BACKGROUND. Although primary care physicians are responsible for providing follow-up care after a positive newborn screen, little is known about their willingness or ability to do so.
METHODS. A national mail survey of a random sample of 350 general pediatricians and 350 family physicians was conducted from April to June 2006.
RESULTS. The response rate was 63% among pediatricians and 50% among family physicians. Most pediatricians (89.7%) and nearly one half of family physicians (44.1%) had had a patient with a positive newborn screen within the past 5 years. Most respondents thought that primary care physicians should be responsible for informing families about a positive newborn screen (73.2%), arranging confirmatory testing (66.0%), and coordinating subspecialty referral (85.3%). However, more than one half (56.2%) would prefer newborn screening programs to provide the initial evaluation of positive newborn screening results. Some respondents (but fewer pediatricians than family physicians) reported that they were not competent to discuss conditions included in newborn screening panels (eg, 22.6% of pediatricians and 53.2% of family physicians for phenylketonuria and 8.8% of pediatricians and 40.4% of family physicians for congenital hypothyroidism). More than one half (58.3%) thought that families with a child diagnosed as having congenital hypothyroidism should receive formal genetic counseling. Respondents were less likely to think that families with a child with sickle cell trait, compared with families with a child who is a cystic fibrosis carrier, should receive formal genetic counseling (69.3% vs 84.1%).
CONCLUSIONS. Many primary care physicians are not prepared to manage the follow-up care of children with a positive newborn screen, including initial counseling, diagnosis, and subspecialty referral. New strategies are needed to ensure appropriate and equitable health care delivery.
The benefits of newborn screening depend on timely follow-up care of children who test positive. Although state newborn screening programs play an important role in coordinating follow-up care, primary care physicians are expected to communicate the results of testing to families, to provide counseling about the meaning of a positive result and the significance of the suspected condition, to ensure that confirmatory testing is performed, to share the results of confirmatory testing with families, to help families navigate the health care system through appropriate referrals, and to monitor the health outcomes associated with the condition throughout childhood.1
These are challenging tasks because the conditions are uncommon and it is difficult to prepare a large workforce to be efficient and effective for rare events.2 Training primary care physicians is difficult because of the state-by-state variability in newborn screening panels3 and because resources, such as available subspecialists4 and genetic counselors, vary significantly among communities. Complicating these challenges even more, the number of conditions in newborn screening panels is increasing. The American College of Medical Genetics (ACMG), in a report commissioned by the federal government, now recommends a panel of 29 conditions.5
To assist primary care physicians, the ACMG developed management guidelines, known as ACTion (ACT) sheets, for each condition.6 These ACT sheets provide a short description of the condition, the differential diagnosis, a list of actions to be taken, methods of confirmatory testing, clinical expectations, and a list of resources. State newborn screening programs can modify these sheets to include a list of local resources. No data are available yet regarding how newborn screening programs plan to use the ACT sheets or how health care providers will receive or use these guidelines.
Although newborn screening has been part of routine care for decades, surprisingly few data are available about the role played by primary care physicians. Previous studies focused on the attitudes of pediatricians toward newborn screening7,8 and communication barriers.9 We are not aware of any studies of the care provided by pediatricians to families after a positive newborn screen. Similarly, we are not aware of any study that has evaluated the practice or attitudes of family physicians. Family physicians provide care for ∼20% of children overall10 and provide an even greater proportion of care in rural areas.11 To begin to understand the potential impact of the new ACMG guidelines, we surveyed primary care physicians about their practice, experience, and attitudes regarding the follow-up care of children with positive newborn screening results.
A national random sample of 350 pediatricians and 350 family physicians was drawn from the American Medical Association Masterfile, a database of all licensed physicians in the United States. The sampling frame included all allopathic and osteopathic physicians in office-based, direct patient care whose board certification and self-described primary specialty was either pediatrics or family medicine. We excluded physicians with a sub-board or secondary specialty, physicians who were ≥70 years of age, resident physicians, and physicians practicing at military or federal facilities. On the basis of survey responses, we also excluded physicians who reported that they did not provide well-child care for children <1 year of age.
Survey Instrument Development
We developed an instrument to assess practice, experience, and attitudes regarding the follow-up care of children with positive newborn screening results. Because of the complexity and heterogeneity of the conditions included in newborn screening panels, we focused on 5 conditions, including 3 conditions that are included in all states' newborn screening panels (congenital hypothyroidism, phenylketonuria [PKU], and sickle cell disease) and 2 conditions for which screening has been recommended but not adopted by all states (cystic fibrosis and medium-chain acyl-coenzyme A dehydrogenase deficiency [MCAD]).3 We attempted to represent the complexity of panels by including metabolic (ie, MCAD and PKU), endocrine (ie, congenital hypothyroidism), hematologic (ie, sickle cell disease), and complex single-gene (ie, cystic fibrosis) conditions, conditions for which the carrier status would be detected unavoidably through screening (ie, cystic fibrosis and sickle cell disease), a condition that is only rarely based on genetic inheritance (ie, congenital hypothyroidism),12 and a condition for which screening has been made possible by tandem mass spectrometry (ie, MCAD).
The survey instrument consisted of 19 questions that explored beliefs about who should be responsible for providing follow-up care after a child has a positive newborn screen, barriers to providing such care, how new guidelines for management should be provided, knowledge regarding the availability of genetic counselors and beliefs about which families should receive counseling, and practice characteristics (eg, including practice ownership or affiliation and rural/suburban/urban location of the respondent's main practice site). To gain additional insight into primary care physicians' knowledge of local resources and knowledge and beliefs about the conditions included in newborn screening panels, we also asked about knowledge regarding the availability of a genetic counselor to whom they could refer families and the indications for such referral. We did not ask questions about other subspecialty referral practices, to minimize the amount of time needed to complete the survey and to ensure adequate response rates.
The instrument was pilot tested by a convenience sample of primary care physicians and was revised to ensure clarity. The instrument consisted primarily of multiple-choice questions and Likert scales of agreement and took <5 minutes to complete.
The first survey mailing, accompanied by a cover letter, a small cash incentive, and a business reply envelope, was sent during April 2006. Two subsequent mailings to nonrespondents were sent at 3-week intervals.
Initially, general frequency responses to all survey items were determined. After this, Pearson χ2 tests of independence were used to test for associations among the categorical variables. We identified whether respondents practiced in states that screen for cystic fibrosis or MCAD on the basis of their mailing addresses. Logistic regression analysis was used to adjust for rural versus suburban or urban differences in selected comparisons of the differences between pediatricians and family physicians. All analyses were performed with Stata 8.2 software (Stata, College Station, TX). We considered P < .05 to be statistically significant. The University of Michigan Medical School institutional review board approved this project.
Response Rates and Practice Characteristics
Of the 700 physicians in this sample, 77 were ineligible (76 did not provide well-child care for infants and 1 was retired) and 9 had undeliverable addresses. The response rate among pediatricians was 63% (207 eligible surveys), and that among family physicians was 50% (143 eligible surveys).13
Respondent characteristics are listed in Table 1. Although practice ownership types were similar between the pediatricians and family physicians, the pediatricians were less likely to practice in rural areas. Nearly all pediatricians but less than one half of the family physicians had had a positive newborn screen in their practice within the past 5 years. Few of the respondents practiced in states that screen for cystic fibrosis. However, most respondents (relatively more pediatricians than family physicians) practiced in states that screen for MCAD.
Roles and Responsibilities
Most pediatricians (75.8%) and family physicians (69.6%) thought that primary care physicians, instead of the newborn screening program, should be principally responsible for informing families about a positive newborn screen (P = .21). More pediatricians than family physicians thought that primary care physicians should be principally responsible for arranging confirmatory testing (75.5% vs 54.0%; P < .001) and for arranging subspecialty evaluation if a condition is confirmed (90.7% vs 78.4%; P = .001).
Barriers to Follow-up Care After Newborn Screening
Table 2 lists the proportions of respondents who identified the selected barriers to follow-up care for each of the conditions included in the survey. For the most part, pediatricians were less likely to report barriers than were family physicians. Among the conditions, MCAD had the most reported barriers. Approximately 20% of pediatricians and one half of family physicians reported that they would not feel competent discussing PKU with families after a positive newborn screen. Approximately 10% of pediatricians and one third of family physicians reported that they would not feel competent discussing congenital hypothyroidism, sickle cell disease, or cystic fibrosis. For each of the conditions in the survey, pediatricians were less uncertain than were family physicians regarding who to refer patients to if a condition was confirmed.
Pediatricians in states with MCAD screening were more likely than pediatricians in states without MCAD screening to be uncertain what confirmatory test to order (71.0% vs 48.3%; P = .02), to be uncertain how to interpret confirmatory tests (52.7% vs 31.0%; P = .03), and to be uncertain who to refer patients to if the condition was confirmed (19.5% vs 3.5%; P = .03). However, there was no statistical difference in the barriers related to competence in discussing MCAD (P = .82) or the availability of a laboratory for confirmatory testing (P = .11). Among family physicians, there were no statistically significant differences in the proportions who reported any of the barriers between those in states with and without MCAD screening (P > .24). We had an insufficient sample size to compare those in states with and without cystic fibrosis screening.
Attitudes Toward New Written Guidelines
Most respondents, regardless of physician type, would prefer to receive the new guidelines by mail (47.3%) or by fax (31.5%). Fewer reported that they would prefer the guidelines to be disseminated electronically (eg, e-mail: 12.0%; Internet: 8.0%). More than one half of the pediatricians (60.0%) and family physicians (53.2%; P = .20) agreed or somewhat agreed that they would prefer a telephone call explaining what to do in the case of a positive newborn screen, instead of receiving a written guideline. Despite the availability of new specific guidelines, approximately one half of the pediatricians (50.7%) and more than one half of the family physicians (64.1%) agreed or somewhat agreed that they would prefer the newborn screening program to provide the initial evaluation (P = .01).
Availability of and Recommendations for Genetic Counseling
Overall, pediatricians were more likely than family physicians to report knowing of a genetic counselor in their practice area (83.9% vs 65.7%; P < .001) and were less likely to be unsure whether there was one available (7.8% vs 18.9%; P < .01). After adjustment for rural versus urban or suburban practice location, pediatricians, compared with family physicians, still had greater odds (2.47 vs 1; P = .001) of knowing and lower odds of being unsure (0.30 vs 1; P = .001) whether there was a genetic counselor in their area.
Table 3 lists the conditions for which the respondents thought that families with affected children should receive formal genetic counseling. The least commonly recommended condition for which families should receive formal genetic counseling was congenital hypothyroidism; however, more than one half recommended such counseling. Among pediatricians, the likelihood of recommending referral for congenital hypothyroidism did not vary according to whether they reported feeling competent discussing the condition with the parents of a child with a positive newborn screen (P = .82). However, family physicians who reported not feeling competent were more likely to refer families (75.0% vs 51.2%; P < .01).
The only significant difference between pediatricians and family physicians was for sickle cell trait. After adjustment for rural versus urban or suburban practice differences and whether a genetic counselor was known to be available, pediatricians had lower odds than family physicians of recommending genetic counseling for families with a child with sickle cell trait (0.58 vs 1; P = .03). Overall, 19.4% of pediatricians and 12.6% of family physicians responded that they would refer a family with a child who is a carrier for cystic fibrosis for formal genetic counseling but would not refer a family with a child who has sickle cell trait (P = .09). In contrast, few pediatricians (1.9%) or family physicians (2.1%) would refer a family with a child with sickle cell trait to a genetic counselor but would not refer a family with a child who is a cystic fibrosis carrier (P = .92).
To be successful, newborn screening must involve a complex relationship between public health programs, primary care providers, subspecialists, genetic counselors, families, and payers of health care services. Although primary care providers are expected to provide follow-up care for children with positive newborn screening results, little formal guidance was available before the availability of the ACT sheets. In fact, we were unable to identify any explicit recommendations within the National Guideline Clearinghouse (a comprehensive database of evidence-based clinical practice guidelines) for any of the conditions now included in newborn screening panels.14 No data are available regarding the degree to which newborn screening programs assist with and coordinate follow-up care for children with positive newborn screening results. Our anecdotal experience is that the availability and completeness of these services vary according to state and condition, ranging from careful care coordination to only state-level tracking of the number of children who are confirmed to be affected. However, the availability of clear concise guidelines may not be sufficient to eliminate inappropriate variations in care.15 For example, although there are explicit guidelines for the care of children with blood lead poisoning, approximately one half of the Medicaid-enrolled children in Michigan with elevated blood lead levels did not receive follow-up testing, although many of them had subsequent medical encounters.16
A key issue is the degree to which variation in care is acceptable. Some variation is acceptable because of incomplete medical knowledge (eg, thresholds for test results or effectiveness of specific interventions or treatment) or the local availability of services. Guidelines for follow-up care of children with positive newborn screening results may help minimize inappropriate variations in care if there is sufficient public health infrastructure to assist primary care providers. In this study, we found that most primary care physicians thought that they were responsible for newborn screening follow-up care. However, more than one half of the respondents would prefer their state's newborn screening program to provide the initial evaluation of positive newborn screening results. Any new public health efforts to support primary care providers will need to consider not only factors related to the conditions in the screening panels but also the availability of local resources and the training and experience of the primary care providers in the community. More effort is needed to ensure that family physicians are included in any educational outreach programs.
Currently, many primary care providers feel unprepared to manage follow-up care for a child with a positive newborn screen. We also found important differences between pediatricians and family physicians. For example, nearly 20% of the pediatricians and one half of the family physicians reported that they were not competent to discuss PKU, a condition that has been part of newborn screening for decades, with a family after a positive newborn screen. We were not surprised that most respondents did not feel competent to discuss MCAD, one of the newest conditions to be added to newborn screening panels, with a family after a positive newborn screen. However, we were surprised that pediatricians in states that screen routinely for MCAD were more likely than those in states that do not offer such screening to be uncertain how to confirm the diagnosis or where to refer a child with confirmed MCAD. It may be that pediatricians in states that screen for MCAD have given more thought to this challenge or have had the experience of not knowing what to do after a patient tests positive for MCAD.
We found evidence of misconceptions about the conditions in newborn screening panels. For example, one half of the respondents would refer families with a child with congenital hypothyroidism for genetic counseling, which is unlikely to be helpful. In contrast, fewer respondents would refer families with a child who has sickle cell trait, compared with families with a child who is a cystic fibrosis carrier, for genetic counseling. Our survey cannot identify the underlying reason for this difference. Possible explanations include differences in how physicians perceive these conditions or differences in how they perceive the impact of genetic counseling for families with these conditions. Interestingly, the new ACT sheets explicitly recommend offering referral for genetic counseling for children with sickle cell trait but have no similar recommendation for those found to be cystic fibrosis carriers.
Although the individual conditions in newborn screening panels are rare, most pediatricians and approximately one half of family physicians had a child with a positive newborn screen in their practice within the past 5 years. The number of both true- and false-positive newborn screening results will continue to increase in the coming years as the panels expand to include more conditions, which will increase the burden on public and private health systems and on families. The most important limitation of this study is that we were unable to assess directly the relationship between the physicians' self-reported attitudes and practices in actual care delivery and individual child and family outcomes. Future studies are needed to assess directly the health care utilization and health outcomes after positive newborn screening results, to assess and to refine the ACT sheets and to develop new interventions to ensure timely, appropriate, equitable health care delivery.
This study was funded by the Child Health Evaluation and Research Unit, Division of General Pediatrics, University of Michigan.
- Accepted July 20, 2006.
- Address correspondence to Alex R. Kemper, MD, MPH, MS, Duke Clinical Research Institute, PO Box 17969, Durham, NC 27715. E-mail:
The authors have indicated they have no financial relationships relevant to this article to disclose.
- ↵Bordley WC, Travers D, Scanlon P, Frush K, Hohenhaus S. Office preparedness for pediatric emergencies: a randomized, controlled trial of an office-based training program. Pediatrics.2003;112 :291– 295
- ↵National Newborn Screening and Genetics Resource Center. National newborn screening status report. Available at: http://genes-r-us.uthscsa.edu/nbsdisorders.htm. Accessed May 17, 2006
- ↵Gruskin A, Williams RG, McCabe ERB, et al. Final report of the FOPE II Pediatric Subspecialists of the Future Workgroup. Pediatrics.2000;106 :1224– 1244
- ↵American College of Medical Genetics. Newborn screening ACT sheets and confirmatory algorithms. Available at: www.acmg.net/resources/policies/ACT/condition-analyte-links.htm. Accessed May 17, 2006
- ↵Acharya K, Ackerman PD, Ross LF. Pediatricians' attitudes toward expanding newborn screening. Pediatrics.2005;116(4) . Available at: www.pediatrics.org/cgi/content/full/116/4/e476
- ↵Koopmans J, Ross LF. Does familiarity breed acceptance? The influence of policy on physicians' attitudes toward newborn screening programs. Pediatrics.2006;117 :1477– 1485
- ↵Desposito F, Lloyd-Puryear MA, Tonniges TF, Rhein F, Mann M. Survey of pediatrician practices in retrieving statewide authorized newborn screening results. Pediatrics.2001;108(2) . Available at: www.pediatrics.org/cgi/content/full/108/2/e22
- ↵Randolph GD, Pathman DE. Trends in the rural-urban distribution of general pediatricians. Pediatrics.2001;107(2) . Available at: www.pediatrics.org/cgi/content/full/107/2/e18
- ↵Kemper AR, Foster CM. Congenital hypothyroidism: a guide for the general pediatrician. Contemp Pediatr.2003;20 :32– 48
- ↵Council of American Survey Research Organizations. On the definition of response rates: CASRO Task Force Special Report, 1982. Available at: www.casro.org/resprates.cfm. Accessed May 17, 2006
- ↵Agency for Healthcare Research and Quality. National guideline clearinghouse. Available at: www.guideline.gov/. Accessed May 17, 2006
- Copyright © 2006 by the American Academy of Pediatrics