Objective. Previous work has focused attention on the prevalence of specific maternal health problems known to affect children, such as smoking or depression. However, the cumulative health burden experienced by mothers and the potential for a practical pediatric health services response have not been examined. The aims of this study were to characterize: 1) the prevalence and cumulative burden of maternal health behaviors and conditions, 2) maternal access to a source of comprehensive adult primary care, and 3) maternal perceptions of a pediatric role in screening and referral.
Methods. We surveyed 559 consecutive women bringing a child 18 months of age or less to one of four pediatric primary care sites between July 1996 and May 1997. The pediatric sites included one outpatient program in an academic hospital, one in a community health center, and two in-staff model practices of a managed care organization (these last two were combined for analysis). The self-administered questionnaire contained previously validated questions to assess health behaviors and conditions (smoking, alcohol abuse, depression, violence, risk for unintended pregnancy, serious illness, self-reported health) and access to care (regular source, regular provider, health insurance, care delayed or not received). Maternal attitudes toward a pediatric role in screening and referral were also elicited.
Results. In the three settings, response rates ranged from 75% to 84%. The average age of the women ranged from 25.1 to 32.1 years and the average age of the children ranged from 6.5 to 8.0 months. Across the settings, the percentage of women reporting at least one health condition (66%–74%) was similarly high, despite significant demographic differences among sites. Many women reported more than one condition (31%–37%); among all women who smoked, 33% also screened positive for alcohol abuse, 31% for emotional or physical abuse, and 48% for depression. Access to comprehensive adult primary care was variable with 23% to 58% of women reporting one or more barriers depending on the site. Across all sites, >85% of mothers reported they would “not mind” or “would welcome” a pediatric role in screening and referral.
Conclusions. Two-thirds of women bringing their children for pediatric care had health problems regardless of the site of care. Many women also reported substantial barriers to comprehensive health care. Most women reported acceptance of a pediatric role in screening and referral. Given the range and depth of maternal health needs, strategies to connect or reconnect mothers to comprehensive adult primary care from a variety of pediatric settings should be explored.
Although research on the relationship between maternal and child health has traditionally focused on pregnancy and the early postpartum period, the relationship between the health of women and that of their children extends well into childhood. The links between maternal smoking and increased rates of sudden infant death syndrome,1 otitis media,2,,3 and asthma4,,5 in children are well established. Maternal depression is associated with childhood behavior problems, poor growth, and accidents.6,,7 Children from unintended pregnancies have been shown to have lower educational attainment8 and a higher risk for physical abuse or neglect9,,10 whereas children who witness domestic violence are susceptible to developmental problems and prone to model their parents' roles.11,,12
Although the prevalence of specific maternal health behaviors and conditions has been documented,13–15 there is little data on the cumulative burden of these conditions. Barriers to health care for women of reproductive age increase the likelihood that these conditions are inadequately addressed. Between 10% and 16% of women report no regular source of care16,,17 and 31% of low income women report no health insurance.18 Moreover, 45% of women aged 18 to 39 years have not had a routine check-up in the past year.19 Whether these figures are accurate for women who have recently delivered a child remains unexplored.
Although comprehensive solutions to the morbidity and access barriers experienced by women must await major changes in health care financing and public policy, the potential to use pediatric settings to enhance maternal and family health has drawn increasing attention.20,,21 The recently released Bright Futures: National Guidelines for Health Supervision of Infants, Children, and Adolescents22 promotes a general family screening approach while strategies to intervene with specific maternal health problems show some effectiveness.23Despite this recent interest, little empirical evidence exists regarding the practical requirements of a pediatric response to the full range of unmet maternal health needs.
The aim of this study was to lay the foundation for such a response by seeking to determine the full scope of health needs and access barriers facing women who seek pediatric care for their children. Specifically, we sought to characterize: 1) the prevalence and cumulative burden of maternal health behaviors and conditions, 2) maternal access to a source of comprehensive adult primary care, and 3) maternal perceptions of a pediatric role in screening and referral. Women with young children were the focus given the strong evidence that links the health of women to that of young children, the high frequency of pediatric contact, and the potentially difficult transition from pregnancy- to nonpregnancy-related health care.
We surveyed mothers of children receiving care at four pediatric sites in greater Boston between July 1996 and May 1997. The sites included one outpatient practice in an academic hospital (AHP), one in a community health center (CHC), and two in-staff model practices of a large managed care organization (MCO).
Women waiting for their child's appointment were consecutively screened by a research assistant in the waiting room. Women were eligible if they spoke English or Spanish, were not currently pregnant, and had a child that was aged 18 months or less that was not acutely ill.
The purpose of the survey was explained to eligible mothers and informed consent was obtained. At the first site (AHP), the survey was designed as a confidential survey to allow comparison of the demographic characteristics of responders with nonresponders. Given concern for underreporting of key issues of interest, eg, domestic violence, an anonymous survey was used for the next three sites. The survey, available in English and Spanish, was self-administered, took ∼10 to 15 minutes, and was completed in either the waiting area or the examination room. The research assistant offered to assist those mothers who were unable to read the survey.
The survey instrument focused on four areas: maternal health status, access to care, maternal perceptions of a pediatric role in screening and referral, and demographic characteristics (Table 1). The instrument was a composite of previously validated screening questions for maternal health status and access to care as well as new questions regarding maternal perceptions of a pediatric role.
Maternal health status was characterized by health conditions, health behaviors, and general health status. We used a three question screen for depression previously validated against the Diagnostic Interview Survey24 and used with mothers in the pediatric setting.15 Questions on verbal and physical abuse were taken from two studies of violence during the childbearing years.25,,26 The survey introduced domestic violence with the following paragraph:
“No matter how well people get along, there are times when they disagree, get annoyed or have fights. They also use many different ways of trying to settle their differences. These questions address whether you have been affected by violence as a way to settle differences.”
The questions themselves were as follows: 1) “Within the last year, have you been emotionally or verbally abused by your partner or someone important to you; for example, has anyone sworn at you, threatened to hit you or threatened to throw something at you?”; and 2) “Within the last year, have you been hit, slapped, kicked, or otherwise physically hurt by your partner or someone important to you?” A single question was used to prompt for significant or serious illnesses, followed by a checklist of conditions including asthma, hypertension, and diabetes.
Among the high-risk health behaviors, current smoking was identified using two questions taken from the National Health Interview Survey.27 Alcohol abuse was assessed using a four question instrument that is 90% sensitive when compared with the Michigan Alcohol Screening Test13 and incorporates two questions on tolerance and binge drinking shown to be particularly sensitive when used with women: 1) “How many drinks does it take to make you feel high or feel a buzz?”28 and 2) “Do you ever have 5 or more drinks at a time?”29 High risk for unintended pregnancy was based on questions assessing: 1) sexual activity in the past 3 months; 2) frequency of use of a method of birth control; and 3) not being pregnant, postpartum, or trying to conceive.10,,30,31 Finally, overall health status was assessed using standard self-report: “In general, would you say your health is: excellent, very good, good, fair, or poor?”32,,33 dichotomized as excellent/very good/good versus fair/poor.
Access to health care questions assessed the presence of one or more financial, structural, and functional barriers to care, using questions from the National Health Interview Survey on health insurance, regular source of care, regular provider at that site of care, and care delayed or not received.27 Three questions gauged maternal perceptions of a potential pediatric role: 1) “How much do you think the following health issues in a mother can affect the baby's health or well-being?” (not at all, some, a lot). 2) “How would you feel if your child's pediatrician or nurse practitioner asked you about the following issues during a well-child visit?” (welcome, not mind at all, mildly annoyed, very annoyed). 3) “If one of these issues actually affected you, how would you feel if your child's pediatrician offered to help you get an appointment with an adult care provider?” (welcome, not mind at all, mildly annoyed, very annoyed). Participants were invited to take the last page of the survey, which was a list of local health care and social support resources.
The primary outcome measures were: 1) the presence of one or more maternal health behaviors or conditions, and 2) the presence of one or more barriers to health care. Potential differences among settings in the epidemiology of specific health needs and access barriers were also examined and should be considered exploratory. The two MCO sites were combined for analysis because of similar sociodemographic profiles (except for race/ethnicity), however, the AHP and CHC had significantly different populations and were analyzed separately. χ2 statistics were used to test for any differences among these three settings. All the sites were combined for a descriptive analysis of maternal attitudes toward a pediatric role in screening and referral. All statistical tests were two-tailed. Given the number of statistical comparisons, P values should be considered significant at the P = .01 level. With a level required for statistical significance of α = 0.01, the study had adequate power (β = 0.20) to detect a difference of 0.20 among sites in the proportion of women affected by at least one health condition. All analyses were done using SAS software (SAS Institute, Cary, NC, Version 6.12).
A total of 559 women completed surveys with site-specific response rates ranging from 75% to 84%. Respondent's average age varied among sites from 25.1 years to 32.1 years with an average time since delivery of 6.5 to 8.0 months (Table 2). One-half of the AHP respondents were black and 60% had a household income under $20 000. The CHC population was primarily Hispanic with almost half having <12 years of education and 87% with income <$20 000. In contrast, the majority of the MCO respondents were white, non-Hispanic, college educated women with household incomes >$40 000. Nonrespondents, assessed only at the AHP site, were not significantly different with respect to the age of the mothers or their children, maternal education, race, or marital status when compared with AHP respondents. Nonrespondents were significantly more likely to have an income under $10 000.
Despite the demographic diversity, the percentage of women with one or more reported health conditions was similarly high across sites, ranging from 65.7% to 73.8% (P = .29) (Table 3). There was, however, variation in the prevalence of specific problems, including depressive symptoms and general health status. There was a trend for the CHC respondents to be less likely to screen positively for alcohol abuse, but more likely to report physical abuse and to screen positively for depression. Fifteen percent to 20% of women reported a serious or significant illness with no difference across sites (P = .44); 11% of all women reported being affected by hypertension or asthma.
There was substantial comorbidity among the women. More than 30% of women across all sites reported being affected by two or more problems, and 8.5% to 21.5% reported three or more (Table 4). For example, women who smoked were also at risk for alcohol problems (33%), unintended pregnancy (25%), emotional or physical abuse (31%), and depression (48%).
The proportion of women reporting specific barriers to health care varied more substantially across sites. Women bringing their children to a managed care site were significantly more likely to have health insurance themselves and to report a regular source of care (Table 5). However, even in the MCO sites 22.5% of mothers reported at least one access barrier. Across all sites, the percentage of women with at least one health need and at least one access barrier ranged from 16.5% to 41.2%.
More than 80% of women at each site recognized the potential impact of the specific health conditions on a child's health and well-being (Fig 1). The data for specific sites are combined, because the differences across sites were clinically small (eg, potential impact of depression, 86% vs 96%, P < .02). Mothers perceived smoking, heavy drinking, and violence between partners to have the greatest potential impact on their children whereas unplanned pregnancy was least often reported as important.
More than 85% of all women said they would welcome or not mind inquiries about each specific condition (Fig 1); the subset of women affected by a given condition were similarly receptive. Finally, more than 90% of women at each site said they would welcome or not mind an offer to assist with appointment-making with an adult care provider if affected by one of these issues. Acceptance of a pediatric referral role did not differ between those women with and without health conditions (97.1% vs 95.6%, P = .63).
Two-thirds of all women bringing their children for pediatric care reported at least one health condition or behavior, many faced financial, structural, or functional barriers to health care, and a significant proportion experienced both a health need and a barrier to care. The prevalence of unmet health needs is particularly striking given the short interval of time between pregnancy-related care and participation in this survey. Women were highly receptive to the notion of a pediatric role in screening and referral to an adult primary care provider.
The prevalence of individual health conditions and behaviors was generally consistent with the data of Kemper and colleagues,14,,34 whose work suggested that these problems are quite common in women with children. In our study the percentage of women reporting depressive symptoms was somewhat higher and the rates of smoking modestly lower. These differences may be accounted for by the younger age of the children in our sample and, with regard to the smoking, to secular trends. Moreover, there was variation in the prevalence of specific conditions between sites. This highlights the need for clinician sensitivity to the unique distribution of these problems in their own practice population.
Despite some site to site variation in the rates of specific conditions, there was considerable uniformity across sites in the cumulative health burden of women bringing their children for care: almost two-thirds of mothers bringing their children to care were affected by at least one important health problem. Our data suggest that these women not only experience relatively high rates of medical conditions and potentially harmful behaviors, but that conditions often coexist. One-third of smoking women screened positive for alcohol problems and abuse and one-half for depression posing a challenge to current strategies that encourage pediatricians to address specific maternal health problems, such as smoking.23 Rather, our data suggest that a more comprehensive response may be required. One important element of this response should be enhanced access to comprehensive medical care because between 15% and 20% of surveyed women reported experiencing potentially serious illness, including hypertension and asthma.
Our data also suggest that significant discontinuities in health care occur for women after they deliver. More than 20% of women in the CHC and academic clinics reported no regular source of care relatively soon after pregnancy. This may be explained in part by health insurance disenrollment for women who receive Medicaid only while pregnant35 and the lack of coordination of obstetrical care with other primary care services. Women in the managed care settings had significantly better access to health care compared with those in the other two sites. However, in a setting where virtually all families are insured and identify providers as part of enrollment, we were surprised that more than one in five women still reported a barrier to care. Although this study examined major barriers to care, more extensive work remains to be done to refine our understanding of unmet health needs. Our design did not allow us to probe maternal awareness of each identified health problem nor could we assess the adequacy of treatment received for those women with a regular doctor. Such information would be extremely relevant to the development of focused and effective interventions.
Our study is strengthened by the broad demographic heterogeneity we achieved in surveying a range of clinical settings. However, because the survey could not reach mothers who do not regularly bring their children to pediatric care, the prevalence of unmet health needs may underestimate that of the general population. In addition, our results should be confirmed in other practice settings, such as small private practices. Nonresponse bias may also have influenced the results; data from a sample of nonrespondents suggest that they had lower incomes that might bias our findings toward an underestimate of unmet health needs. Moreover, because the data relied on self-report for such sensitive issues as domestic violence and alcohol abuse, there exists the potential for underreporting by affected individuals. Conversely, the reported acceptance by women of pediatric screening and referrals may be overstated because of the survey's administration in a pediatric setting. Nevertheless, the magnitude of the principal findings make it unlikely that these limitations undermine the basic conclusions reported here.
This study not only found that women bringing their children for pediatric care have major health needs and face access barriers, it found that these women almost uniformly accepted a pediatric role in screening and referral to adult primary care. This underscores the opportunity afforded by the pediatric visit to address maternal health and should ease fears physicians may have of offending families when screening for behavioral or psychosocial issues in a clinical encounter.36,,37 Moreover, the development of a referral linkage between pediatric and adult health clinicians might offer a system of care to complement the work of family physicians in ensuring health care for all members of a family.
It is essential to recognize that the clinical response to these findings must be shaped by the serious time constraints clinicians already face. Although the parental health needs explored in this study are entirely consistent with those raised by Bright Futures,22 the development of intervention strategies that respect the demands of a busy clinical practice remain an urgent challenge. By outlining the depth of maternal health needs and the receptiveness to a pediatric response, we hope to attract the attention and creative energy of the practicing community, which in many ways represents the best guarantee that any ameliorative strategies will prove both effective and intensely practical. In addition, full recognition of parental health needs will best ensure that our advocacy for improvements in health care for children will be linked to broader reforms essential to family health and well-being.
Given the range of health problems and comorbidity involved, we believe one practical response may be an effort to connect or reconnect women to comprehensive adult health care. A screening and enhanced referral approach could supplement other strategies, including counseling or selected treatment in the pediatric setting. The precise components of such referral efforts should reflect the strengths and limitations of a given pediatric setting. Such considerations might include the colocation of child and adult health services in clinical facilities, availability of office support staff, emerging partnerships between academic hospitals and private practices, and shared delivery and information systems of MCOs. The refinement of effective systems of referral, however, will require a full assessment of whether referred women follow through with arranged appointments and, if so, whether the identified health problems are addressed effectively.
Child health professionals could contribute substantially to the health of women and children by ensuring that women who bring their children for care have their own health needs addressed in the context of comprehensive adult health care. In this manner, use of the pediatric visit to address maternal health need not imply a greatly expanded expertise, but merely a more practical expression of a longstanding commitment to family health.
This work was funded by the Ambulatory Pediatric Association, the Harvard Pilgrim Health Care Foundation, Partner's Healthcare, the Peabody Foundation, and the Klingenstein Foundation. Dr Kahn was a fellow in a National Research Service Award Primary Care Research Training Program at Children's Hospital, Boston, when this study was undertaken.
We thank Drs Howard Bauchner and Barry Zuckerman for their helpful comments. The research assistance of Rachel Bird was invaluable to the project. The contributions of the clinical staff at Pediatric Health Associates, Harvard Pilgrim Health Care and Chelsea Health Care Center were also much appreciated.
- Received January 7, 1998.
- Accepted August 5, 1998.
Reprint requests to (R.S.K.) Department of Pediatrics, Maternity Bldg Room 414, Boston University School of Medicine/Boston Medical Center, 91 East Concord St, Boston, MA 02118.
- AHP =
- academic hospital practice •
- CHC =
- community health center •
- MCO =
- managed care organization
- Ey J,
- Holberg C,
- Aldous M,
- et al.
- DiFranza JR,
- Lew RA
- ↵Environmental Protection Agency. Respiratory Health Effects of Passive Smoking: Lung Cancer and Other Disorders. Washington, DC: Government Printing Office; 1992
- Zuckerman BS,
- Beardslee WR
- Beardslee WR,
- Keller MB,
- Lavori PW,
- Staley J,
- Sacks N
- ↵Institute of Medicine. In: Eisenberg L, Brown SS, eds. The Best Intentions: Unintended Pregnancy and the Well-being of Children and Families. Washington, DC: National Academy Press; 1995
- Wolfe DA,
- Korsch B
- Kemper KJ
- ↵Cornelius L, Beauregard K, Cohen J. Usual Sources of Medical Care and Their Characteristics. National Medical Expenditure Survey Research Findings 11. Rockville, MD: Public Health Service; 1991. (AHCPR Publ. No. 91–0042)
- ↵Leiman JM, Meyer JE, Rothschild N, Simon LJ. Selected Facts on US Women's Health: A Chart Book. New York, NY: Commonwealth Fund; 1997
- ↵Brown ER, Wyn R, Cumberland WG, et al. Women's Health-Related Behaviors and Use of Clinical Preventive Services. Los Angeles, CA: Commonwealth Fund; 1995
- Zuckerman B,
- Parker S
- Klerman L,
- Reynolds DW
- ↵Green M. Bright Futures: National Guidelines for Health Supervision of Infants, Children, and Adolescents. Arlington, VA: National Center for Education in Maternal and Child Health; 1994
- Wall MA,
- Severson HH,
- Andrews JA,
- Lichtenstein G,
- Zoref L
- Benson V,
- Marano MA
- Waterson EJ,
- Murray-Lyon IM
- Peterson LS
- Kaplan GA,
- Camacho T
- ↵Ware JE, Snow KK, Kosinski M, Gandek B. SF-36 Health Survey. Manual and Interpretation Guide. Boston, MA: Nimrod Press; 1993
- ↵Short P. Medicaid's Role in Insuring Low-Income Women. Washington, DC: RAND; 1996
- Frankowski BL,
- Weaver SO,
- Secker-Walker RH
- Copyright © 1999 American Academy of Pediatrics