Doctor-parent-child relationships: a 'pas de trois'.
A girl with asthma spent years undergoing unnecessary, harmful, and potentially lethal treatment because doctors trusted her parents too much. Doctor-parent-child relationships: a 'pas de trois'. Tates K(1), Elbers E, Meeuwesen L, Bensing J. Author information: (1)Netherlands Institute for Health Services. Article · Figures & Data · Supplemental · Info & Metrics · Comments *Children's Mercy Bioethics Center and Department of Pediatrics, University of Missouri- Kansas, . It could also seriously damage or end the physician-parent relationship.
Two subdomains of longitudinality that cover these facets of the patient-provider relationship are strength of affiliation with the provider and the interpersonal relationship. Child, adult, and provider versions of the instrument are available, and the reliability and validity are reported elsewhere. The subdomain was scored with an algorithm based on 3 questions. A higher value was assigned if the provider was 1 identified as the regular source of care and was also 2 the provider who knew the child best and 3 the provider from whom care would be sought for a new problem.
Respondents who identified a provider for only 1 question were considered to have poor strength of affiliation and were given the minimum score of 1, as were respondents who identified different providers for each question. Those identifying the same provider for all 3 questions were considered to have high strength of affiliation and were given the maximum score of 4. Respondents who identified the same provider for 2 questions were given a score of 3 if the provider was the regular source of care and knew the child best and a 2 for any other combination.
The interpersonal relationship subdomain was evaluated through 14 questions asking parents to report on the interactions between the child, the family, and the health care provider. Questions asked, for example, about 2-way communication, knowledge of the family, and the interaction between provider and child. Responses were averaged to produce a single scale score.
A complete list of the questions is included Figure 1. View Large Download Complete list of survey questions on the relationships among children, families, and primary care providers.
Interpretation of the interpersonal relationship domain is slightly complex because of the small range of the scale scores Because of the large number of items in the scale and the small scale range, even a relatively small change in the mean score reflects a substantive difference in interpersonal relationship experience. Though the ability of the Pediatric Primary Care Assessment Tool to predict future outcomes has not yet been established, differences of a single point 1.
Though this may seem insignificant at first glance, most of the scores fall in the range of 2. Covariates We controlled for socioeconomic status family income, maternal employment, and maternal educationcharacteristics of the health care system provider specialty and cost sharingand demographics child's age and sex and public vs private insurance coverage.
Covariates were selected from previous studies examining primary care quality, use, and satisfaction—factors inherent to a sustained patient-provider relationship. The effects of also controlling for health status and naming a provider vs a place as a regular source of care are described in the "Comment" section.
Analysis The analyses tested the significance of the association among the 3 managed care policies and longitudinality scores for each of the 4 racial and ethnic groups separately, while adjusting for study covariates. To further explore which racial group was most affected by the managed care policies, we then tested the difference in mean longitudinality scores across the racial and ethnic groups for both self-reported and validated managed care characteristics.
First, frequencies of the study variables including socioeconomic status, health system factors, and demographic characteristics were assessed across racial and ethnic groups. Mean longitudinality subdomain scores were reported across racial and ethnic groups, and the significance was assessed using a standard t test. Second, generalized linear model procedures were used separately for each racial and ethnic group to assess differences in longitudinality across managed care characteristics, while adjusting for covariates.
Bonferroni t tests were used to test the significance of the difference in mean scores between "yes" and "no" responses to a particular managed care policy, while accounting for multiple comparisons.
These analyses were done separately for the self-reported and validated managed care data. Health status and having a provider vs a place as the primary source of care were controlled for in additional models described in the "Comment" section. Third, we assessed differences in mean longitudinality scores across racial and ethnic groups, adjusting for covariates. To control for the effects of managed care, we used the same generalized linear model procedures to compare scores across races, but only for those who reported "yes" to a managed care policy.
This was done for each policy, as well as for both self-reported and validated insurance information, using white race as the reference group. The mean values that are presented in the tables for across-policy comparisons naturally differ from the mean values of across-race analyses because of the process of adjustment. These differences do not affect the direction or overall magnitude of the findings; therefore, only the across-policy means are presented.
Results Table 1 shows the unadjusted demographic characteristics, socioeconomic status, and health system factors of our analytic sample by racial and ethnic group. Most of the respondents Asians and whites, however, were more likely to report a particular provider Black respondents had the lowest proportion reporting any cost sharing Asians were the most likely to report cost sharing There were no significant differences in child's age, sex, or health status across groups.
Table 1 also compares the frequencies of self-reported and validated managed care policies and longitudinality scores by race and ethnicity. Hispanics were the most likely When compared with the validated data, all racial and ethnic groups overestimated the prevalence of this policy. There were no significant differences in self-reporting of the requirement to seek a referral for specialty care or to stay within the provider network. Compared with the validated data, however, all racial and ethnic groups underestimated the prevalence of the policies.
The validated data suggest that Hispanics and blacks were the racial and ethnic groups most likely to be required to seek referrals and stay within a provider network. Hispanics had the highest frequency of the 2 policies Verification of the self-reported insurance information with the validated data revealed that The longitudinality subdomain scores also varied according to race and ethnicity.
Whites had the greatest strength of affiliation and interpersonal relationship scores 3. Hispanics, followed by blacks, reported significantly lower affiliation scores 3.
And Asians and Hispanics had the lowest relationship scores 3. More detailed analyses not shown reveal that racial disparities are particularly large for 2 specific items in the interpersonal relationship domain. Asian, black, and Hispanic respondents, compared with whites, had shorter-duration relationships with their regular sources of care. Similarly, Asians and Hispanics were, on average, leaning toward "probably" changing providers if it were easy to do mean scores, 2.
Table 2 compares the associations among the 2 measures of longitudinality and the 3 managed care policies for each race and ethnicity, adjusting for socioeconomic status, health system factors, and demographic characteristics. When data were reported by respondents, there were no significant differences in the affiliation subdomain for any racial and ethnic group across these policies. Although the differences were not significant, there was a consistent overall trend of lower affiliation scores associated with reporting "yes" to any of the restrictive policies for all groups.
These 3 policies were particularly correlated with reductions in the relationship scores for Asian and Hispanic children. Requiring patients to sign up with a PCP was associated with a 0. The requirement to stay within a provider network was associated with nearly equivalent reductions for Asians 0.
When parents disagree with doctors on a child's treatment, who should have the final say?
Asians further reported a 0. Surprisingly, both white and black children did not experience a reduction in the relationship score when the managed care policies were present. With regard to data validated by health plan representatives, there were still few significant differences in the affiliation domain Table 2. Black children were a major exception, reporting a 0.
What is Munchausen syndrome by proxy?
Though still nonsignificant for most races, the trend of lower affiliation being associated with more restrictive policies continued and was slightly enhanced with the validated insurance data. According to the validated data, managed care policies are strongly negatively associated with minority but not white children's interpersonal relationships with their physicians.
The specialist referral policy was associated with reductions in relationship scores of 0. Similarly, the requirement to stay within a provider network was associated with deficits in relationship scores of 0.
Doctor-parent-child relationships: a 'pas de trois'.
In addition, black children who were required to sign up with a PCP reported a 0. Table 2 demonstrates that, controlling for managed care policies, racial and ethnic minorities experience poorer longitudinality of care.
The self-reported insurance data show that Hispanics and blacks report much lower strength of affiliation than whites, regardless of the managed care policies. Asian and Hispanic children also experience poorer interpersonal relationships with their primary care physicians than whites. Comment The findings of our community-based study advance the literature on MCOs and the patient-provider relationship for children by demonstrating that managed care policies restricting patient flexibility in seeking medical care are negatively correlated with the parent-reported interpersonal relationship between child and provider.
Racial and ethnic minority children experienced much larger deficits in longitudinality under these policies than white children do, even after controlling for socioeconomic status, health system factors, and demographic characteristics. The policy requiring patients to seek care within a network of physicians was associated with the largest deficits in longitudinality of care.
Both self-reported and validated insurance data revealed significant deficiencies for minority children in the relationship scale. The policy may affect this relationship by limiting the ability of families, and minorities in particular, to seek and develop relationships with culturally appropriate providers. This may be the case particularly if there are language difficulties, or if prior relationships are disrupted by forced entry into managed care, changing of health plans, or the frequent turnover of physicians in MCOs.
The requirements to select a network PCP and to seek referrals for specialty care were also associated with significant, but slightly smaller, deficits in longitudinality of care.
The policy requiring patients to select a PCP from a network was associated with reductions in the relationship domain for Asians and Hispanics using self-reported data and blacks using validated data. The perception of limited choice, in this case, is more negatively correlated with a sustained relationship than the actual presence of the policy. Requiring referral was associated with reductions in parent reporting of the interpersonal relationship between minority children and physicians, primarily when validated insurance data were used.
Perception of the referral policy, in this case, appears to be less detrimental than the actual policy and the corresponding physician behaviors it inspires with regard to the patient-provider relationship.
Policies undermining patient autonomy in seeking care or emphasizing the gatekeeping role of the PCP rather than the coordination of care role may lead to distrust of the physician and a poorer overall relationship. In one case, parents declined artificial feeding for an undernourished child with a disabilitypreferring the child remained lighter for lifting. And in other instances, parents of children with cancer have refused conventional treatment in favour of alternative therapies. Parental decision-making often involves weighing up the interests of different family members.
There are two main problems with this approach. There is no straightforward way of calculating well-being and comparing it across treatment options to identify which would be best. And there are often several possible courses of action, each of which would benefit the child in different ways.
Second, the best interests approach may not adequately acknowledge parental autonomy. Parental decision-making often involves weighing up the interests of different family members — which are ethically appropriate considerations — and parents may be choosing against the best possible outcome for a particular child in order to care for the family as a whole.
Striking the right balance Parents refusing treatment for their child is one type of situation for which doctors seek clinical ethics advice in paediatric hospitals. Although clinical ethics support services are widespread and longstanding in the United States and United Kingdom, in Australia they are in their infancy.
Over the past decade, clinical ethics support has been established at a few large metropolitan hospitals and more hospitals are working to make these services available to their staff. The outer boundary of the zone of parental discretion is harm to the child. Parents are not ethically entitled to choose options that may harm the child. For example, if parents are refusing physiotherapy necessary to ensure their child will walk again, this choice is outside the zone of parental discretion and should be overriden.
The zone of parental discretion is a conceptual tool, suited both to formal clinical ethics case discussions and more informal reflection and decision-making.