Communication Study
Physician–child interaction: When children answer physicians’ questions in routine medical encounters

https://doi.org/10.1016/j.pec.2011.07.007Get rights and content

Abstract

Objective

The objective of the study is to examine predictors of children answering questions during primary care pediatric visits.

Methods

Relying on a sample of 322 video-taped community practice encounters, this study identifies predictors of when children answer physicians’ questions. Multi-level multivariate regressions were used to model the relationships among communication and socio-demographic variables and whether or not children answered questions pediatricians asked them.

Results

Whereas race and education predict whether physicians select children to answer questions, these factors are not associated with whether children answer physicians’ questions. Instead, a child's performance is associated with communication practices specific to physician–child interaction such as the grammatical type of question and doctor gaze.

Conclusion

Children are less responsive to physicians’ questions than their parents but their failure to answer is predictable and thus can be improved. By increasing their participation in the visit, physicians may (a) secure more information about children's health and (b) socialize children to be more pro-active patients.

Practice implications

Physicians can improve the likelihood that children will answer their questions by (a) asking them social questions early in the visit, (b) phrasing their questions as yes–no questions, and (c) and directing their gaze at the children during each question.

Introduction

In pediatric encounters, direct communication between physicians and children is important: it builds rapport, trust and is valued by child patients [1], [2]; it socializes the child into the role of patient—what they should know, in particular [3], [4]; and children can contribute substantially to these medical encounters [4], [5]. Children are uniquely positioned to provide particular types of information to physicians [6], [7], [8], [9], [10]. Although some studies argue that parents may be partially responsible for inhibiting their children's participation [4], [11], most studies show that outpatient health care providers spend a meager amount of time interacting with their child patients, with or without the collusion of a parent [5], [11], [12], [13], [14], [15], [16], [17]. Moreover, relatively little effort has gone into developing an account for the circumstances under which physicians become more or less likely to involve children in the medical encounter. Stivers and Majid, relying on a multi-method analysis of video-taped pediatric encounters, argue that physicians involve some children more than others and argue that physicians may be exhibiting an implicit bias against children of particular race and class backgrounds [15]. They argue that differences in children's sociodemographic backgrounds account for a substantial amount of the variation in whether children are selected by physicians to answer physicians’ questions.

Still, that study leaves open the question of whether physician question asking patterns are primarily the result of generalizations from different interactional experiences with children and parents from certain racial and socioeconomic backgrounds. More concretely, are children from lower SES backgrounds or of a minority race simply less likely to answer questions posed to them than their counterparts? If this were the case, physicians would likely internalize this and stop posing questions to children from these backgrounds leading to the findings offered by Stivers and Majid. By contrast, if children are no more likely to answer based on socio-economic or racial background, then we are left with a puzzle and should investigate what other sorts of predictors might explain child participation. This study identifies some of the factors that predict when children answer questions physicians pose to them and offers several ways in which children's participation can be increased.

Section snippets

Design

A nested cross-sectional design was employed, consisting of 570 videotaped pediatric encounters for children with upper respiratory tract infection symptoms clustered within 38 pediatricians in 27 community pediatric practices around Los Angeles County conducted between October, 2000 and June, 2001. In the course of analyzing the data with regard to how parents and physicians arrive at a treatment decision [18], [19], [20] – the reason for the constraint on URTIs – it became clear that the

When children answer questions

The results of the multivariate logistic regression are shown in Table 6 as odds ratios with 95% confidence intervals. Of the different levels that were tested, all were significant. Thus, just as Stivers & Majid found that some physicians are more likely to ask children questions, so too are some physicians more likely to secure answers from children. The visit level significance suggests that if a child answers a question s/he may be more likely to continue answering questions. There were six

Discussion

In two main respects physicians pitch their questions to interlocutors who are more likely to answer their questions: child age was a predictor of whether a physician selected a child to answer a question, of whether a child answered a question posed to them, and of whether a parent was willing to answer questions about them. These findings, combined with the observation that children were categorically less likely than adults to answer questions asked of them, supports the analysis that was

References (25)

  • P. Alderson et al.

    Health care choices: making decisions with children

    (1996)
  • K. Aronsson et al.

    Child discourse and parental control in pediatric consultations

    Text

    (1988)
  • Cited by (0)

    Author note: Portions of this paper were presented at the UCLA Center for Language and Culture 2007 Symposium on Socialization, Interaction and Culture and the 2007 International Meeting on Conversation Analysis in Clinical Encounters at the University of Exeter. Thanks to Ignasi Clemente and John Heritage for comments on earlier drafts of this manuscript.

    View full text