Physician Perceptions of Children’s Coping with Asthma are Associated with Children’s Psychosocial and Disease Functioning

Abstract

Introduction

Low-income, ethnic minority children disproportionately face poor asthma control, and poorly controlled asthma is related to psychosocial difficulties. This study assessed physician reports of coping in child patients, and examined associations between physician reports of child coping and parent and child reports of children’s coping, psychosocial and asthma outcomes (asthma-related stress, emotional and behavioral problems, asthma control, and school missed due to asthma).

Methods

Physicians reported on coping in their patients (N=67) ages 5–17 with asthma. Parents reported on child coping, asthma-related stress, emotional and behavioral problems, asthma control, and school missed due to asthma. Children ages 9–17 provided self-reports.

Results

Physicians’ reports of primary control coping (e.g., problem solving) and secondary control coping (e.g., cognitive restructuring) were not associated with parent ratings of corresponding coping strategies, but physician reports of disengagement coping (e.g., avoidance) were correlated with parent reports of disengagement and secondary control coping. Physician perceptions of higher child primary control, and lower disengagement, were correlated with less parent-reported stress, better asthma control, and for primary control, fewer partial days of school missed. Physician reports were not correlated with child reports of coping, but physician reports of disengagement were correlated with child-reported conduct problems.

Discussion

Findings suggest that physician reports of child coping provide independent information from parent and child reports of coping, and could be leveraged to identify and intervene with patients who are at elevated risk for poor outcomes.

Asthma is a chronic respiratory disease that causes inflammation and narrowing of the airways and leads to wheezing, shortness of breath, coughing, and chest tightness (NIH, 2014). Over 25 million people in the U.S. are affected by asthma, and nearly 10% of children in the U.S. have the disease (Akinbami et al., 2012), with the highest rates among non-White children living in inner-city, low-income environments (CDC, 2011; Claudio et al., 2006; Gupta et al., 2008). While not curable, asthma symptoms can be controlled with medication and children can participate in daily activities without limitation. Children with asthma are at risk for psychosocial difficulties (i.e., difficulties in multiple domains of child functioning, including emotional, behavioral, social, and academic domains), particularly when asthma is not well controlled. Childhood asthma is associated with stress (Wright et al., 1998) and emotional and conduct problems (Halterman et al., 2006; Katon et al., 2007; McQuaid et al., 2011), and almost half of children with asthma report missing at least one school day each year (CDC, 2013). Psychosocial difficulties also contribute to poorer asthma control and morbidity (Hasler et al., 2005; Tibosch et al., 2011; Weil et al., 1999). Low-income children of color have high rates of uncontrolled asthma (CDC, 2013), but psychosocial research with this population is limited.

Multi-informant Approaches to Assessing Child Functioning

Multi-informant data is considered the gold standard for assessing children’s psychosocial functioning (Achenbach, 2006; Achenbach et al., 1987; De Los Reyes & Kazdin, 2005). Clinicians may have unique observations of child psychosocial functioning that may be useful to identify patients at risk for more negative outcomes, independently from parent and child reports. Research suggests that clinicians generally show only small to medium levels of agreement with child and parent ratings of child behavior. A seminal meta-analysis showed the average correlation between parent and mental health clinician ratings of child behavior was r = .24, and between child and clinician ratings was r = .27 (Achenbach et al., 1987).

Physician reports of child functioning

Research on physician perceptions of patient coping has been limited, but suggests that agreement between physicians and patients (or patients’ parents) is relatively low for coping and other psychosocial variables such as quality of life and emotional distress (Braido et al., 2012; Janse et al., 2008). For children with asthma, studies suggest low agreement between parent and physician ratings of asthma control, and physician ratings are more strongly correlated with physiological measures (Hammer et al., 2008; Levy et al., 2004; Uysal Soyer et al., 2012). These discrepancies between physicians and patients/parents may indicate that each reporter has a unique perspective on child functioning, each of which is valuable for identifying risk and intervention targets (De Los Reyes & Kazdin, 2005). Despite the potential benefits of utilizing physician reports, they are rarely examined as a potential source of information about children’s psychosocial functioning, and little research has examined physician reports in relation to child and parent reports.

Children’s Coping with Asthma

Children’s coping is particularly relevant for their psychosocial and asthma outcomes. Coping refers to conscious, volitional efforts to regulate oneself and/or the environment in response to stress (Compas et al., 2001). Contemporary research supports an empirically-validated theoretical model containing three dimensions of coping: primary control, secondary control, and disengagement (Compas et al., 2006; Connor-Smith et al., 2000; Wadsworth et al., 2005; Yao et al., 2010). Primary control coping involves directly changing the stressor or one’s emotional response to the stressor (e.g., problem solving to establish a regular schedule for taking asthma medication), secondary control coping involves adaptation and accommodation to the stressor (e.g., cognitive reappraisal, such as thinking about something positive that has come from having asthma), and disengagement coping involves avoiding the stressor or one’s reactions to it (e.g., avoiding people or things that remind you of having asthma). In children, primary and secondary control have been associated with more positive asthma management behaviors and outcomes (Chen et al., 2011; Koinis-Mitchell & Murdock, 2002; Schreier & Chen, 2008) and better psychosocial functioning (Rodríguez, Kumar, Draeger, & Sánchez-Johnsen, 2016). Secondary control appears to be especially beneficial for urban, low-income, ethnic minority youth with asthma (Chen et al., 2011; Rodríguez et al., 2016). Primary control may also be beneficial, with certain primary control strategies (e.g., problem solving) associated with better asthma management practices (Koinis-Mitchell & Murdock, 2002; Schreier & Chen, 2008). Findings on disengagement coping indicate it is associated with poorer psychological functioning and health-related quality of life (Marsac et al., 2007; Peeters et al., 2008), although in conditions of low stress or when combined with other coping strategies, disengagement has been associated with more positive outcomes (Greene et al., 2006; Koinis-Mitchell & Murdock, 2002). Overall, coping is an important indicator of risk for psychosocial and disease outcomes in children with asthma, but current research is based solely on parent and child report.

The Current Study

This study examined physician ratings of children’s coping with asthma. The children in our sample were primarily low-income, urban, and racially/ethnically diverse, representing a population at elevated risk for poorer asthma outcomes. Our research questions were: 1) What are the correlations between physician ratings of children’s coping, and child and parent ratings of child coping? and 2) Are physician ratings of child coping significantly correlated with child and parent reports of child psychosocial and asthma outcomes? Based on previous research on cross-informant correlations, we expected that physician-parent and physician-child correlations would be small to medium. We also hypothesized that physician ratings of higher primary and secondary control coping, and lower disengagement coping, would be correlated with fewer child emotional and behavioral problems, lower asthma-related stress, better asthma control, and less school missed due to asthma, even after accounting for demographic factors associated with physician reports of coping.

Method

Participants

Parents

Parents were drawn from a larger cross-sectional study of 78 families that examined adjustment to asthma; details regarding recruitment methods and consent rates are provided elsewhere (Rodriguez, Kumar, Alba-Suarez & Sánchez-Johnsen, in press; Rodriguez et al., 2017). Sixty-seven parents/caregivers of children ages 5 to 17 years old with asthma were included in the current study (see Table 1 for demographic information). The current sample is slightly smaller than the overall cross-sectional study because we were not able to obtain physician reports for eleven patients (we could not distribute forms to physicians right after these patients’ appointments if the physician was not available).

Table 1.

Participant Characteristics

Characteristics	Child (N = 37)	Parent (N = 67)
Age	M= 12.1 years SD=2.4	M=38.01 years SD=8.3
Female	35% N=13	91% N=67
Race
Black	46% N=17	49% N=32
White	16% N=6	17% N=11
Asian	3% N=1	3% N=2
Native American	3% N=1	1% N=1
Other/More than One/Did not specify	32% N=12	30% N=20
Ethnicity
Latino	41% N=15	39% N=26
Specific Latino Ethnicity
Mexican	74% N=11	81% N=21
Puerto Rican	8% N=2	8% N=2
Honduran	0% N=0	3% N=1
Other/More than One/Did not specify	8% N=3	8% N=2
Family Income
<$25,000		63% N=40
$25,000–50,000		33% N=21
>$50,000		4% N=3
Language Completed Questionnaire
Spanish	5% N=2	19% N=13
English	95% N=35	81% N=54
Marital Status
Single/Divorced/Separated		58% N=39
Married/Remarried/Living with Partner		42% N=28

Note. Percentages are approximated to total 100%.

Children

In the overall cross-sectional study (Rodriguez et al., 2017), we collected self-report data from 42 children ages 9 to 17. We excluded children under age 9 from providing self-report data because some of our self-report measures did not have evidence for reliability with younger children. Of these 42 children, we included 37 children in the current study because physicians did not provide data for 5 children from the overall cross-sectional study.

Physicians

The participating physicians (N = 3) represented all of the board certified pediatric pulmonologists/allergists treating children at an asthma and allergy specialty clinic.

Procedure

We recruited parents and children from an urban university hospital outpatient clinic specializing in pediatric asthma and allergies in the Midwestern United States from February to April 2014. Children are typically referred to this clinic by their primary care clinician due to uncontrolled moderate to severe persistent asthma. Parents were eligible if they had a child between the ages of 5–17 years old who had a diagnosis of asthma, the child was receiving treatment at the clinic, and they were able to read or speak English or Spanish; their children ages 9–17 were also eligible. Adult participants provided informed consent; children provided assent. We compensated parents $15 and gave children a $5 gift card for participating. Parents and children completed the study at the clinic before or after the child’s outpatient appointment. Once the parent and child had completed the study, we provided the child’s physician with a form to rate that child’s coping. Typically, the physician completed the form the same day as the child’s appointment. The university’s Institutional Review Board approved the study.

Parents completed measures on demographic information, child coping, child asthma-related stress, child emotional and conduct problems, asthma symptom control, and school attendance. Children completed self-report measures on asthma-related stress, coping, emotional and conduct problems, and asthma control. Physicians completed reports of child coping. We gave parents and children the option to complete measures on their own as written questionnaires or orally as an interview with a research assistant; most chose to complete measures in writing. Physicians completed all forms as written questionnaires.

Measures

Family demographics

We collected child age, race, and ethnicity and parent age, race, ethnicity, family income, marital status, and years of education on a demographic form.

Coping

The Responses to Stress Questionnaire-Pediatric Asthma version (RSQ; Connor-Smith et al., 2000) measured children’s coping with asthma. The RSQ includes 30 items in which participants indicate on a 4-point scale how much they use various coping strategies. The RSQ yields three coping scales: primary control, secondary control, and disengagement. Items include “I try to think of different ways to change or fix the situation” (primary control), “I tell myself that things could be worse” (secondary control), and “I try not to think about it, to forget all about it” (disengagement). Internal consistencies for the current sample were good: α = .82 for primary control, α = .90 for secondary control, and α = .82 for disengagement for the parent report and α = .84 for primary control, α = .82 for secondary control, and α = .88 for disengagement for the child report.

Measure adaptation for physicians

Based on the RSQ, we designed a brief measure for physicians to report child coping. To keep the form brief, we grouped the 30 coping items from the original RSQ into 10 broader items about the use of these coping strategies. We kept the item wording the same as the original RSQ, but listed multiple RSQ items as bullet points within a single item. For example, on the original RSQ there are three items about problem solving; on the physician form, these 3 items were listed as bullet points within a single item, and we asked physicians to rate the frequency of these multiple problem solving strategies as a whole. The questionnaire was designed this way so physicians could complete it in a shorter amount of time for each patient. As noted above, all physicians in the clinic agreed to participate and they were able to complete the measure for the majority (86%; i.e., 67 of 78) of patients from the larger cross-sectional study, providing some initial evidence that the adapted measure was acceptable and feasible. However, the adaptations made to the original RSQ likely contributed to lower internal consistency for the physician version (α = .60 for primary control, α = .58 for secondary control, and α = .64 for disengagement).

We used proportion scores (dividing the score for each factor by the total coping score) in the current analyses to indicate the proportional use of primary control, secondary control, and disengagement (see Connor-Smith et al., 2000; Vitaliano et al., 1990).

Asthma-related Stress

The stressor items from the RSQ measured the child’s level of asthma-related stress (internal consistency was α = .87 for parents, and α = .85 for children). This includes a list of 10 asthma-related stressors (e.g. “not being able to do things because of asthma;” “having to go to the hospital, clinic, or doctor’s office so often;”), rated on how stressful each item has been recently on a scale from 1 (Not at all) to 4 (Very).

Emotional and Conduct Problems

We measured emotional and conduct problems using these two subscales of the Strength and Difficulties Questionnaire, a brief and well-validated questionnaire to assess mental health difficulties in children within the past six months (SDQ; Goodman, 1997). Internal consistencies were α = .74 for each scale for parents, and α = .65 for emotional problems and α = .55 for conduct problems for children. Both scales consist of five items, with higher scores indicating greater problems. Emotional items include: “worries a lot,” and “often unhappy, depressed or tearful.” Conduct items include: “get[s] very angry and often lose[s] temper,” and “often accused of lying or cheating.”

Asthma symptom control

We used the symptom control subscale of the Asthma Therapy Assessment Questionnaire (ATAQ; Skinner et al., 2004) to assess parent-reported asthma control (e.g., “wheezing or difficulty breathing” in the last four weeks). Higher scores indicate poorer asthma control. Internal consistency for the current sample was good (α = .81). We used the 5-item Asthma Control Test (ACT; Nathan et al., 2004) to assess child self-reported symptom control and functional impairment due to asthma (e.g., “how much of the time did your asthma keep you from getting as much done at work, school or at home”). Higher scores indicate better asthma control on the ACT. Internal consistency of the ACT for the current sample was α = .70.

Asthma-related school absenteeism

Parents reported the number of full days and partial days (i.e., left early or arrived late) of school that their child missed in the past three months due to asthma.

Translation of Measures

All measures were translated into Spanish, and Spanish versions were pretested and refined based on feedback of bilingual research staff. The demographic questionnaire and the RSQ were adapted from similar versions translated and backtranslated by the first author and other bilingual researchers using the iterative process described by Brislin (1970). The Spanish versions of the SDQ (Gómez-Beneyto et al., 2013; Hernandez et al., 2012) and the ACT (Rodrigo et al., 2008; Vega et al., 2007), were previously translated and validated with Spanish-speaking participants. Merck & Company, Inc. distributes the Spanish version of the ATAQ.

Data Analyses and Statistical Power

We used SPSS version 23 to conduct our analyses. We first examined relations between demographic variables and physician ratings of child coping. Because we collected children’s race (i.e., White, Black, Asian, Native American, or Other/More than one race) and ethnicity (i.e., Latino or non-Latino) as separate variables, we recoded these separate variables into one race/ethnicity category. Latino participants of any race were coded as “Latino;” if participants were non-Latino, they were coded as their race. To test our hypotheses, we conducted bivariate correlations to examine associations between physician and parent and child ratings of child coping, and associations between physician coping ratings and parent and child ratings of child stress, emotional and conduct problems, asthma control, and school missed. This analytic approach is consistent with prior research on multi-informant perspectives (Achenbach et al., 1987). If physician reports of coping were associated with demographic variables, we conducted regression analyses in which we controlled for demographic variables at Step 1 of the model, and entered physician ratings at Step 2. To examine whether the non-independence of physician ratings (i.e., the same physician rating multiple patients) impacted results, we also conducted analyses controlling for individual physician; our results were the same as when not controlling for this variable. Therefore, to maintain statistical power, we did not include individual physician as a control variable. Post-hoc power analyses conducted with G*Power version 3.1 (Faul, Erdfelder, Buchner & Lang, 2009) indicated that our sample of 67 parents yielded power of .93 to detect statistical significance with α = .05 for two-tailed correlations of ≥ 0.4, and .70 to detect correlations of ≥ 0.3. Our sample of 37 children yielded power of .71 to detect significance for two-tailed correlations of ≥ 0.4, but .44 to detect smaller correlations (≥ 0.3).

Results

Associations between Physician reports of Child Coping and Demographic Variables

Physician reports of secondary control coping were positively correlated with child age (r = .24, p < .05). There were no other significant associations between physician reports of coping and demographic variables.

Correlations Among Physician, Parent and Child Reports of Child Coping

Physician reports of primary control and secondary control were not significantly correlated with corresponding types of coping based on parent or child report (r’s from .02 to .20; p’s > .05). Physician reports of disengagement coping were positively correlated with parent reports of disengagement (r = .28, p < .05) and negatively correlated with parental reports of secondary control (r = −.30, p < .05), but not child reports of coping (see Table 2).

Table 2.

Descriptives and Correlations Among Physician, Parent, and Child Reports of Children’s Coping

	M (SD)	1	2	3	4	5	6	7	8
1. MD Report PCC	.30 (.06)	–
2. MD Report SCC	.43 (.07)	−.37**	–
3. MD Report DC	.28 (.07)	−.47**	−.66**	–
4. Parent Report PCC	.32 (.04)	.02	−.05	.03	–
5. Parent Report SCC	.42 (.05)	.20	.15	−.30*	−.38**	–
6. Parent Report DC	.26 (.06)	−.19	−.13	.28*	−.43**	−.67**	–
7. Child Report PCC	.30 (.04)	−.08	−.03	.09	.31+	−.11	−.18	–
8. Child Report SCC	.43 (.05)	−.02	.13	−.10	−.18	.34*	−.15	−.43**
9. Child Report DC	.26 (.05)	.08	−.10	.03	−.09	−.24	.31+	−.43**	−.63**

Note. MD = Physician. PCC = primary control coping. SCC = secondary control coping. DC = disengagement coping. Coping scores are proportion scores. N = 67 for parent-physician correlations; N = 37 for child-parent and child-physician correlations.

⁺p < .10;

^*p < .05;

^**p < .01.

Associations Between Physician Perceptions of Coping and Child Psychosocial Adjustment and Asthma Outcomes

Physician reports of primary control were negatively correlated with parental reports of children’s asthma-related stressors (r = −.24, p < .05), poorer asthma control (r = −.33, p < .01), and partial days of school missed due to asthma (r = −.29, p < .05). Physician reports of disengagement were positively correlated with parental reports of asthma-related stressors (r = .29, p < .05) and poorer asthma control (r = .30, p < .05), and child self-reports of conduct problems (r = .37, p < .05). Physician reports of secondary control were not significantly correlated with child outcomes (see Table 3). Since physician reports of child secondary control were correlated with child age, we also conducted regression analyses to examine the association of secondary control with each outcome after accounting for child age. Physician reports of child secondary control were not significantly associated with psychosocial or asthma outcomes after accounting for child age in the regression models.

Table 3.

Descriptive Values of Parent and Child Reports of Children’s Psychosocial Adjustment and Asthma Outcomes and Correlations With Physician Reports of Children’s Coping

Parent Report of Child Outcomes		Physician Report of Child Coping
	M (SD)	Primary Control	Secondary Control	Disengagement
Asthma-related stress	1.92 (0.70)	−.24*	−.10	.29*
Emotional problems	2.39 (2.23)	−.15	−.04	.16
Conduct problems	1.69 (2.02)	−.12	−.11	.20
Asthma control (ATAQ)	2.68 (2.44)	−.33**	−.03	.30*
Full School Days Missed	2.44 (3.11)	−.15	.13	.01
Partial School Days Missed	0.85 (1.43)	−.29*	.05	.19
Child Self-Report of Outcomes
Asthma-related stress	1.86 (0.68)	−.21	−.02	.18
Emotional problems	3.14 (2.33)	−.06	−.22	.26
Conduct problems	2.22 (1.83)	−.27	−.16	.37*
Asthma control (ACT)	18.38 (4.21)	.27	−.21	.00

Note. N = 67 for parent-physician correlations; N = 37 for child-parent and child-physician correlations. ATAQ = Asthma Therapy Assessment Questionnaire; higher scores indicate poorer control. ACT = Asthma Control Test; higher scores indicate better control.

^*p < .05;

^**p < .01.

Discussion

Discrepancies among different reporters are often noted in child-focused research for a range of constructs, including youth stressors (Kusher & Tackett, 2017), symptoms of disorders (McDonald et al., 2016; Narad et al., 2015; Nugent, Kline, Thompson, Reeves & Shiffman, 2013) and target problems in mental health treatment (Hawley & Weisz, 2003). Consistent with this literature, we found a significant medium-sized positive correlation between physician and parent reports of disengagement coping, and a negative correlation between physician reports of disengagement and parent reports of secondary control, while physician reports of children’s primary and secondary control coping showed small to medium non-significant correlations with child and parent reports. This result suggests that disengagement coping may be subject to less discrepancy among informants than other types of coping. Similar to research showing greater agreement for externalizing compared to internalizing problems among different informants (Achenbach et al., 1987), it may be that disengagement coping behaviors are more naturally observable than primary or secondary control behaviors, or that the connection between disengagement coping and poorer functional outcomes is more apparent to physicians.

In addition, we found that physician-reported disengagement coping was associated with poorer asthma control, higher asthma-related stress, and conduct problems, while physician perceptions of higher primary control coping were associated with better asthma control, lower asthma-related stress, and fewer partial school days missed. These findings indicate that physician perceptions of children’s coping are associated with clinical outcomes. Physician judgments about their patients’ coping could be used to identify high-risk patients as well as patient strengths (e.g., high levels of primary control) to be supported within a patient-centered integrated care model. For example, children identified by physicians as low in primary control or high in disengagement could be referred for targeted interventions such as problem solving therapy (Malouff, Thorsteinsson & Schutte, 2007). Those already high in primary control could receive brief supportive interventions (i.e., embedded within a primary care visit) as needed, and could even be trained to provide intervention to higher risk patients as part of peer-based programs (e.g., peer coping-skills training; Prinz, Blechman, & Dumas, 1994).).

Strengths of this study include the multi-informant design, the innovative use of physician reports of child coping, the diverse sample, and the direct implications for clinical practice. Limitations include the small child sample, the limited number of physicians who participated, the low internal consistency of some child and physician measures, the reliance on parents for attendance data (which may have been subject to recall bias), and the lack of physiological measures of asthma control. Future studies can address these limitations by assessing a larger sample of youth and physicians from both primary and specialty care, and incorporating data from schools to enhance the validity of attendance data. Future research should also seek to improve the internal consistencies of the physician report measure using measurement development techniques (e.g., item response theory to select the most relevant items) and assess feasibility and acceptability of the measure in greater detail (e.g., by administering it to a larger sample of physicians and examining rates of survey completion and satisfaction).

Conclusions

Our results suggest that physician perceptions of patient coping are meaningful, are largely independent from parent and child perceptions of coping, and have clinical utility. Future research should examine how physicians learn to judge their patients’ coping (e.g., the questions they ask and the behaviors they observe during appointments) and incorporate this training into medical education. Mental health services are increasingly integrated into primary and tertiary care settings. Physicians’ ability to identify high-risk youth may facilitate screening and intervention, and benefit both psychosocial well-being and disease outcomes for these youth.