[Skip to Content]
Access to paid content on this site is currently suspended due to excessive activity being detected from your IP address Please contact the publisher to request reinstatement.
Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
[Skip to Content Landing]
March 1998

Listening to ParentsA National Survey of Parents With Young Children

Author Affiliations

From The Commonwealth Fund, New York, NY (Drs Young and Davis, and Ms Schoen); and the Division of Developmental and Behavioral Pediatrics, Boston Medical Center, Boston, Mass (Dr Parker).

Arch Pediatr Adolesc Med. 1998;152(3):255-262. doi:10.1001/archpedi.152.3.255

Objective  To document the child-rearing needs and pediatric health care experiences of parents with children from birth to 3 years old.

Design  A nationally representative sample of 2017 parents with children younger than 3 years using a 25-minute structured telephone questionnaire. Interviews were completed by 68% of the screened eligible respondents. The margin of sampling error for results at the 95% confidence level was ±3 percentage points.

Results  Seventy-six percent of children younger than 3 years were reported by parents to be in excellent health; 88% had a regular source of pediatric health care. Seventy-one percent of parents who received special pediatric services rated their child's physician as excellent in providing good health care. Seventy-nine percent of parents reported they could use more information in at least 1 of 6 areas of child rearing, and 53% wanted information in at least 3 areas. Forty-two percent had talked with their child's physician about "nonmedical" concerns; 39% of parents read to or looked at a picture book with their child on a daily basis; 51% of parents set daily routines for meals, naps, and bedtime. Breast-feeding and reading to the child on a daily basis were much more likely if a physician encouraged parents to do so.

Conclusions  Most parents view the pediatric health care system as meeting the physical health needs of their young children. Parents want more information and support on child-rearing concerns, yet pediatric clinicians often fail to discuss nonmedical questions with them. The interventions of pediatric clinicians can positively affect parental behavior. Pediatric practices should consider creative ways to reconstitute and augment their current services and systems of care.

THE IMPORTANCE of child-behavior and parenting concerns within pediatric practice has increased with the decline in morbidity and mortality from childhood infectious diseases and the rapid pace of social changes affecting family life. In 1958, a pediatrician reported that parents' developmental concerns about their children accounted for less than 2% of his practice time.1 By the mid 1960s, reports from individual primary care practices found that 45% of mothers were more concerned about their children's development and behavior than all other issues.2

More recently, studies have reported that parents continue to have such questions and are eager for specific information on child behavior and parenting. In fact, these nonmedical issues frequently are the parents' primary concern. McCune et al3 reported that 81% of parental questions for pediatricians concerned psychosocial issues. Hickson and associates4 confirmed this finding. In their study, parenting issues were parents' predominant concern: 70% of mothers were more worried about some aspect of their parenting or their child's behavior than they were about their child's physical health.

Despite these concerns, only 28% of mothers indicated that they had discussed them with their pediatrician.4 Sharp et al5 reported that opportunities existed to discuss psychosocial concerns in 88% of health supervision visits, but in 60% of these visits the physician ignored the concern or provided no information or guidance to the parent about it. Reisinger and Bires6 found that pediatricians spent an average of less than 11/2 minutes on anticipatory guidance during a routine visit with a young child.

In sum, the majority of parents are concerned about and would like their pediatric clinicians to address child-behavior and parenting issues. Building from the findings in the literature, the objectives of this study were 2-fold: first, to document the child-rearing concerns and child health care experiences of a national sample of parents of children from birth to 3 years old; and second, to learn from parents what child health care providers might do to support parents with young children. Based on our findings, we make suggestions for pediatricians and pediatric practices in a changing health care system.


This study was based on a series of 25-minute, in-depth, structured telephone interviews, using a stratified random-digit-dial sample design to obtain a nationally representative sample of parents with children aged birth to 3 years. Eighty-one percent of the respondents completed the screening, and 68% of those screened and eligible completed the interview. (The Commonwealth Fund Survey of Parents With Young Children7 was designed and analyzed by Princeton Survey Research Associates, Princeton, NJ, in collaboration with the Fund. Interviewing was conducted by DataStat, Inc, Ann Arbor, Mich.) Eligible parents were told that the interview had the support of the American Academy of Pediatrics. They were also told that we hoped to learn about the kinds of guidance and help parents need in rearing children and to determine what the health system could do to support parents with infants and toddlers.

A representative sample of 2017 parents with children younger than 3 years was obtained. An oversample (a sample of larger numbers of a group than their proportion in the population) of 397 non-Hispanic African American parents and 419 Hispanic parents were interviewed to ensure an accurate representation of parents' views among these groups. The results were weighted to reflect the overall population distribution. The margin of sampling error for results at the 95% confidence level, based on the total sample, was ±3 percentage points.

The sample included both mothers and fathers who were biological parents, adoptive parents, stepparents, or guardians of a child younger than 3 years living in their household, but excluded parents who did not live with their child. The child's mother or father was randomly selected to complete the interview. More mothers than fathers were interviewed (1320 mothers vs 697 fathers).

Two limitations to this study should be acknowledged. First, mothers' and fathers' responses reflect their own perceptions of their parenting behaviors, the health of the child, and the performance of health care providers. The study design did not allow for validation of parents' responses. Second, the survey asked parents about the "child's doctor or nurse." No attempt was made to determine whether the practitioners were pediatricians, family physicians, or nurse practitioners.


Today's children younger than 3 years are growing up in families that vary in many ways. In our sample, 72% of children younger than 3 years were living with married parents, 10% were living with both parents as a family (even though the parents are unmarried), and 18% were living with 1 parent (typically their mother). Only 6% of parents were younger than 20 years; 18% were older than 35 years. The median family income was about $30000 (Table 1).

Table 1. 
Family Demographics (n=2017)*
Family Demographics (n=2017)*

The majority of babies and toddlers, as reported by their parents, were in excellent health and have a regular source of pediatric health care (Table 2). Children in low-income families were less likely to be reported in excellent health and to have a regular source of care (P<.001). Two thirds of parents whose child had a regular source of care reported that their child's physician did an "excellent" job of providing overall pediatric care and listening carefully and answering questions. (The survey used the general term "doctor" and did not ask the parent to specify whether their child saw a pediatrician, a family practitioner, or a general practitioner.)

Table 2. 
Pediatric Health Status and Care by Income (n=2017)*
Pediatric Health Status and Care by Income (n=2017)*

Many parents, however, were less satisfied with the extent to which their child's regular physician helped them understand their child's growth and development and how to care for the child. Parents were also less satisfied with their ability to reach their child's physician or nurse by telephone (Table 3).

Table 3. 
Parent Satisfaction With Pediatric Care (n=1488)*
Parent Satisfaction With Pediatric Care (n=1488)*

Special pediatric services, such as home visits, developmental assessments, telephone information lines focused on child behavior and development, and records to chronicle the child's health and development, were valued by parents. However, most parents were not receiving such services from their child's physician or health plan. For example, only 20% of parents reported a home visit by a nurse or other health professional to teach them about newborn care. Two thirds (67%) of parents who received a follow-up home visit found the visit to be very useful (Table 4).

Table 4. 
Special Pediatric Services (n=1488)*
Special Pediatric Services (n=1488)*

Among parents who received additional services, a majority found them useful. Parents who received information and comprehensive pediatric services report significantly higher levels of satisfaction with their child's physician than parents who did not receive these services (Table 5). Almost three quarters of parents who received 3 or more special pediatric services rated their child's physician as excellent on providing good health care and guidance and in helping them understand their child's growth. Fewer than half of those who received none of these services give their pediatrician such a rating (P<.001). Whether or not parents received such services, almost two thirds said they would be willing to pay $10 a month for them.

Table 5. 
Parents Who Rate Pediatric Care Professionals as Excellent*
Parents Who Rate Pediatric Care Professionals as Excellent*

A majority of parents reported using multiple sources for information on child development and child-rearing practices. Almost three quarters (74%) of the parents sometimes used books, magazines, television, and videos to help answer their questions. Participation in parenting classes, however, was far less common: only 35% of parents had attended such a class or discussion.

Even so, many parents desire expert guidance and information on a wide range of nonmedical child-rearing topics (Table 6). Parents of infants and toddlers were eager to have more expert information on "encouraging their young child to learn." They also expressed a strong interest in receiving more information on other behavioral aspects of child rearing, such as discipline, toilet training, sleep patterns, and responding to a crying baby. However, fewer than one quarter of parents with infants and toddlers actually talked with their pediatric clinician about how to help their child learn or about how to discipline their child, and fewer than one third talked with their physician or nurse about toilet training.

Table 6. 
Child Rearing*
Child Rearing*

Most parents (79%) reported they could use more information in at least 1 of 6 areas of child rearing: newborn care, sleep patterns, how to respond to a crying baby, toilet training, discipline, and encouraging early learning. More than half (53%) wanted more information or help in at least 3 areas.


Scientific research emphasizes the rapid brain development in children younger than 3 years, and that everyday parental activities such as reading, singing, and being affectionate are important influences in a child's healthy cognitive and psychosocial development.8 Our results demonstrated, however, that many parents are missing opportunities to foster their infant's or toddler's early development (Table 7). For example, only 39% of the parents reported reading or looking at a picture book with their child on a daily basis. Although book-sharing activities were higher for toddlers, rates remained low: 48% of parents in the study with children aged 1 to 3 years read to or showed a picture book to their child at least once a day (Table 8). In addition, reading to a child once a day or more correlated with education and income (Table 9). Our findings also reveal that parents who speak with their physician or nurse about encouraging their child to learn are more likely to read to their child on a daily basis, compared with parents who do not discuss learning with their child's physician (47% vs 37%, P<.001).

Table 7. 
Parent Activities With Child in the Week Before the Survey (n=2017)*
Parent Activities With Child in the Week Before the Survey (n=2017)*
Table 8. 
Reading to Child by Age of Child (n=2017)*
Reading to Child by Age of Child (n=2017)*
Table 9. 
Reading to Child by Income and Education (n=2017)*
Reading to Child by Income and Education (n=2017)*

While most pediatric clinicians stress that regular routines are important for healthy social and behavioral development,9 in this study, only 51% of parents set daily routines for meals, naps, and bedtimes (Table 10). Regular routines were less common among single parents (P<.05) and among low-income families (P<.001) (Table 11).

Table 10. 
Daily Routines (n=2017)*
Daily Routines (n=2017)*
Table 11. 
Children's Daily Routine by Income and Family Type*
Children's Daily Routine by Income and Family Type*

Nine percent of mothers and 4% of fathers reported having experienced 3 to 5 depressive symptoms at least some of the time during the week before the survey. These included "feeling depressed," "feeling sad," or "feeling disliked"; "experiencing crying spells"; and "not finding life enjoyable." Mothers and fathers who experienced 3 to 5 depressive symptoms were significantly more likely than parents with no depressive symptoms to say that they became frequently frustrated with their child's behavior in a typical day (P<.001). They were less likely to maintain daily nap, meal, and bedtime routines or to read to their child daily; and they were more likely to yell at their young child (Table 12). In addition, parents who reported that they were physically or sexually abused as a child, compared with parents who were not abused, were more likely to exhibit depressive symptoms (12% vs 5%) and were more likely to become frustrated frequently with their child (30% vs 25%).

Table 12. 
Depressive Symptoms and Daily Child-Rearing Practices*
Depressive Symptoms and Daily Child-Rearing Practices*

Despite the medical evidence about the favorable consequences of breast-feeding, 46% of mothers in this study did not breast-feed their babies. When they did breast-feed, many stopped within a month. Breast-feeding varied among mothers, with mothers who were younger, lower income, and African American being less likely to breast-feed than other mothers. (Table 13). Breast-feeding was much more likely if mothers were encouraged to do so by a physician or nurse: 74% of mothers who were encouraged to breast-feed their babies actually did, whereas only 45% of those who did not receive encouragement chose to breast-feed their infant (P<.001).

Table 13. 
Breast-feeding by Parent Characteristics (n=1283)*
Breast-feeding by Parent Characteristics (n=1283)*

Almost all children younger than 3 years have medical insurance. However, almost a quarter of families in this study with annual incomes less than $40000 reported having difficulty paying for their child's medical expenses (Table 14). More than half of the parents had insurance coverage for their child that requires parents to pay part of the coverage themselves. Medicaid was a major source of health insurance coverage, covering 28% of children younger than 3 years. Medicaid was more likely to cover preventive services, such as immunizations and routine well-child visits, and visits when the child was sick, than was private insurance (Table 15).

Table 14. 
Child's Health Insurance Coverage, Source, and Type by Income (n=2017)*
Child's Health Insurance Coverage, Source, and Type by Income (n=2017)*
Table 15. 
Services Paid in Full by Health Insurance*
Services Paid in Full by Health Insurance*

This report contains important implications for pediatric clinicians. It represents the collective voice of 2000 families with young children—families from all over the country and from all walks of life. The use of an in-depth structured questionnaire allowed the interviewers the luxury of extended conversation with mothers and fathers. Parents responded by telling us of their concerns, how they are faring, and what they would like from their pediatric clinicians.

Their message to us is a complex one, at once reassuring and discomfiting. First, the encouraging news: the glass is half full.

1. The vast majority of children have an identified and stable point of pediatric care. Two thirds of parents said their child's physician or nurse is doing an "excellent" job in providing health care for their child and rated as "excellent" their physician's or nurse's ability to listen and answer questions. It seems that most families view their pediatric health care providers as meeting their child's physical health needs, and they are generally satisfied with the physician or nurse who provides that care.

2. Most children are reported as healthy. A very small percentage (2%) of parents described their child's health as "only fair." The rest of the children were described as in "excellent"or "good" health. In some ways the children of America have never been so healthy.

3. The interventions of pediatric clinicians positively affect parental behavior. Parents told us that if pediatric clinicians discuss certain things with them, they respond, for example, by breast-feeding more (74% vs 45%, P<.001) and by reading more frequently (47% vs 37%, P<.001) to their children. The authority and voice of pediatric clinicians still have clout with parents. The trustful encouragement and advice of pediatric clinicians seems to leave its mark on parents' behavior. So too, silence represents a missed opportunity to help families with young children.

But in counterpoint to these encouraging findings lie challenges: the glass is half empty.

1. Parents want more information and support for their concerns about child rearing. To be sure, there is plenty of information available to families, and parents use it. Yet, despite the extensive sources of child-rearing information, more than half of parents would like more information, especially about discipline, toilet training, sleep issues, crying, and how to help their child to learn.

Why then do parents want more support and information? Child rearing is a tricky business. There are rarely "one-size-fits-all" solutions to subtle psychosocial concerns. It is generally in the context of a supportive relationship that information can be not only transmitted but also discussed, personalized, and crafted to the family's specific needs. As often the only professional interacting with the family in the first years of their child's life, the pediatric clinician is ideally poised to provide such information in the context of an ongoing supportive relationship. When pediatricians talk with parents of infants and toddlers about their nonmedical concerns, they can elicit important information for identifying children with behavioral and developmental problems, such as language or motor delays. Individualizing the content of anticipatory guidance and responding to a parent-led agenda is not necessarily more time consuming, but it is likely to be more effective in addressing parents' needs.10

We were surprised to find "helping and encouraging children to learn" at the top of the list of parental desire for more information. This is not an area typically addressed by pediatric clinicians and is rarely taught in pediatric training. In the early days of the "new morbidity," psychosocial and behavioral "problems" were emphasized. As a result, some pediatric clinicians have learned new strategies to address thorny behavioral issues and sticky family dilemmas.

More recently, attention has been directed toward anticipatory guidance and the promotion of "wellness." But the world has continued to change, and it seems that parents are telling us that pediatric practice has not kept up with their current concerns. The pace of change in our society is escalating; children will have to make their way in a technological society that bears little resemblance to the world of their parents' childhood. In that context, it makes sense that the number one concern of mothers and fathers would be the child's ability to learn, that is, adapt and succeed in the ever-changing (and difficult-to-anticipate) circumstances of the 21st century. Perhaps the pediatric agenda for routine health care maintenance should once again consider reconfiguring itself to address the new priority concerns of mothers and fathers.

2. Pediatric clinicians often fail to discuss the nonmedical questions on parents' minds. While parents in this study were happy with their pediatric clinician's ability to provide good health care, only slightly more than half of them were happy with the pediatric clinician's guidance on how to care for their child and help them understand how their child is growing and developing. Physicians rarely provide help on even routine behavioral and developmental issues; only a few parents reported ever having discussed discipline, crying, and toilet training with their pediatric clinician.

The ability of pediatric clinicians to address psychosocial concerns—and make parents aware that they are capable of addressing those concerns—is critical to successful health supervision visits. Addressing them can often relieve parents' anxiety, improve interaction skills of parents and their young children, and potentially avoid having parents return again for the same concern.5

Increasingly, there is wide support for behavioral pediatrics in pediatrics and family practice training programs to improve the training, knowledge, and skills of physicians. The Task Force on Pediatric Education, the Guidelines for Child Health Supervision II,11 and Bright Futures: Guidelines for Health Supervision of Infants, Children and Adolescents12 all conclude that behavioral and social concerns require more attention in primary care and in education for primary pediatric practice. Major reform, however, has just now begun to take place, with changes occurring in both graduate medical education and residency programs. As a result, the majority of pediatricians in practice today do not feel well prepared to deal with the [not so new] new morbidity.

We must also be realistic in realizing that despite all these efforts there are still limitations on a clinician's office visit time within the current system of care. After all, very real limits exist on what can be accomplished within the constraints of a 15- to 20-minute visit.

To adequately address parents' nonmedical concerns, changes in how pediatric care is practiced and in priorities of the routine visit may be required. Greater use of nonphysician specialists and the introduction of group well-child visits could provide more time for teaching, anticipatory guidance, and discussion of behavioral issues.13,14 In addition, while many offices have designated call-in numbers for health questions, a carefully supervised designated call-in number for dealing with parents' child-rearing and behavioral concerns could also become a routine part of primary pediatric care.

3. Mothers and fathers do not report engaging in important activities with their children on a regular basis. For example, fewer than 1 in 5 parents in this study read or looked at a book with their child more than once a day; only slightly more than one third sang or played music more than once a day; and almost one third of parents did not even play with their child more than once a day. Parents are busy, to be sure, and there just does not seem to be enough time to do even more, especially if the importance of these activities in the child's development is undervalued. Since the respondents said that they were more likely to read to their child if encouraged to do so by their pediatric clinician, this represents another area for pediatric clinicians potentially to address.

4. Some parents are having a hard time of it. This study revealed a clinically important and troubling association between parental depressive symptoms and daily child-rearing activities, such as meal and bedtime routines, being affectionate, playing with the child, and engaging in activities such as reading or playing music. What should be the pediatric clinician's role with regard to parental depressive symptoms? After pregnancy, the pediatric primary care system may be the only health system with which parents are consistently involved; as such, it affords a critical opportunity to address parents' mental health.15 Yet, pediatric clinicians are not expert in recognizing the signs of depression in parents or children. If clinicians are to support families, addressing nontraditional issues such as parental depression may need to become a routine part of a "2-generation approach" to pediatric health care maintenance.


These findings offer a workable blueprint for pediatric clinicians who want to provide the assistance families are looking for. Feasible measures include:

1. Pediatric clinicians should focus more on helping children to learn. In some ways, this could become an organizing goal of all of pediatrics: to help raise children who are sufficiently healthy, happy, motivated, and confident, so they can learn to succeed in the world.16 This means that anything that could affect learning might potentially become the province of the pediatric clinician—from the promotion of breast-feeding (for its positive effect on health and, possibly, cognition), to early reading and music to promote active brain development, to addressing family issues, to promoting the child's self-esteem. The focus for each child and each family would, of course, be different. For some, medical issues will predominate; for others, family issues; and for others, more subtle aspects of enhancing motivation and joy in learning.

2. Pediatric practices should consider creative ways to reconstitute and augment their current services with the support of administrative and financial systems. In this study, mothers and fathers voiced their interest in enhanced pediatric services, such as home visiting, developmental assessments, special telephone advice lines, and child health and developmental records. They reported more satisfaction with their physician when they received these services.

With the advent of managed care and capitation, such services will also have to demonstrate value to the insurers. Data from this study suggest that parents are willing to pay for expanded services; perhaps even more would "vote with their feet" and offer a competitive advantage to practices that provide more services. If the nonmedical needs of the family are to be met by the pediatric clinician, it can only happen when plan administrators provide operational and financial support for these services in the context of systemic change.

In today's competitive health care environment, managed care has the potential to advance goals that support children and families. Health plans can provide opportunities for families to receive comprehensive services from a physician who knows the child and parents; to reduce fragmentation and unnecessary visits to the emergency department; and to establish accountability and reduce costs. Managed care plans have market incentives to build strong relationships with families, educate parents to keep children healthy and accident free, and link pediatrics to prenatal care and other adult health services. Despite the rhetoric in support of preventive care and wellness, no new resources, to our knowledge, have been committed to preventive services under managed care.

The time is right to begin reconceptualizing how the needs of young children and their families may be met. With the expansion of managed care, there is both the opportunity and the challenge to creatively reconfigure the current model of pediatric practice. Can a solo clinician really be expected to address all of the family needs in a capitated system? Competition and financial pressures will require pediatricians to work with each other and other nonphysician specialists such as nurses, early childhood educators, and psychologists in expanded roles of delivering pediatric care.15,17,18 In addition, and as mentioned earlier in this article, pediatric practices might find it more cost-effective to offer group well-child visits and a designated telephone line to discuss mothers' and fathers' behavioral concerns.

These proposed new interdisciplinary partnerships and ways of delivering well-child care could be viewed as a threat to some pediatricians. But if properly carried out within organizations that provide clinical autonomy and financial accountability under the leadership of pediatricians and other health care providers, these new models of practice could be a boon to children's health and development.17 They could also improve the quality and efficiency of pediatric care.

The methods for delivering pediatric care are changing. What parents in this study have provided is a possible blueprint for action to help guide the creation of new care models that would better address the concerns of families in the 21st century.

Back to top
Article Information

Accepted for publication October 28, 1997.

We wish to acknowledge Shawn V. LaFrance, MPH, Linda Greenberg, PhD, and Tammi Troy for their contributions in preparing the manuscript. We also wish to thank Paul Dworkin, MD, Joel Alpert, MD, and Barry Zuckerman, MD, for their thoughtful comments in reviewing the manuscript.

Corresponding author: Kathryn Taaffe Young, PhD, The Commonwealth Fund, One East 75th St, New York, NY 10021 (e-mail: kty@cmwf.org).

Boulware  JR The composition of private pediatric practices in a small community in the south of the United States. Pediatrics. 1958;225- 18
Deisher  RWEngel  WISpielholz  R  et al.  Mothers' opinions of their pediatric care. Pediatrics. 1965;3582
McCune  YDRichardson  MMPowell  JA Pyschosocial health issues in pediatric practices: parents' knowledge and concerns. Pediatrics. 1984;74183- 190
Hickson  GBAltemeler  WAO'Connor  S Concerns of mothers seeking care in private pediatric offices: opportunities for expanding services. Pediatrics. 1983;72619- 624
Sharp  LPantell  RHMurphy  LOLewis  CC Psychosocial problems during child health supervision visits: eliciting, then what? Pediatrics. 1992;89619- 623
Reisinger  KBires  J Anticipatory Guidance in Pediatric Practice. Pediatrics. 1980;66889- 892
Young  KTDavis  KSchoen  C The Commonwealth Fund Survey of Parents With Young Children.  New York, NY The Commonwealth Fund1996;
Shore  R Rethinking the Brain: New Insights Into Early Development.  New York, NY Families & Work Institute1997;
Parker  SedZuckerman  Bed Behavioral and Developmental Pediatrics.  Boston, Mass Little Brown & Co Inc1995;
Glasco  EPDworkin  PH Obstacles to developmental surveillance. J Dev Behav Pediatr. 1993;14344- 349
American Academy of Pediatrics, Committee on Psychsocial Aspects of Child and Family Health, Guidelines for Child Health Supervision II.  Elk Grove Village, Ill American Academy of Pediatrics1988;
Green  Med Bright Futures: Guidelines for Health Supervision of Infants, Children and Adolescents.  Arlington, Va National Center for Education in Maternal and Child Health1994;
Dodds  MNicholson  LMuse  BOsborn  LM Group health supervision visits: more effective than individual visits in developing health care information. Pediatrics. 1993;91668- 670
Osborn  LMWooley  FR Use of groups in well child care. Pediatrics. 1981;67701- 706
Zuckerman  BParker  S Preventive pediatrics—new models of providing needed health services. Pediatrics. 1995;95758- 762
Dworkin  PH Ready to learn: a mandate for pediatrics. J Dev Behav Pediatr. 1993;14192- 196Article
Lewin  LS Adapting your pediatric practice to the changing health care system. Pediatrics. 1995;96799- 803
Nazarian  LF A look at the private practice of the future. Pediatrics. 1995;96812- 815