Objectives. We examined parental preferences for locus of service delivery for their teenager's congenital heart disease (CHD) and the influence of disease severity, sociodemographic factors, and insurance on these preferences.
Methods. A consecutive sample of parents of teenagers followed in a pediatric cardiology clinic completed a mailed questionnaire. Disease severity was classified as low (≤1 cardiovascular procedure), moderate (>1 cardiovascular procedure), and high (cyanosis or single ventricle physiology).
Results. Eighty-six of 148 parents responded (58%): 40, low severity; 36, moderate severity; and 10, high severity of illness. Parents preferred using primary care providers (PCPs) as a point of first contact for all 11 of 11 general health concerns and 5 of 7 potential cardiovascular-related concerns: chest pain (52%), syncope (73%), seeming seriously ill (79%), sports physical examination (79%), and endocarditis prophylactic antibiotics (94%). Increasing disease severity was significantly associated with preferring cardiologists for 6 of 7 cardiovascular-related concerns. Overall, 58% of parents viewed their care as a PCP–cardiologist comanagement model versus a cardiologist-dominated model. Lower family income (odds ratio [OR]: 1.5; confidence interval [CI]: 1.0–2.2) and severity of illness (OR: 2.1; CI: 1.0–4.4) were associated with a comanagement model of health care versus a cardiologist-dominated model.
Conclusions. This study suggests that the majority of parents of teenagers with CHD prefer to use their teenager's PCP for all routine health care needs and many cardiovascular health needs. Severity of illness and family income are positively associated with greater preference for cardiologist care.
- CHD =
- congenital heart disease •
- PCP =
- primary care provider •
- HMO =
- health maintenance organization •
- CI =
- confidence interval •
- OR =
- odds ratio
Attributable in part to the advances in medical and surgical care over the last 40 years, there are an increasing number of children with chronic medical conditions. Depending on the classification criteria used, it is estimated that between 5% and 30% of children in the United States have a chronic illness.1,2 Contained in these estimates is congenital heart disease (CHD), with an incidence rate of nearly 1 in every 100 newborns.3
Unfortunately, the organization of health care services for children with chronic illness is fraught with fragmental service delivery across settings and providers.4–6 For example, research on children with chronic arthropathies found that neither primary care providers (PCPs) nor specialists addressed issues such as parental advice for schooling and emotional issues, future family planning, genetic counseling, and coordination of care.7 It has also been reported that nearly 40% of children with chronic illness have health concerns, such as nervousness, headaches, and skin problems, that they had never discussed with any of their providers.8 Many factors likely play a role in this problem. For example, it has been reported that there is disagreement between PCPs and specialists on the degree to which specialists are asked to provide primary pediatric care, indicating a lack of clear role definitions.9 It has also been found that the availability of medical resources in the community, accessibility to tertiary care centers, insurance status, and gatekeeping arrangements all influence the interaction between PCPs and specialists.10–12 Specific to gatekeeping, it has been reported that physicians in gatekeeping plans were less likely to schedule an appointment or communicate with specialists for their referred patients.12
From the parental perspective, a recent study found that nearly one half of parents who self- referred their children to specialists reported that the child's regular physician was unconcerned about the medical problems.13 Because self-referral without knowledge of the child's regular physician can lead to redundant or unnecessary care and difficulty in arranging follow-up care, this parental perception and response additionally compound the fragmentation of care for chronic disease.13
With respect to CHD, research has suggested that there is little involvement of PCPs in providing health care for any aspect of children's heart disease. More specifically, only 30% of parents of children with CHD reported discussing aspects of the chronic illness, such as psychosocial, family, economic, or genetic issues, with their PCP.14 Additionally, the parents in this study expressed a low level of confidence in the ability of their PCP to meet many of their child's medical care needs.14
Increasingly, children with chronic conditions are enrolling in managed health care plans. Rules regarding the relationships between PCPs and specialists are highly variable across plans. The relevant issues include ease of ability to bypass gatekeepers, restrictions on access to specialists through utilization review mechanisms, and selective contracting with providers that may exclude some types of pediatric subspecialists. Unaccounted for in these rules are the parental preferences for chronic illness care distribution. Little is known concerning which providers parents of children with chronic illness view as accountable for delivery of general and disease-specific health care. Understanding the preferences of parents of children with chronic illness for locus of service delivery is essential to improve the responsiveness of the health care industry to patient/parent expectations.15–17
Examination of adult patient-related factors leading to differences in utilization of PCPs versus specialists found that gender, race, and insurance status result in differential utilization of PCPs and specialists.18 Not addressed in this study is the influence of disease severity on physician utilization. Furthermore, examination of care distribution preferences and of the influences of patient-related factors among childhood illnesses has not been performed.
We conducted a survey of parents of teenagers with CHD to examine their preferences for locus of service delivery. We had 3 hypotheses in conducting this research: 1) parental preferences for locus of service delivery regarding their teenager's potential cardiovascular-related concerns would favor use of a cardiologist unlike their preferences regarding general health concerns; 2) parental views about the current overall distribution of care responsibility between PCPs and specialists in the care of their teenagers' CHD would favor a cardiologist-dominated model; and 3) greater disease severity, patient sociodemographic factors, and insurance status would influence these parental preferences.
Case Selection and Survey Administration
A questionnaire was mailed to a consecutive sample of parents of teenagers 11 to 19 years old who were seen for follow-up evaluation in the Pediatric Cardiology Clinic at Johns Hopkins Hospital between July 1, 1996 and May 31, 1998. This clinic involved 7 attending cardiologists and did not use a routine policy governing counseling families with respect to type of provider to contact for specific concerns. Because of an additional component of the overall project requiring teenager responses, teenagers who were unable to read at an 11-year-old level were excluded because of their inability to independently complete a survey. Additionally, teenagers with other chronic disease were excluded to obtain a subject population with only CHD as a chronic health concern. Significant other chronic disease was defined as disease involving specialist management, in addition to the cardiologist, as determined from chart review.
The questionnaire had 18 items addressing service delivery preferences for both general health concerns and potential cardiovascular-related concerns. Attributable to an additional component of the overall project that required the teenagers to complete 87 other questions not reported here, the teenagers were not surveyed regarding their service delivery preferences because of concerns about respondent burden on the teenagers.19 Parental results for each of these 18 areas are reported separately because the interpretation of a summary score from the first report of this tool would be difficult.
The range of health-related concerns on the questionnaire encompassed preventive services, acute signs/symptoms, acute injuries, and psychosocial issues. Relevant items were selected based on review of the literature.6,14 The stem for each of these questions read, “Thinking about your teenager, who would you call or make a visit with first if … ,” with response options of either the cardiologist or pediatrician/family physician. The general health concerns that were addressed included signs/symptoms (cold symptoms and new skin rash), acute injuries (needing stitches, broken bone, and injured back), preventive health issues (needing immunizations and nutritional advice), and psychosocial issues (emotional/behavioral concerns, school performance concerns, sadness/depression concerns, and sexuality questions). The potential cardiovascular concerns addressed included sign/symptoms (chest pain, syncope, and seemed seriously ill) and preventive health issues (refill of heart medication, sports participation question, sports participation examination, and antibiotics for dental visit). In addition, 1 question was asked regarding current parental views of their teenager's health care. This read as follows: “Which of these doctors is mainly responsible for taking care of your teenager's heart condition?” The response options were the cardiologist, the pediatrician/family physician, or both the cardiologist and pediatrician/family physician. The survey also asked 13 questions about demographics and socioeconomic issues. Lastly, because care delivery preferences may be significantly altered by acute versus chronic disease status, we asked 6 questions regarding health care utilization related to CHD within the last year as a marker of disease acuity. Three weeks after the initial mailing, all participants who had not responded were contacted by telephone as a reminder. The estimated time for completion of this survey was 15 minutes.
Disease Severity Index
To obtain information about disease severity, a chart review was completed for all survey respondents. The severity index categorized the teenagers as low severity (≤1 cardiovascular surgery or catheterization procedure to date), moderate severity (>1 cardiovascular surgery or catheterization to date), and high severity (persistent cyanosis, <92% oxygen saturation at rest, or single ventricle physiology). Only diagnostic catheterizations performed in the newborn period were excluded in determining disease severity, because today newborn diagnosis of CHD is typically accomplished by echocardiography. This severity index has shown construct validity in terms of being able to distinguish subjects based on left ventricular function, use of cardiovascular medications, health-related quality of life, and complexity categorization of surgical intervention.19
The objectives of the data analyses were threefold. First, we sought to compare the proportion of parents preferring to see PCPs, compared with those preferring to see cardiologists for each specific general health concern and potential cardiovascular- related concern examined. Second, we examined parental views regarding their current overall health care model, choosing between a comanagement model with shared cardiovascular care between PCPs and cardiologists and a cardiologist-dominated model. Finally, we evaluated the impact of disease severity, patient sociodemographic factors, and insurance status on care distribution preferences.
One-sample tests of proportions were used to evaluate the care distribution preferences for the specific health concerns examined and for examination of the overall health care model. Univariate logistic regression analysis was performed to examine the linear impact of disease severity, family income, and insurance status on care distribution preferences. Multivariable logistic regression analysis, adjusting for disease severity, family income, and the age and gender of the teenager, was performed to evaluate associations with preferences for specific general health and potential cardiovascular-related issues and with views regarding the type of specialist involvement—specifically, a comanagement model of care versus a cardiologist-dominated model. Because of the small number of respondents with Medicaid insurance and the high collinearity between insurance status and income, insurance status was omitted from the regression analysis.
All care distribution preferences were binary, comparing PCPs with cardiologists and comparing a comanagement model with a cardiologist-dominated model. Those respondents who viewed the pediatrician alone as primarily responsible for their teenager's CHD were classified with the respondents who answered that both the cardiologist and PCP were responsible, because, by study inclusion criteria, all respondents had involvement of cardiologists. Insurance status was obtained from clinic administrative records and was classified into indemnity plans (Blue Cross Blue Shield and other commercial fee-for-service plans), health maintenance organizations (HMOs [including HMOs, point-of-service plans, and preferred provider plans]), and Medicaid (Medicaid recipients in managed care organizations). Our study population contained no families without insurance or with a self-pay status for other reasons. For multivariable regression analysis, 2 indicator variables were used for disease severity, while family income was dichotomized into <$60 000 per year versus ≥$60 000 per year. The age of the teenager was defined as the nearest year at the time of survey completion. All analyses were completed using STATA, Version 5.0 (STATA Corporation, College Station, TX).
A level of P ≤ .05 was used as the point of statistical significance and all confidence intervals (CIs) reported are at the 95% level. Regression analysis results are reported as odds ratios (ORs).
Survey Responses and Subject Characteristics
Of the 202 potentially eligible families, 148 were actually enrolled. The reasons for exclusion were: 3 teenagers with mental retardation not identified on initial chart review, 2 teenagers died since the last clinical examination, and 49 families were unreachable because of outdated address and telephone information. From the 148 families to whom surveys were mailed, responses were received from 86 families for a 58% response rate. There were no significant differences among the severity of illness categories for sociodemographic factors, type of insurance, and health care utilization within the last year as a marker of acute versus chronic CHD status (Table 1). Additionally, there was no significant difference in the distribution of disease severity among survey respondents and nonrespondents. The lesions represented in our study subjects encompassed 18 different forms of structural heart disease3 (Table 2).
Health Care Preferences and Views
Parents of teenagers with CHD nearly uniformly preferred using the PCP as a point of first contact for all 11 of 11 specific issues regarding general health concerns, such as acute signs/symptoms, acute injuries, preventive health issues, and psychosocial issues. With respect to 4 of 7 specific potential cardiovascular-related concerns, significantly more parents preferred to see the PCP as a point of first contact as opposed to the cardiologist (Table 3). Specifically, a significantly greater percentage of parents preferred to see PCPs when asking about the potential cardiovascular-related signs of syncope (P < .001) and seeming seriously ill (P < .001), as well as the cardiovascular preventive health concerns of a sports participation examination (P < .001) and obtaining antibiotics for endocarditis prophylaxis (P < .001). Conversely, for the specific issues of heart medication prescription refills (P < .001) and for a specific sports participation question (P = .001), significantly more parents preferred to see a cardiologist first.
When asked about their view regarding an overall health care model, 58% of parents considered their teenager's CHD care responsibility to occur as a comanagement model, shared between a PCP and cardiologist, as opposed to a cardiologist-dominated model (P = .003). Additionally, parents of teenagers in comanagement models reported a consistently greater preference for seeing a PCP for specific potential cardiovascular-related concerns.
Health Care Preferences and Views Stratified by Disease Severity
Increasing disease severity was associated with a consistent trend of decreased parental preference for PCP care delivery for 6 of 7 specific potential cardiovascular-related health concerns (Table 4). With respect to acute signs and symptoms, this trend was significant for chest pain (P= .02), syncope (P = .003), and seeming seriously ill (P = .04). Comparably, this trend was significant for the preventive health issues of specific sports participation questions (P = .04), sports participation examinations (P = .02), and obtaining antibiotics for endocarditis prophylaxis (P = .02). Conversely, disease severity was not associated with current views on the type of specialist involvement, namely comanagement with a PCP versus specialist-dominated care.
Health Care Preferences and Views Stratified by Family Income and Insurance Status
Family income and insurance status were not associated in univariate analysis with differences in PCP versus cardiologist preferences for all specific potential cardiovascular-related issues examined. Furthermore, family income and insurance status were not associated with parental views about the current type of specialist involvement model for their teenager's CHD care. All 7 respondents with Medicaid insurance reported comanagement model views, although this may primarily reflect chance given the small sample size.
Multivariable Regression Analysis
Adjusting for family income, the age and gender of the teenager, and disease severity, there was a consistent finding of decreased preference for use of a PCP with greater disease severity for all 7 specific cardiovascular concerns. Moreover, despite the relatively small sample size, this finding was statistically significant for 5 of the concerns: chest pain (P = .02), syncope (P = .003), sports participation questions (P = .04), sports participation examinations (P = .03), and obtaining prophylactic antibiotics (P = .04). The covariates of income, age, and gender had varied impacts on preferences for using a PCP compared with using a cardiologist, never reaching significance for any of the 7 specific cardiovascular concerns.
In terms of current views of overall health care models, covariates significantly associated with a comanagement model as opposed to a cardiologist-dominated model were lower disease severity (OR: 2.1; CI: 1.0–4.5) and lower family income (OR: 2.8; CI: 1.0–7.7). Low disease severity compared with high disease severity was more strongly associated with a comanagement model view (OR: 5.9; CI: 1.1–32.4). There was not a statistically significant association for current comanagement model views when comparing moderate severity disease with high severity disease (OR: 3.4; CI: .6–18.4).
The majority of parents of teenagers with CHD preferred using their PCP rather than a cardiologist as a point of first contact for all general health concerns and many specific potential cardiovascular-related concerns. Furthermore, we consistently found a decrease in preference for PCP use with increasing disease severity. Overall, the majority of parents viewed their teenager's CHD care to involve a comanagement health care model, defined as a sharing of responsibility for cardiovascular care between the PCP and cardiologist, compared with a cardiologist-dominated model of care. This study also found that teenagers with lower disease severity and low family income were more likely to consider themselves to be in a comanagement model of care than were their counterparts.
Several limitations of this study should be noted. First, the stated preferences of the parents concerning locus of service delivery may differ from actual use. Comparably, although it may seem from our survey results that the parents have a great deal of confidence in their teenager's PCP, unlike some previous reports, such an extrapolation of our data needs to be done with caution. Even so, our results suggest that parental satisfaction with the health care delivered to their teenagers with chronic illness will be enhanced when their experiences match their expectations for a shared PCP–cardiologist model of patient management. Second, response bias is a threat to the study's internal validity. We achieved a 58% response rate and importantly found no significant difference in disease severity distribution among respondents and nonrespondents. Third, our categorization of the financing and organizational features of the insurance plan did not directly account for restrictions on access to specialists. It is unlikely that these issues had any differential effects in our study sample because the distribution of insurance types across the disease severity categories was comparable. Fourth, our study had a relatively small sample size, especially in the high severity of illness group. Although this fact may have obscured some findings, multiple unique and significant findings were still evident. Finally, the responses of parents in this study may differ from the responses of parents with teenagers with other chronic illnesses. Likewise, responses of parents of children with multiple medical problems may also differ from this sample of children with only CHD. It is likely that having multiple medical problems would be associated with greater preference for specialist care.
Not answered by this study is the important question of why severity of illness and family income are associated with decreased preference for PCP use. Perhaps it is appropriate that parents of teenagers with greater disease severity prefer specialist involvement because their teenagers need more intensive services. Previous research specific to CHD has reported that parent's of children with CHD have a low level of confidence in the ability of the PCP to meet the medical needs of their children.14 This lack of confidence may be exacerbated by greater disease severity. Among adult patients, affluence has been shown to be associated with greater self-referral to specialists.18 Our family income findings may represent the pediatric extension of this same association. The reasons behind this association are unclear and likely multifactorial.
A recently published study examining the value that adult patients place on their PCP as first-contact care found that most patients preferred PCP involvement in decisions about obtaining care from specialists and that most patients preferred to initiate care for new medical problems with their PCP.20 Additionally, this study found that patients who had difficulty in obtaining specialty referrals were more likely to report low trust, confidence, and satisfaction with their PCP.20 Therefore, the authors concluded that managed care policies that emphasize PCPs as gatekeepers to impede access to specialists actually undermine patients' trust and confidence in PCPs. Our finding that parents of teenagers also place value on PCPs as first-contact care mirrors the data on adults. We would speculate, therefore, that managed care policies to impede specialist access would have the comparable effect of diminishing trust and confidence in pediatric PCPs. This may have a differential effect on the health status of children with high medical needs because of their greater reliance on specialists.
Overall, our study reaffirms the need for a functional sharing of health care responsibilities not only as the model of choice by managed care organizations, but also as the model of choice by a majority of parents of teenagers with chronic illness. Furthermore, our study identified 2 patient-related factors, disease severity and family income, which may independently influence a functional collaboration between PCPs and specialists.
Dr Miller was supported by the Hayden Whitney Smith Research Scholarship.
- Received August 30, 1999.
- Accepted December 10, 1999.
Reprint requests to (M.R.M.) Agency for Health Care Policy and Research, Center for Quality Measurement and Improvement, 2101 E Jefferson St, Suite 502, Rockville, MD 20852. E-mail:
The opinions and assertions contained herein are the private views of the authors. No statement in this article should be construed as an official position of the Agency for Health Care Policy and Research or the US Department of Health and Human Services.
- Stein REK
- Perrin EC,
- Newacheck P,
- Pless IB,
- et al.
- Ferencz C, Rubin JD, Loffredo CA, Magee CA, eds. Epidemiology of Congenital Heart Disease: The Baltimore-Washington Infant Study, 1981–1989, Perspectives in Pediatric Cardiology, IV. New York, NY: Futura Publishing Company; 1993
- Pless IB,
- Satterwhite B,
- Van Vechten D
- Kanthor H,
- Pless B,
- Satterwhite B,
- Myers G
- Palfrey JS,
- Levy JC,
- Gilbert KL
- Blancquaert IR,
- Zvagulis I,
- Gray-Donald K,
- Pless IB
- Forrest CB,
- Glade GB,
- Starfield B,
- Baker AE,
- Kang M,
- Reid RJ
- Young PC,
- Shyr Y,
- Schork MA
- Miller MR, Forrest CB, Kan JS. A disease severity index and health-related quality of life for congenital heart disease. Manuscript in press
- Copyright © 2000 American Academy of Pediatrics