OBJECTIVE. The goal was to examine pediatricians' views about whether and how well-child care for children 0 to 5 years of age should be changed.
METHODS. A mail survey of a national random sample of 1000 general pediatricians was performed with a survey instrument that examined pediatricians' attitudes and behaviors toward our current way and an ideal way of providing well-child care. Results were analyzed for the following 3 major domains of change in well-child care: provider type, visit format, and visit location.
RESULTS. Sixty percent (n = 502) of eligible subjects responded to the survey. Nearly all respondents (97%) rated the current US system as excellent or good in providing well-child care. Most pediatricians (85%–91%) reported that they are currently the main providers of anticipatory guidance, developmental screening, and psychosocial screening. However, a majority (54%–60%) reported that, in an ideal system that maximized the effectiveness and efficiency of care, nonphysicians would provide these services. Fewer pediatricians (24%) reported that ideally nonphysicians should provide the physical examination. The majority of respondents (79%–93%) reported that at least some anticipatory guidance, minor acute care, and chronic care services could be conducted through telephone or e-mail communication, and 55% stated that at least some well-child care services should be provided in alternative locations, such as day care centers. In multivariate analysis, support for these changes was distributed widely across pediatricians with varying personal and practice characteristics.
CONCLUSIONS. Although most pediatricians are generally satisfied with our current way of providing well-child care, a majority think that a system that is less reliant on physicians and face-to-face office visits would be a more effective and efficient way to provide care.
The American Academy of Pediatrics recommends that all children receive comprehensive preventive care that includes screening, counseling, and guidance directed or provided by a physician.1 Over 25% of all physician visits by children are for these well-child care visits, for which pediatricians serve as the main providers.2,3
There are systematic problems with well-child care, ranging from disparities in access to care to wide variations in the quality of care.4–13 There have also been proposals for redesigning care and improving the quality of care,14–27 including improving the delivery of developmental services, increasing the efficiency of office visits, and using group formats for well-child care.15,20,24,28
The views of pediatricians are important in considering redesign, because of the intimate familiarity of pediatricians with how well-child care is being delivered currently and the importance of their involvement in any efforts at redesign. However, we know very little about the views of pediatricians regarding how our system of care could be designed for the optimal delivery of preventive services. Therefore, we conducted a national survey of primary care pediatricians, to discover whether and how they think well-child care should be restructured.
Subjects and Design
We sampled randomly 1000 pediatricians from the American Medical Association Masterfile of office- and hospital-based pediatricians who identified pediatrics as their primary specialty, had completed training, and were <70 years of age (N = 42593). This sample size was chosen to provide national estimates of pediatricians' beliefs about changes to well-child care with a precision of ±5%, assuming a 10% rate of undeliverable addresses, a 10% rate of ineligibility, and a 50% response rate. During August 2005, a survey was mailed to each subject, with a cover letter describing the purpose of the survey. Pediatricians who did not respond initially were sent follow-up letters and replacement surveys in 2 mailings at 6- to 8-week intervals. A $5 cash incentive was included in the third mailing, and respondents in all rounds were included in a raffle for a $1000 prize. Subjects were considered ineligible if their mailing address was undeliverable or if they reported that they did not practice general outpatient pediatrics. The study was approved by the University of Chicago institutional review board.
We developed a survey instrument with the help of an advisory board of community and academic pediatricians. This advisory board was selected to represent a diverse group of perspectives and focused on considering advantages and disadvantages of our current system of care, specific areas in need of change, and alternatives to our current way of providing care. To maximize the psychometric rigor of the instrument, the survey was reviewed with experts in survey design, revised, pretested, and revised again.
The survey included closed-ended items to assess pediatricians' attitudes about their current way of providing well-child care and about what they believed, from their own perspective, would be an ideal way of providing care. The survey defined ideal as maximizing both the effectiveness and efficiency of care. Because we wanted respondents to focus on quality of care, rather than the effect of changes on the finances of their practice, they were asked to assume that their income would not be affected by the proposed changes. Although the survey focused on well-child care services, we included other aspects of primary care (eg, minor acute care and chronic care) that often are addressed during well-child care visits.
The survey asked respondents to rate, on a 4-point scale, how well our current system (defined as a system “where the physician is the primary provider of care”) works for providing various services within primary care. Respondents rated well-child care visits as a whole, as well as individual services within well-child care. We then asked respondents about 3 areas of possible change in well-child care. First, we asked respondents which type of providers currently provided each of 4 well-child care services (anticipatory guidance, developmental screening, psychosocial screening, and physical examination) for healthy children 0 to 5 years of age in their practices and which provider type (pediatricians, nurse practitioners/physician assistants, or registered nurses/medical assistants or other staff members) they thought ideally should provide each service.
Second, we asked whether various services (anticipatory guidance, acute care, and chronic care management) could be provided by using alternative formats such as e-mail and telephone. Third, we asked whether respondents thought that some well-child care services should be provided in alternative locations. We included schools, day care centers, and grocery stores together as examples of alternative locations, to illustrate a wide range of options for respondents, some used currently for well-child care visits and some used rarely. For alternative visit formats and visit locations, pediatricians were asked to respond with the number of visits that could be handled in these alternative ways (many, some, few, or none).
We also collected information on the personal and practice characteristics of respondents. The survey focused on care for children 0 to 5 years of age.
We used simple descriptive statistics to examine pediatricians' attitudes toward redesign, focusing on (1) how they rated the current system of care, (2) who currently provided well-child care services in their own practices and who they thought ideally should do so, and (3) whether some visits for primary care services should be provided through alternative formats and in alternative settings. Then, χ2 and t tests were used to examine bivariate associations of personal and practice demographic features with survey responses and main outcome variables.
We performed regression analyses to discover whether certain respondents were more likely to favor changes. We examined the multivariate associations between 8 outcome variables measuring pediatrician support for changes and predictor variables that we hypothesized would be associated with pediatricians' attitudes toward changes. Four outcome variables measured support for using nonphysicians as primary providers for (1) anticipatory guidance, (2) developmental screening, (3) psychosocial screening, and (4) physical examination. Three variables measured support for telephone and e-mail communication for (1) anticipatory guidance, (2) acute care, and (3) chronic care management. The last outcome variable measured support for the use of alternative locations for well-child care visits. Predictor variables were respondent's age, gender, practice setting, board certification status, amount of personal income from salary, reported proportion of patients with Medicaid insurance, and number of nurse practitioners and physician assistants at the respondent's main practice site.
We performed the regression analyses in 2 different ways. First we used maximum likelihood ordered logistic regression, which allowed us to preserve the ordinal nature of the outcome variables, for example, the ideal provider type as (1) physician, (2) nurse practitioner/physician assistant, or (3) registered nurse/medical assistant. Then we repeated the analysis by using logistic regression, which required that we dichotomize the outcome variables, for example, provider type as physician or nonphysician. We used a theoretically derived model that included all predictor variables hypothesized to be associated with our outcomes. Because the logistic and ordered logistic regression results were very similar, we present the results from the logistic regression analysis for ease of interpretation. All analyses were conducted with Stata SE 9 software (Stata Corp, College Station, TX).
Of the 1000 pediatricians to whom surveys were sent, 84 (8%) had an undeliverable address and 83 (8%) did not provide outpatient primary care. Of the remaining 833 eligible pediatricians, 502 returned a survey (response rate: 60%). Table 1 presents respondent and nonrespondent characteristics. Compared with nonrespondents, respondents were slightly more likely to have board certification.
Attitudes Toward the Current System
The vast majority of respondents rated the current system of care as excellent or good for providing well-child care for healthy children (97%), chronic illness management (93%), and anticipatory guidance (88%) (Table 2). In contrast, only 55% rated the system as excellent or good for providing psychosocial screening.
Well-Child Care in an Ideal System
Who Should Provide Care?
Despite their generally positive views of our current system, many respondents favored changes when asked how care ideally should be provided (Table 3). For example, although 9 of 10 respondents stated that currently they were the main providers of most well-child care services, 55% to 60% thought that ideally anticipatory guidance, developmental screening, and psychosocial screening should be provided by nonphysician providers, such as nurse practitioners, physician assistants, nurses, and medical assistants.
Where and How Should Care Be Provided?
One fourth of pediatricians reported that many or some services should be provided at locations other than the doctor's office, such as schools, day care centers, or grocery stores (data not shown). An additional 30% stated that a few services should be provided there. Sixty-four percent of pediatricians stated that much or some anticipatory guidance could be provided by telephone or e-mail; an additional 29% stated that a few such services could be provided that way. Similar proportions of respondents reported that visits for minor acute care could be replaced by telephone or e-mail contacts (many or some visits: 57%; a few visits: 33%), and fewer, but still a majority, thought so for chronic care management (many or some visits: 38%; a few visits: 41%).
Which Pediatricians Favor Change?
In multivariate analysis, respondent age was the only respondent characteristic that demonstrated a consistent association with any of the 8 outcome variables. Compared with respondents <40 years of age, respondents >60 years of age were more likely to favor the use of nonphysicians to perform routine physical examinations (odds ratio [OR]: 4.5; 95% confidence interval [CI]: 1.9–10.4; P = .001) but less likely to state that e-mail and telephone communication could replace some visits for anticipatory guidance (OR: 0.3; 95% CI: 0.1–0.5; P < .000), chronic care (OR: 0.3; 95% CI: 0.2–0.7; P = .003), and acute care (OR: 0.4; 95% CI: 0.2–0.9; P = .031) or that care should be provided in alternative locations (OR: 0.3; 95% CI: 0.1–0.6; P = .002) (data not shown).
Despite their satisfaction with the current system of well-child care, many pediatricians reported that it should be structured quite differently. Respondents were interested in using nonphysician staff members to provide many well-child care services, using telephone and e-mail communication with parents to replace in-office visits, and allowing at least some visits to be provided outside the physician's office. There were few physician or practice characteristics associated with beliefs about these changes to care, which suggests that our findings are shared by a diverse group of primary care pediatricians.
Although pediatricians gave high ratings to our current system for providing well-child care, many reported that an ideal system would be structured quite differently. There are a few possible explanations for this apparent contradiction. First, respondents' lower ratings of certain services, such as psychosocial screening, may contribute to their attitudes toward change. Next, it is possible that many respondents simply believe that, although the current system is good, it could be better. Lastly, it may be that respondents report satisfaction initially but with more probing questions they can identify areas for change.
In other national surveys, many pediatricians reported that they do not have adequate time or training or receive adequate reimbursement to provide comprehensive well-child care services, including anticipatory guidance, developmental screening, and psychosocial screening.29–31 This supports our finding that a majority of respondents thought that these well-child care services should be provided by nonphysician professionals, with the notable exception of the physical examination. The possibility of using nonphysician professionals to provide these services was described in the American Academy of Pediatrics guidelines for well-child care published in 1967.32 Since then, innovative programs to improve care have used such resources. For example, the Healthy Steps for Young Children Program uses developmental specialists to enhance developmental and behavioral services through in-office well-child care visits, home visits, parent groups, and telephone advice lines.24
Pediatricians >60 years of age were unique in that they were >4 times more likely to think that nonphysicians should be the main providers of the physical examination. It seems that physicians with more years of experience were more comfortable delegating this aspect of care, which younger physicians were reluctant to relinquish. In contrast, their years of experience seemed to have made them less likely to relinquish more of the traditional in-person office visits to telephone or e-mail communication or visits in other settings.
This study has several limitations. First, because there were no previous published or available survey instruments that addressed our research questions, we developed a novel survey instrument that had no measure of reliability. Next, we used a small cash incentive ($5) that, because of financial constraints, was offered only in the third mailing. This use of differential incentives might be seen as inequitable treatment of early responders.33
Although respondents and nonrespondents were similar in most ways, our respondents were more likely to be board certified, which might limit the generalizability of our results. When we asked how care should be provided ideally, we defined ideal as being the most effective and efficient way to provide care. However, some respondents might have interpreted “ideal” processes of care in a different manner (eg, what they most enjoy doing). It is also possible that the respondents answered questions on the basis of concerns about preservation of their income and current way of practicing. However, such a tendency would understate our findings, and we did find that a majority of pediatricians reported an ideal system of primary care that was quite different from their current practice. Finally, our findings were based on a closed-ended question format, which limited discovery of pediatricians' views to only the changes we addressed in the survey.
Our findings have policy, research, and practice implications for well-child care and primary care in general. First, payors for care might consider creating reimbursement systems that give practices the flexibility to be innovative in the way they deliver care. In doing so, payors might also be supporting possible changes that could optimize the efficiency of care. Second, researchers could evaluate different models of well-child care by measuring changes in patient outcomes and receipt of services, overall costs, and parent and provider satisfaction. Lastly, the use of alternative providers and locations for care would require more extensive coordination of care between various providers and locations and more extensive efforts to maintain the doctor-parent relationship. Both of these could be achieved through a medical home, in which children receive comprehensive care that may be provided in a number of community settings by various providers, managed by a well-trained physician.1
To our knowledge, this is the first study to report pediatricians' attitudes toward redesigning the system of well-child care. The findings from this national survey of pediatricians suggest that a majority of primary care pediatricians working in diverse settings think that a system of care that is less dependent on physicians and face-to-face office visits may enhance the effectiveness and efficiency of care.
This study was supported by the Robert Wood Johnson Clinical Scholars Program.
We gratefully acknowledge the project advisory board (Herbert Abelson, Donald Brown, Karen Goldstein, Arthur Kohrman, James Mitchell, and Beth Volin). We also acknowledge the generous advice given by Ted Karrison, Diane Lauderdale, Colm O’Muircheartaigh, Heena Santry, Anirban Basu, and Ronald Thisted in statistical methods and survey planning. Lastly, we thank Sumerah Backsh, Holly Stotler, and Terri Rossi for survey coordination and all of the physicians who participated in the survey.
- Accepted July 17, 2006.
- Address correspondence to Tumaini Coker, MD, MBA, University of Chicago, 5841 S Maryland, MC 2007, Chicago, IL 60637. E-mail:
The funding source had no role in the collection, analysis, interpretation, or reporting of the data or in the decision to submit the manuscript for publication.
The authors have indicated they have no financial relationships relevant to this article to disclose.
- ↵American Academy of Pediatrics, Medical Home Initiatives for Children With Special Health Care Needs. Policy statement: organizational principles to guide and define the child health care system and/or improve the health of all children. Pediatrics.2004;113 :1545– 1547
- ↵Hing E, Cherry DK, Woodwell DA. National Ambulatory Medical Care Survey: 2004 summary. Adv Data.2006;(374):1–36
- Triggs EG, Perrin EC. Listening carefully: improving communication about behavior and development: recognizing parental concerns. Clin Pediatr (Phila).1989;28 :185– 192
- Reisinger K, Bires JA. Anticipatory guidance in pediatric practice. Pediatrics.1980;66 :889– 892
- Leatherman S, McCarthy D. Quality of Health Care for Children and Adolescents: A Chartbook. New York, NY: The Commonwealth Fund; 2004
- ↵Schor EL. Rethinking well-child care. Pediatrics.2004;114 :210– 216
- Dodds M, Nicholson L, Muse B, Osborn LM. Group health supervision visits more effective than individual visits in delivering health care information. Pediatrics.1993;91 :668– 670
- Rice RL, Slater CJ. An analysis of group versus individual child health supervision. Clin Pediatr (Phila).1997;36 :685– 689
- Taylor JA, Davis RL, Kemper KJ. A randomized controlled trial of group versus individual well child care for high-risk children: maternal-child interaction and developmental outcomes. Pediatrics.1997;99 (6). Available at: www.pediatrics.org/cgi/content/full/99/6/e9
- Zuckerman B, Parker S. Preventive pediatrics: new models of providing needed health services. Pediatrics.1995;95 :758– 762
- Margolis PA, Stevens R, Bordley WC, et al. From concept to application: the impact of a community-wide intervention to improve the delivery of preventive services to children. Pediatrics.2001;108 (3). Available at: www.pediatrics.org/cgi/content/full/108/3/e42
- Bundy DG, Randolph GD, Murray M, Anderson J, Margolis PA. Open access in primary care: results of a North Carolina pilot project. Pediatrics.2005;116 :82– 87
- ↵Zuckerman B, Parker S, Kaplan-Sanoff M, Augustyn M, Barth MC. Healthy Steps: a case study of innovation in pediatric practice. Pediatrics.2004;114 :820– 826
- ↵Hochstein M, Sareen H, Olson LM, O'Connor KG, Inkelas M, Halfon N. A comparison of barriers to the provision of developmental assessments and psychosocial screenings during pediatric health supervision. Presented at: annual meeting of the Pediatric Academic Societies; April 28–May 1, 2001; Baltimore, MD
- ↵O'Connor K, Duncan P, Hagan J. What to say and when? Prioritizing and prompting preventive care services. Presented at: annual meeting of the Pediatric Academic Societies; May 14–17, 2005; Washington, DC
- ↵Council on Pediatric Practice. Standards of Child Health Care. Evanston, IL: American Academy of Pediatrics; 1967
- Copyright © 2006 by the American Academy of Pediatrics