al., 1978, 1980). Five of the items assess competencies in the positive direction and 5 in the negative direction and are summed separately to derive a total positive self-esteem score and a total negative self-esteem scale. Because we wished to assess the construct of perceived source(s) control in children as young as 4 years of age, the developmental appropriateness of all items were evaluated by two experts in developmental psychology. Following additional editing to further refine the developmental appropriateness of items, 100% agreement between these two experts was achieved. The final form of the modified scale was piloted on 10 children between the ages of 4 and 7 to assess
Perceived Control
765
Anticipatory Procedure-Related Distress The assessment of children's anticipatory procedure-related distress was obtained from three sources: Children's Ratings. Children's anticipatory self-ratings of fear and pain with respect to the impending medical procedure were measured using the Children's Global Rating Scale (CGRS; Carpenter, 1990). Because of the difficulties in measuring complex psychological constructs, such as fear and pain, across diverse developmental populations, different measuring techniques have often been used for younger and older children posing serious problems to the internal and external validity of study results. To address this methodological problem, we developed the Children's Global Rating Scale, a self-report methodology for measuring children's perceptions of such constructs as pain and fear (Carpenter, 1990). A series of studies (Carpenter, 1990, 1991) with 145 children between the ages of 4 and 17 undergoing invasive medical procedures examined: (a) the comparability of the CGRS with another commonly used self-report method for assessing pain and fear in young children—the Faces Scale; (b) the convergent and predictive validity of the CGRS; and (c) applicability of the CGRS for assessing the constructs of pain and fear across the developmental span. The results of these investigations indicated that (a) children's CGRS fear and pain ratings had consistently significantly (p < .05) greater overall Pearson correlations (ranging from .30 to .54) with other measures of procedure-related distress obtained from parents, clinicians, and a trained clinical observer than their similar ratings using the Faces Scale; (b) children's anticipatory CGRS fear ratings with respect to an impending medical procedure emerged as a significant predic-
Downloaded from http://jpepsy.oxfordjournals.org/ at University of Winnipeg on August 19, 2015
the applicability and validity of administering it to young children. Because the scale asks the same question in two different ways for each perceived source of control assessed, the pilot administration revealed that children as young as 4 years of age were capable of validly reporting their attributions regarding perceived source of control if the items are understandable to them. A study by Zeltzer et al. (1988) indicated that young children are capable of using rating scales. We also administered the modified scale in a standardized, yet flexible, manner. All children were administered the scale on an individual basis by a trained research assistant with the child's parent(s) instructed not to coach or answer. Because of the time limitations placed on us by the Phlebotomy Department, all scale items were presented orally to every child participating in the study. A standardized response format card was given to each child to assist in remembering all possible response choices and to control for response preservation.
766
Carpenter
Downloaded from http://jpepsy.oxfordjournals.org/ at University of Winnipeg on August 19, 2015
tor of those children would extend or interfere with the timely administration of the medical procedure; and (c) no significant differences were found between the CGRS fear and pain ratings of children s 7 years of age and children 5: 8 years of age. The rationale, development, administration, and initial empirical evaluation of the CGRS has been discussed in detail elsewhere (Carpenter, 1990). Parent Ratings. Parents' rated their perceptions of their children's anticipatory fear on a 5-point scale ranging from not at all (0) to extremely (4). Parents also rated how scared or apprehensive they were themselves regarding the medical procedure using the same 5-point scale. This global rating scales is identical to the one used in our other investigations of children's distress during medical procedures (Carpenter, 1990). Clinical Observer. A behavior checklist, comprising 21 behavioral items grouped within six domains: Verbal (e.g., I'm scared); Vocal Nonlanguage (e.g., crying), Facial Expressions (e.g., scowling); Body Movements (e.g., flail); Physiological Manifestations (e.g., clenched fist); and General Response to Environment (e.g., hypersensitivity to environmental stimuli) was used to assess the behavioral components of children's distress in the anticipatory phase of the medical procedure. The anticipatory phase began when a child entered the waiting room of the Pediatric Phlebotomy Department and ended when he/she was called by a phlebotomist to enter the procedure room. To assure comparability, the specific behavioral content contained in other behavior observation scales for assessing children's behavioral distress during invasive medical procedures was incorporated into the behavior checklist used in this study (Elliott, Jay, & Woody, 1987; Jay, Ozolins, Elliott, & Caldwell, 1983; Katz, Kellerman, & Siegel, 1980; LeBaron & Zeltzer, 1984). The scoring schema used for our behavioral checklist is identical to that employed by Katz et al. (1980). Specifically, the 21 behavioral items are coded with respect to their observed presence or absence (yes = 1; no = 0). A total behavior index score is calculated by summing of all the observed behaviors within a specific time phase. An interrater reliability check of this behavioral checklist with the initial 20 subjects revealed a statistically acceptable level of interrater agreement (r = .92). Although more sophisticated behavioral sampling techniques and scoring schema have been used, the results of a study by Jay and Elliott (1984) revealed no significant improvement in either the sensitivity or validity of behavioral checklists using more sophisticated methodologies. Actual Procedure-Related Distress Consistent with our previous research in this investigative area, procedurerelated distress was operationalized as the extent to which a child's fearful or avoidance behavior interferes with or extends a specific medical procedure (Car-
Perceived Control
767
penter, 1991). In accordance with this operational definition, the phlebotomist conducting the medical procedure rated the extent to which a child's behavior interfered with or extended the time it normally takes to draw blood in children of a similar age independent of the time the child was in the procedure room using a 5-point scale ranging from not at all (0) to extremely (4). Our previous work has shown clinician's ratings of children's distress during the actual medical procedure to be significantly correlated with children's, parents', and a trained observer's ratings of anticipatory procedure-related distress (all rs, p > .30). Procedure
RESULTS Preliminary Analyses A median split was used to dichotomize the distribution of all 73 children's responses to each of the three sources of perceived control (Personal, Powerful Other, and Unknown) into a high/low group. The combination of high/low scores for each of the three sources of control placed each child into one of the
Downloaded from http://jpepsy.oxfordjournals.org/ at University of Winnipeg on August 19, 2015
The study protocol was designed to assess two phases of the target medical procedure: (a) Anticipatory Phase-—the time the child first entered the Phlebotomy Department; and (b) Actual Procedure Phase—time during which the procedure is performed. Upon entering the Pediatric Outpatient Phlebotomy Departments, children and their parents were approached by a research technician and asked to participate in the study. The study was explained as an investigation of the relationship between children's self-reported perceived control and procedure-related distress. Of the 80 children approached, 73 consented to participate. Time constraints were the reason the other 7 subjects declined to participate. Those who consented were entered on the protocol which was approved by the institutional IRB committee. Prior to having blood drawn, the child's parent(s) were asked some basic questions regarding their child's demographic background and whether or not their child was told about the medical procedure in advance. They also rated their child's anticipatory fear just prior to undergoing the medical procedure as outlined above. Parents were also asked to rate how apprehensive they were before the procedure. The research technician then administered the Perceived Control Questionnaire and the CGRS to the child assessing his/her anticipatory pain and fear. The clinical observer completed the preprocedural behavioral checklist. The phlebotomist completed a rating form once the procedure was over.
768
Carpenter
four perceived control types outlined in Table I. The results of this classification (see Table II) revealed that 19 (26%) children fell into the Mutual Control Group; 7 (10%) into Control From Powerful Others; 20 (27%) Personal Control; and 27 (37%) Unknown source of control. Table II presents the study data for selected demographic variables and measures of anticipatory and actual procedure-related distress obtained from children, parents, a trained clinical observer, and phlebotomist performing the medical procedure for the total sample (n = 73) and for each perceived control type. Main Analyses
Downloaded from http://jpepsy.oxfordjournals.org/ at University of Winnipeg on August 19, 2015
Because the literature in this area has consistently found children's age to be a strong predictor of observable procedure-related distress (Jay, 1988), a group (Mutual, Powerful Other, Personal, Unknown) by age (^7; s 8 ) analytic model was used in the evaluation of all study data examining the relationships between procedure-related distress and children's perceived sources of control. We chose to dichotomize age in this fashion for comparability with other reports in the literature (e.g., see Jay et al., 1983). Analysis of variance (ANOVA) was applied to the statistical analysis of variables with a continuous level of measurements and the chi-square method with a categorical level of measurement (Table III). Demographic Differences. The results revealed no statistically significant differences between the four perceived control types and the selected demographic variables of age, sex, or race. Main Effects for Age. Children's CGRS fear rating was the only variable significantly related to age. Specifically, this one finding revealed that children 8 years of age. Main Effects for Group. As summarized in Table III, the statistical analysis of study data revealed main effects for group for all variables except the negative self-competence items and parents' own fear ratings. Statistical analysis (r and chi-square tests) of the specific differences between the Unknown perceived control group and the Mutual, Powerful Other, and Self perceived control groups were additionally performed to evaluate significant main effects for group. These results revealed that a significantly (p < .01) greater proportion of children in the Unknown perceived control group (41%) were rated by the phlebotomist performing the medical procedure as extending or interfering with the timely or efficient administration of the procedure than any of the other three perceived control types (Mutual = 21%; Powerful Other = 12%; Personal = 5%). In addition, children in the Unknown perceived control group rated themselves, were rated by their parents, and were observed to experience significantly (all ps < .05) greater anticipatory procedure-related distress than children in any of the
Demographic Age (in years) M (SD) Range Male [n (%)] Ethnic minority n (%) Children's self-report [M {SD)] Fear-CGRS Pain-CGRS Self-competence Positive items Negative items Parents' self-report Told child about procedure in advance [n (%)] Child's fear [M (SD)] Parental fear [M (SD)] Trained clinical observer Total [M (SD)] Phlebotomists' ratings Extend/interfere with the medical procedure [n (%)]
Variable
Downloaded from http://jpepsy.oxfordjournals.org/ at University of Winnipeg on August 19, 2015
17 (23)
1.36(2.5)
56 (77) 1.42 (1.3) 1.03(1.1)
17.31 (2.9) 11.25 (3.3)
1.21 (1.5) 1.56 (1.4)
10.00 (3.8) 4-18 38 (52) 17 (23)
Total sample (N = 73)
4(21)
0.37 (0.8)
16 (84) 1.17 (1.0) 1.11 (1.1)
18.7(1.7) 10.8 (2.9)
1.05 (1.4) 1.37(1.3)
8.2 (3.2) 4-16 8(42) 2(11)
Mutual control (n = 19)
1 (12)
1.71 (2.2)
6 (86) 1.29(1.6) 1.14 (1.7)
17.0 (2.2) 11.1 (3.9)
0.86(1.5) 2.00 (1.4)
10.4 (2.6) 7-14 2(29) 2(28)
Control from powerful other (n = 7)
Perceived control types
Table II. Descriptive Statistics for all Variables
1 (5)
0.70 (2.2)
18 (90) 0.82 (1.0) 0.65 (0.9)
17.9 (2.2) 11.3 (2.9)
0.65 (1.1) 0.80 (0.9)
10.9(3.5) 6-18 11 (55) 6(30)
Personal control (n = 20)
11 (41)
2.44 (3.4)
16 (59) 2.09 (1.3) 1.22 (1.1)
15.4 (3.7) 11.7 (3.9)
1.85 (1.6) 2.15(1.5)
10.5 (4.2) 4-18 17 (63) 7(26)
Unknown control (n = 27)
3
1
Conti
770
Carpenter Table m . Analysis for Group, Age, and Interaction Effects'1 Effects* Variable
Age
Group x Age
3.26* 4.45'
4.78rf 0.00
3.00 0.45
5.11' 0.87
0.00 0.27
0.49 9.99
11.86' 4.14' 0.76
0.24 1.27 1.09
2.64C 1.63 1.01
1.05
0.43
0.01
8.23
Perceived Control as a Predictor of Distress in Children Undergoing Invasive Medical Procedures1 Paul J. Carpenter2 University of Rochester Medical Center
Children's anticipatory attributional assessment of the source(s) of perceived control is hypothesized to play an important role in the etiology of distress in children undergoing invasive medical procedures. Four perceived control types, based on learned helplessness theory, are specified by a conceptual model that guides this research: Mutual, Powerful Other, Personal, and Unknown. Among 73 children between the ages of 4 and 18 having their blood drawn, it was predicted that children with an attributional analysis of unknown perceived source of control prior to the impending medical procedures would experience a heightened level of anxiety (procedure-related distress). As predicted, children with an anticipatory attributional assessment of unknown perceived control interfered with or extended the medical procedure significantly more (41%) than children who could attribute some perceived source of control (13%). They were also rated by themselves, the parents, and a trained clinical observer as manifesting significantly greater (p < .05) anticipatory procedural distress using both cognitive (subjective) and behavioral (objective) assessment perspectives. These findings were independent of children's age. This paper supports the need for additional theory-driven research and the importance of investigating the role of attributional variables in the etiology of procedure-related distress in children. KEY WORDS: medical procedures; distress; perceived control; learned helplessness. 'The author thanks Axel Kaires and his staff in the Outpatient Phlebotomy Department at the University of Rochester Medical Center for their support and help with this research. The author also thanks Yvette Auger for her valuable contribution in the collection of study data, Barbara Cichetti for the preparation of the manuscript, and Gary R. Morrow for his helpful critique of an earlier draft of this paper. Portions of this paper were presented at the annual meeting of the Society of Behavioral Medicine, Chicago, April 20, 1990. This research and the preparation of the manuscript were supported in part by grants from the American Cancer Society (IN-18-31; PBR52). 2 AU correspondence should be addressed to Paul J. Carpenter, Box 704, Behavioral Medicine Unit, University of Rochester Cancer Center, 601 Elmwood Avenue, Rochester, New York 14642. 757 0I46-8693/92/I200-O757M6.5O/0 © 1992 Plenum Publishing Coiporalion
Downloaded from http://jpepsy.oxfordjournals.org/ at University of Winnipeg on August 19, 2015
Received September 7, 1991; accepted March 25, 1992
758
Carpenter
Downloaded from http://jpepsy.oxfordjournals.org/ at University of Winnipeg on August 19, 2015
The safe and efficient administration of invasive medical procedures poses one of the greatest challenges to the pediatric health care delivery system (Carpenter, 1991; Peterson, 1989). As a result, a fair amount of research has been focused on this investigative area. The need for investigating the efficacy of behavioral interventions for addressing this clinical problem is paramount. However, before additional behavioral intervention programs are evaluated and recommended for clinical use, more research focused on the etiology of procedure-related distress is needed. As we (Carpenter, 1991) and others (Peterson, 1989) have underscored, scientific progress in this has been slow due to a host of theoretical, conceptual, and methodological issues. Clearly there is a need for conceptual consistency and a greater emphasis on theory-derived investigations. In an effort to help stimulate progress in this investigative area, I made a distinction between strategy-based versus model-based research intervention approaches (Carpenter, 1991). To summarize, the strategy-based research approach asks the empirical question: What behavioral strategy or combination of strategies is most effective in helping children to cope successfully with invasive medical procedures? The accumulation of research using this empirical approach indicates that any behavioral strategy is better than none in helping to ameliorate the noxious physical and psychological effects of invasive medical procedures (e.g. Blount, Davis, Powers & Roberts, 1991; Ludwick-Rosenthal & Neufeld, 1988). Despite these positive effects, the results also suggest that many children do not benefit from the universal application of specific behavioral strategies. Furthermore, within the strategy-based research approach, investigations assessing the efficacy of matching specific coping styles of children (e.g., repressor, sensitizer) with specific behavioral strategies thought consistent with a particular coping trait have yielded results that are inconsistent and sometimes opposite of that predicted (cf. Blount et al., 1991). The absence of conceptual models for the development and empirical evaluation of many of these investigations have limited their potential clinical usefulness. In contrast, the model-based research approach asks the empirical question: What specific factors are helpful in facilitating the efficacy of specific behavioral strategies in helping specific children cope with medical procedures? Within this approach, the identification of specific intervention strategies is not the primary objective. Rather, the effective management of procedure-related distress involves a complex interaction between the process by which a specific behavioral strategy is facilitated and the specific factors that are predictive of children who manifest procedure-related distress. Although several reviews of the literature pertaining to procedure-related distress in children (e.g., Blount et al., 1991; Jay, 1988; Ludwick-Rosenthal & Neufeld, 1988; Peterson, 1989; Zeltzer, Jay, & Fisher, 1989) and the report of a subcommittee of a conference focused on the management of pain associated with procedures in children with cancer (Zeltzer
Perceived Control
759
et al., 1990) have consistently underscored the need for a greater emphasis on model-based research investigations and clinical trials, this approach continues to be rare. The literature in experimental psychology and other basic sciences provides a rich source of empirically tested models that may apply to the investigation of this clinical problem, however, there has been minimal transfer of this technology to this investigative area. The integration of basic and applied research in this investigative area is an important goal.
APPLICATION OF COGNITIVE/ATTRIBUTION THEORY TO INVESTIGATING THE ETIOLOGY OF PROCEDURE-RELATED DISTRESS IN CHILDREN Downloaded from http://jpepsy.oxfordjournals.org/ at University of Winnipeg on August 19, 2015
Although the construct of "control" has been discussed repeatedly in the literature (e.g., Fosterling, 1988; Steptoe, 1989), the importance of this variable in conceptualizing the etiology of procedure-related distress in children has not received much systematic investigation (Steptoe, 1989). Because recent research has revealed that "giving control" to children during a medical procedure was associated with greater distress (Blount et al., 1991), conceptual clarity between the behavioral and attributional definition of control is warranted. The behavioral definition of control refers to a child's ability to make a decision or choice regarding how or when a medical procedure is performed. For example, a parent or the clinician performing the medical procedure may transfer decisional control to a child regarding when the procedure might begin or which arm he/she prefers to use. The cognitive/attributional perspective, on the other hand, defines control as a child's perception that he/she can successfully cope with the physical and psychological stressors of an impending medical procedure. This perception involves the positive integration and trust in the following ingredients: (a) an understanding of what will happen, what will be experienced, and when it will happen; (b) access to a perceived effective coping strategy; (c) if needed, having a trusted significant other (e.g., parent) to provide active emotional support and coping guidance; and (d) cognitive-behavioral rehearsal to work through anticipatory anxiety. Thus, we hypothesize that children's attributional assessment of perceived control to an impending aversive event, such as an invasive medical procedure, plays a significant role in the etiology of distress behaviors elicited during these stressful events. Attributional theory associated with the model of learned helplessness provides a fundamental structure for conceptualizing individual differences in childrens' distress responses to invasive medical procedures, and for potentially developing interventions to ameliorate such distress among pediatric patients. According to the original theoretical formulation of learned helplessness
760
Carpenter
Downloaded from http://jpepsy.oxfordjournals.org/ at University of Winnipeg on August 19, 2015
(Seligman & Maier, 1967), which has been summarized by Nolen-Hoeksema, Seligman, and Girgus (1986), experience with uncontrollable stressors can lead to the expectation that no responses in one's repertoire of responses will control future outcomes. The model hypothesizes that this expectation or attributional assessment results in specific behavioral (decreased response initiation and persistence), cognitive (inability to perceive existing alternatives to control outcomes) and, in humans, emotional (depression and lowered self-competence) deficits or symptoms. Four primary theoretical inadequacies of the original model of learned helplessness have been raised: (a) It does not explain when helplessness deficits would be stable in time and when they would be unstable; (b) it does not explain why helplessness deficits would generalize to multiple domains of outcomes and when they would be specific to one domain; (c) it does not explain why people would lose self-competence when they perceived they were helpless; and (d) it does not account for individual differences in humans' susceptibility to helplessness. In response to these theoretical issues, Abramson and her colleagues (Abramson, Seligman, & Tiesdale, 1978; Abramson, Garber, & Seligman, 1980) proposed a reformulation. According to this reformulation, the attributional assessment that people formulate for good and bad outcomes influence their expectations about future outcomes, and thereby influence their reactions. The reformulation incorporates three dimensions that are hypothesized to influence the symptoms of learned helplessness that individuals experience following an event: (a) Stability—causes can be stable in time or they can be unstable. If a person explains a bad event by a cause that is stable rather than unstable in time, he or she will expect bad events to reoccur in the future and helplessness deficits will be chronic; (b) Generality—causes can have effects in many areas of an individual's life, or they can effect only one area. If an individual explains a bad event by a cause that has global effects instead of by a cause that influences only that specific event, he or she will expect bad events to occur in multiple domains and helplessness deficits will generalize across domains; and (c) Locus of control—causes can either be internal or external to the individual. If a person explains a bad event by a cause internal to him/herself rather than external, he or she will be more likely to show lowered self-competence. To explain individual differences in vulnerability to learned helplessness, Abramson et al. (1978, 1980) hypothesized that people who habitually attribute bad events to internal, stable, and global causes will be more likely to experience the general and longlasting symptoms of helplessness than people with the opposite attributional style. We have drawn from learned helplessness theory to conceptualize and empirically investigate the role of children's cognitive appraisals or attributional assessments in predicting individual differences in the magnitude of distress observed during the course of an invasive medical procedure. Our previous
761
Perceived Control
Table I. Role of Perceived Source of Control in Conceptualizing the Etiology of Distress in Children Undergoing Invasive Medical Procedures Role of Child
Role of Powerful Other
Perceived Control By Child (Child perceives that effective coping is dependent on some strategy used by him/her)
Perceived Control From Powerful Other (Child perceives that effective coping is dependent on some strategy used by a powerful other—e.g., parent)
No Perceived Control From Powerful Other (Child perceives that effective coping is independent of any strategy used by a powerful other)
I Mutual Control (Effective coping is in both my control and the control of a powerful other)
Personal Control (Effective coping is only in my control)
"Table adapted from Abramson et al. (1978, 1980).
No Perceived Control By Child (Child perceives that effective coping is independent of any strategy used by him/her)
Control From Powerful Other (Effective coping is only in the control of a powerful other)
Unknown Control/Helplessness (Effective coping is not in anyone's control)
Downloaded from http://jpepsy.oxfordjournals.org/ at University of Winnipeg on August 19, 2015
research has demonstrated the potential importance of children's subjective selfreport of anticipatory fear and pain as potent predictors of clinician rated procedure-related distress (Carpenter, 1990). As an initial step in applying the multifaceted, theoretical dimensions associated with learned helplessness theory for evaluating individual differences in children's emotional/behavioral responses to invasive medical procedures, a conceptual model that focuses on the dimension of control versus external sources is formulated. Future investigations will examine the additional influences specific to the dimensions of stability and generality. Table I presents a conceptual typology of the construct of children's perceived source(s) of control in predicting the severity of distress experienced by children during the course of actual invasive medical procedures. Based on this conceptual typology, we predict that children who attribute an unknown source of perceived control to an impending medical procedure will experience an acute state of helplessness that triggers capacitating anxiety resulting in excessive avoidance behavior that interferes with or extends the safe and efficient administration of a medical procedure. The literature pertaining to learned helplessness theory had devoted limited attention to the construct of anxiety and the symptoms associated with this
762
Carpenter
Downloaded from http://jpepsy.oxfordjournals.org/ at University of Winnipeg on August 19, 2015
affective state. Therefore, literature pertaining to individual's response patterns to the experience of acute versus chronic pain is helpful for clarifying the importance of this construct. This literature also supports the rationale for the hypothesis of increased anxiety and avoidance behavior proposed in the present investigation. In fact, the learned helplessness literature has been repeatedly criticized for an overemphasis on depressive symptomology as the primary outcome of learned helplessness (Fosterling, 1988). The pain literature suggests that the response patterns of individuals exposed to an acute as opposed to a chronic stressor and who also experience a perceived loss of control or helplessness are different (Amtz & Kelly, 1989). In addition, this literature also suggests that other mediating constructs, such as loss of control/lack of control and helplessness/hopelessness, might also be important conceptual components of the learned helplessness syndrome (Mineka & Kelly, 1989). More specifically, knowledge of a specific impending acute painful event (e.g., having blood drawn) may trigger an attributional assessment in which an individual perceives no source of control in successfully coping. This attributional assessment leads to a state of acute helplessness that can result in a response pattern during the painful event characterized by increased anxiety and avoidance behavior. On the other hand, chronic pain may trigger an attributional assessment in which the individual perceives a generalized inability to ever cope with the pain that can lead to a state of both chronic helplessness and hopelessness resulting in a response pattern characteristic of depressive symtomatology (Arntz & Kelly, 1989). With respect to children, perceived source(s) of control becomes even more conceptually important. The helplessness syndrome in children can result from the absence of a parent, overstimulation of a parent, and/or nonresponsive parenting. If parents do not give the child adequate cues or warning signals about impending stressful or aversive events, the child becomes extremely anxious at inappropriate times. The child cannot predict what will happen and therefore experiences helplessness. A child's excessive fear is frequently the result of inadequate communication or signals from parents or significant others. Children can cope with unpleasant events (e.g., invasive medical procedures) better in the long run if they are given a warning and an honest explanation. The child whose parent is nonresponsive is deprived of control over emotional stimulation. On the other hand, the parent who is overresponsive to his/her child and does not balance the intensity or frequency of his/her own responses to those emitted by his/her child is also out of synchrony. Thus, lack of synchronized responsiveness leads to helplessness. The key to a child's coping mastery, then, is a sensitive, synchronized response of a parent or significant other. Thus, the importance of perceived mutual control is stressed by its inclusion within the conceptual model presented in Table I.
Perceived Control
763
EMPIRICAL VALIDATION OF THE CONCEPTUAL MODEL To empirically test the hypothesis that children who attribute unknown perceived source(s) of control to an impending medical procedure are at greatest risk for experiencing procedure-related distress, we systematically examined the relationship between children's anticipatory cognitive appraisal of perceived source(s) of control with respect to a specific medical procedure (phlebotomy) and the magnitude of avoidance behaviors manifested during the actual procedure.
METHODS
A total of 73 children who were having their blood drawn in the pediatric section of the Outpatient Phlebotomy Department of the University of Rochester Medical Center were studied. The children ranged in age from 4 to 18 with an average age of 10.0 (SD = 3.8). Thirty percent (« = 72) were between the ages of 4 and 7, 19% (n = 14) between the ages of 8 and 9, 25% (n = 18) between the ages of 10 and 12, and 26% (n = 19) between the ages of 13 and 18. Fifty-two percent (« = 38) of the children were male and 77% (n = 56) Caucasian. Fifteen percent (n = 11) of the children were having their blood drawn for the first time, 54% (n = 39) have their blood drawn on a regular basis (7% annually, 34% monthly, and 13% weekly) and 31% (n = 23) have had their blood drawn on an irregular basis. Eighty-three percent (n = 60) of the children were accompanied by their mothers, 9% (n = 7) were accompanied by their fathers, and 8% (n = 6) by both their mothers and fathers, and 77% (n = 56) were told in advance that they would be having their blood drawn.
Measures Children's Perceived Control In reviewing the literature specific to the assessment of children's perceived control, we were particularly drawn to the methodology developed by Connell (1985) becauses of its conceptual and scientific rigor. The Multidimensional Measure of Children's Perceptions of Control emerged from this work. Furthermore, in comparison to other existing measures of children's perceived control (e.g., Children's Multidimensional Health Locus of Control Scale) which pri-
Downloaded from http://jpepsy.oxfordjournals.org/ at University of Winnipeg on August 19, 2015
Subjects
764
Carpenter
Downloaded from http://jpepsy.oxfordjournals.org/ at University of Winnipeg on August 19, 2015
marily assesses generalized expectancy of perceived control, Connell's methodology was designed to assess children's domain-specific perceptions of control. The primary focus of Connell's research using this assessment methodology has been on the investigation of school-aged children's perceptions of perceived control specific to domains within the school environment (i.e., cognitive, social, and physical; Connell, 1985). Connell encouraged investigators to modify his items to reflect the specific domain being assessed. Thus, the Multidimensional Measure of Children's Perceptions of Control represents a structured methodology rather than a fixed scale. More specifically, this methodology standardizes the assessment of three sources of perceived control (Internal, Powerful Other, and Unknown) for any specified domain of interest. Two items are used to measure each perceived source of control. The items are worded identically across each specific domain of interest and are different only in reference to the specific domain being sampled. A response set ranging from 1 to 4 (very true, sort of true, not very true, not at all true) is used in scoring all items. A set of flexible, yet standardized, guidelines for the administration of the assessment methodology is recommended to ensure maximum rapport. Connell's experience in administering this scale indicates that the most valid results emerge when a flexible approach is used. A modification of the Multidimensional Measure of Children's Perceptions of Control consisting of 22 items, using the methodology described above, was used to assess children's source(s) of perceived control to the impending medical procedure. The modification involved a rewording of each of the two items within each source of perceived control (Personal, Powerful Other, and Unknown) specific to the target medical procedure being assessed (e.g., when my mother or father are in the room when I get my blood drawn, I can stay still). Consistent with the way we operationalize procedure-related distress (see discussion below for more detail) all perceived control items target "staying still" as the primary coping behavior. In addition, the 10-item Rosenberg Self-Esteem Scale (Rosenberg, 1979) is embedded into the questionnaire to assess children's perceived self-competence. Assessment of self-competency was included because of the importance of this construct within learned helplessness theory (Abramson et al., 1978, 1980). Five of the items assess competencies in the positive direction and 5 in the negative direction and are summed separately to derive a total positive self-esteem score and a total negative self-esteem scale. Because we wished to assess the construct of perceived source(s) control in children as young as 4 years of age, the developmental appropriateness of all items were evaluated by two experts in developmental psychology. Following additional editing to further refine the developmental appropriateness of items, 100% agreement between these two experts was achieved. The final form of the modified scale was piloted on 10 children between the ages of 4 and 7 to assess
Perceived Control
765
Anticipatory Procedure-Related Distress The assessment of children's anticipatory procedure-related distress was obtained from three sources: Children's Ratings. Children's anticipatory self-ratings of fear and pain with respect to the impending medical procedure were measured using the Children's Global Rating Scale (CGRS; Carpenter, 1990). Because of the difficulties in measuring complex psychological constructs, such as fear and pain, across diverse developmental populations, different measuring techniques have often been used for younger and older children posing serious problems to the internal and external validity of study results. To address this methodological problem, we developed the Children's Global Rating Scale, a self-report methodology for measuring children's perceptions of such constructs as pain and fear (Carpenter, 1990). A series of studies (Carpenter, 1990, 1991) with 145 children between the ages of 4 and 17 undergoing invasive medical procedures examined: (a) the comparability of the CGRS with another commonly used self-report method for assessing pain and fear in young children—the Faces Scale; (b) the convergent and predictive validity of the CGRS; and (c) applicability of the CGRS for assessing the constructs of pain and fear across the developmental span. The results of these investigations indicated that (a) children's CGRS fear and pain ratings had consistently significantly (p < .05) greater overall Pearson correlations (ranging from .30 to .54) with other measures of procedure-related distress obtained from parents, clinicians, and a trained clinical observer than their similar ratings using the Faces Scale; (b) children's anticipatory CGRS fear ratings with respect to an impending medical procedure emerged as a significant predic-
Downloaded from http://jpepsy.oxfordjournals.org/ at University of Winnipeg on August 19, 2015
the applicability and validity of administering it to young children. Because the scale asks the same question in two different ways for each perceived source of control assessed, the pilot administration revealed that children as young as 4 years of age were capable of validly reporting their attributions regarding perceived source of control if the items are understandable to them. A study by Zeltzer et al. (1988) indicated that young children are capable of using rating scales. We also administered the modified scale in a standardized, yet flexible, manner. All children were administered the scale on an individual basis by a trained research assistant with the child's parent(s) instructed not to coach or answer. Because of the time limitations placed on us by the Phlebotomy Department, all scale items were presented orally to every child participating in the study. A standardized response format card was given to each child to assist in remembering all possible response choices and to control for response preservation.
766
Carpenter
Downloaded from http://jpepsy.oxfordjournals.org/ at University of Winnipeg on August 19, 2015
tor of those children would extend or interfere with the timely administration of the medical procedure; and (c) no significant differences were found between the CGRS fear and pain ratings of children s 7 years of age and children 5: 8 years of age. The rationale, development, administration, and initial empirical evaluation of the CGRS has been discussed in detail elsewhere (Carpenter, 1990). Parent Ratings. Parents' rated their perceptions of their children's anticipatory fear on a 5-point scale ranging from not at all (0) to extremely (4). Parents also rated how scared or apprehensive they were themselves regarding the medical procedure using the same 5-point scale. This global rating scales is identical to the one used in our other investigations of children's distress during medical procedures (Carpenter, 1990). Clinical Observer. A behavior checklist, comprising 21 behavioral items grouped within six domains: Verbal (e.g., I'm scared); Vocal Nonlanguage (e.g., crying), Facial Expressions (e.g., scowling); Body Movements (e.g., flail); Physiological Manifestations (e.g., clenched fist); and General Response to Environment (e.g., hypersensitivity to environmental stimuli) was used to assess the behavioral components of children's distress in the anticipatory phase of the medical procedure. The anticipatory phase began when a child entered the waiting room of the Pediatric Phlebotomy Department and ended when he/she was called by a phlebotomist to enter the procedure room. To assure comparability, the specific behavioral content contained in other behavior observation scales for assessing children's behavioral distress during invasive medical procedures was incorporated into the behavior checklist used in this study (Elliott, Jay, & Woody, 1987; Jay, Ozolins, Elliott, & Caldwell, 1983; Katz, Kellerman, & Siegel, 1980; LeBaron & Zeltzer, 1984). The scoring schema used for our behavioral checklist is identical to that employed by Katz et al. (1980). Specifically, the 21 behavioral items are coded with respect to their observed presence or absence (yes = 1; no = 0). A total behavior index score is calculated by summing of all the observed behaviors within a specific time phase. An interrater reliability check of this behavioral checklist with the initial 20 subjects revealed a statistically acceptable level of interrater agreement (r = .92). Although more sophisticated behavioral sampling techniques and scoring schema have been used, the results of a study by Jay and Elliott (1984) revealed no significant improvement in either the sensitivity or validity of behavioral checklists using more sophisticated methodologies. Actual Procedure-Related Distress Consistent with our previous research in this investigative area, procedurerelated distress was operationalized as the extent to which a child's fearful or avoidance behavior interferes with or extends a specific medical procedure (Car-
Perceived Control
767
penter, 1991). In accordance with this operational definition, the phlebotomist conducting the medical procedure rated the extent to which a child's behavior interfered with or extended the time it normally takes to draw blood in children of a similar age independent of the time the child was in the procedure room using a 5-point scale ranging from not at all (0) to extremely (4). Our previous work has shown clinician's ratings of children's distress during the actual medical procedure to be significantly correlated with children's, parents', and a trained observer's ratings of anticipatory procedure-related distress (all rs, p > .30). Procedure
RESULTS Preliminary Analyses A median split was used to dichotomize the distribution of all 73 children's responses to each of the three sources of perceived control (Personal, Powerful Other, and Unknown) into a high/low group. The combination of high/low scores for each of the three sources of control placed each child into one of the
Downloaded from http://jpepsy.oxfordjournals.org/ at University of Winnipeg on August 19, 2015
The study protocol was designed to assess two phases of the target medical procedure: (a) Anticipatory Phase-—the time the child first entered the Phlebotomy Department; and (b) Actual Procedure Phase—time during which the procedure is performed. Upon entering the Pediatric Outpatient Phlebotomy Departments, children and their parents were approached by a research technician and asked to participate in the study. The study was explained as an investigation of the relationship between children's self-reported perceived control and procedure-related distress. Of the 80 children approached, 73 consented to participate. Time constraints were the reason the other 7 subjects declined to participate. Those who consented were entered on the protocol which was approved by the institutional IRB committee. Prior to having blood drawn, the child's parent(s) were asked some basic questions regarding their child's demographic background and whether or not their child was told about the medical procedure in advance. They also rated their child's anticipatory fear just prior to undergoing the medical procedure as outlined above. Parents were also asked to rate how apprehensive they were before the procedure. The research technician then administered the Perceived Control Questionnaire and the CGRS to the child assessing his/her anticipatory pain and fear. The clinical observer completed the preprocedural behavioral checklist. The phlebotomist completed a rating form once the procedure was over.
768
Carpenter
four perceived control types outlined in Table I. The results of this classification (see Table II) revealed that 19 (26%) children fell into the Mutual Control Group; 7 (10%) into Control From Powerful Others; 20 (27%) Personal Control; and 27 (37%) Unknown source of control. Table II presents the study data for selected demographic variables and measures of anticipatory and actual procedure-related distress obtained from children, parents, a trained clinical observer, and phlebotomist performing the medical procedure for the total sample (n = 73) and for each perceived control type. Main Analyses
Downloaded from http://jpepsy.oxfordjournals.org/ at University of Winnipeg on August 19, 2015
Because the literature in this area has consistently found children's age to be a strong predictor of observable procedure-related distress (Jay, 1988), a group (Mutual, Powerful Other, Personal, Unknown) by age (^7; s 8 ) analytic model was used in the evaluation of all study data examining the relationships between procedure-related distress and children's perceived sources of control. We chose to dichotomize age in this fashion for comparability with other reports in the literature (e.g., see Jay et al., 1983). Analysis of variance (ANOVA) was applied to the statistical analysis of variables with a continuous level of measurements and the chi-square method with a categorical level of measurement (Table III). Demographic Differences. The results revealed no statistically significant differences between the four perceived control types and the selected demographic variables of age, sex, or race. Main Effects for Age. Children's CGRS fear rating was the only variable significantly related to age. Specifically, this one finding revealed that children 8 years of age. Main Effects for Group. As summarized in Table III, the statistical analysis of study data revealed main effects for group for all variables except the negative self-competence items and parents' own fear ratings. Statistical analysis (r and chi-square tests) of the specific differences between the Unknown perceived control group and the Mutual, Powerful Other, and Self perceived control groups were additionally performed to evaluate significant main effects for group. These results revealed that a significantly (p < .01) greater proportion of children in the Unknown perceived control group (41%) were rated by the phlebotomist performing the medical procedure as extending or interfering with the timely or efficient administration of the procedure than any of the other three perceived control types (Mutual = 21%; Powerful Other = 12%; Personal = 5%). In addition, children in the Unknown perceived control group rated themselves, were rated by their parents, and were observed to experience significantly (all ps < .05) greater anticipatory procedure-related distress than children in any of the
Demographic Age (in years) M (SD) Range Male [n (%)] Ethnic minority n (%) Children's self-report [M {SD)] Fear-CGRS Pain-CGRS Self-competence Positive items Negative items Parents' self-report Told child about procedure in advance [n (%)] Child's fear [M (SD)] Parental fear [M (SD)] Trained clinical observer Total [M (SD)] Phlebotomists' ratings Extend/interfere with the medical procedure [n (%)]
Variable
Downloaded from http://jpepsy.oxfordjournals.org/ at University of Winnipeg on August 19, 2015
17 (23)
1.36(2.5)
56 (77) 1.42 (1.3) 1.03(1.1)
17.31 (2.9) 11.25 (3.3)
1.21 (1.5) 1.56 (1.4)
10.00 (3.8) 4-18 38 (52) 17 (23)
Total sample (N = 73)
4(21)
0.37 (0.8)
16 (84) 1.17 (1.0) 1.11 (1.1)
18.7(1.7) 10.8 (2.9)
1.05 (1.4) 1.37(1.3)
8.2 (3.2) 4-16 8(42) 2(11)
Mutual control (n = 19)
1 (12)
1.71 (2.2)
6 (86) 1.29(1.6) 1.14 (1.7)
17.0 (2.2) 11.1 (3.9)
0.86(1.5) 2.00 (1.4)
10.4 (2.6) 7-14 2(29) 2(28)
Control from powerful other (n = 7)
Perceived control types
Table II. Descriptive Statistics for all Variables
1 (5)
0.70 (2.2)
18 (90) 0.82 (1.0) 0.65 (0.9)
17.9 (2.2) 11.3 (2.9)
0.65 (1.1) 0.80 (0.9)
10.9(3.5) 6-18 11 (55) 6(30)
Personal control (n = 20)
11 (41)
2.44 (3.4)
16 (59) 2.09 (1.3) 1.22 (1.1)
15.4 (3.7) 11.7 (3.9)
1.85 (1.6) 2.15(1.5)
10.5 (4.2) 4-18 17 (63) 7(26)
Unknown control (n = 27)
3
1
Conti
770
Carpenter Table m . Analysis for Group, Age, and Interaction Effects'1 Effects* Variable
Age
Group x Age
3.26* 4.45'
4.78rf 0.00
3.00 0.45
5.11' 0.87
0.00 0.27
0.49 9.99
11.86' 4.14' 0.76
0.24 1.27 1.09
2.64C 1.63 1.01
1.05
0.43
0.01
8.23
