Poverty, insurance, and well-baby care among Mainland Puerto Rican children

Poverty, insurance, and well-baby care among Mainland Puerto Rican children

Gorman, Bridget K

ABSTRACT: Using data from the Puerto Rican Maternal and Infant Health Study, we investigate the implications of family income and insurance status for well-baby care among mainland Puerto Ricans. Given the socioeconomic disadvantage of Puerto Ricans, it is critical to understand the extent to which low income and lack of health insurance create barriers to well-baby care and result in low utilization. The analysis shows that the income-to-needs ratio is related to barriers to well-baby care, and a key intervening factor is insurance status. The odds of reporting any barriers to care are lowest among those with both adequate income and private health insurance. Access to insurance is also vital in achieving adequate well-baby care. Uninsured children receive inadequate care more often than children with public or private insurance, especially when their income is also low. Children with public insurance are as likely as children with private insurance to receive an adequate number of well-baby visits, despite the fact that their mothers report more barriers to care.

INTRODUCTION

The benefits of preventative health care for young children are widely acknowledged. Preventative health care provides the opportunity to monitor children’s health, to immunize children, and to diagnose various acute and chronic health conditions (American Academy of Pediatrics, 1995; Children’s Defense Fund, 2000; Perrin, Guyer, and Lawrence, 1992). Two of the most important determinants of access to preventative health care are income and insurance status (Brooks-Gunn, Duncan, and Maritato, 1997; Monheit and Cunningham, 1992; St. Peter, Newacheck, and Halfon, 1992). Children in low-income families underutilize preventative health services, and their low utilization is at least partially attributable to inadequate insurance coverage. Nonetheless, some research shows that poor children who are covered by public insurance (i.e., Medicaid) have access to preventative health care that is equivalent to that of nonpoor children (St. Peter, Newacheck, and Halfon, 1992; Wolfe, 1995).

A substantial number of children in the United States live in poverty and/or lack sufficient insurance coverage (Children’s Defense Fund, 2000). In 1998, approximately 13.5 million U.S. children– nearly one in five-lived in poverty. Among those, 26 percent lived in extreme poverty (i.e., below half of the federal poverty line). Roughly 21 million children are enrolled in Medicaid, but approximately 11.9 million children are not covered by any form of health insurance (Children’s Defense Fund, 2000). This situation is particularly distressing, given that poor children are substantially more likely than nonpoor children to suffer from health problems, and thus have the greatest need for access to regular, high-quality health care.

The implications of poverty for preventative health care and health outcomes are especially consequential for the growing population of Latino children. Poverty rates for Latino families are twice as high as for non-Latino white families (Oropesa and Landale, 1997)and while Latinos comprise only 11 percent of the population, they represent nearly one-quarter of uninsured Americans (del Pinal and Singer, 1997). Among Latinos, Puerto Ricans have the highest poverty rate and the worst health outcomes for children (del Pinal and Singer, 1997; Landale and Lichter, 1997; Mendoza, et al., 1991). Thus, it is critical to understand the relationships between poverty, insurance coverage, and health care among Puerto Rican children.

In this paper, we address several issues related to the socioeconomic determinants of well-baby care among Puerto Rican children. Data from the Puerto Rican Maternal and Infant Health Study (PRMIHS) are used to examine the implications of poverty for barriers to well– baby care and the adequacy of the care received. The role of insurance (particularly government insurance, or Medicaid) in minimizing the potentially detrimental effects of poverty on well-baby care is also assessed.

INCOME AND PREVENTATIVE HEALTH CARE FOR CHILDREN

Why are preventative health services important for young children? First, children are a vulnerable population. Their health and well-being depends upon their parents’ ability to provide a healthy environment and appropriate medical care. Moreover, children are especially susceptible to the social and environmental risks for poor health that accompany poverty. “Compared with adults, children are more susceptible to toxins and other environmental hazards. For their body size and weight, children drink more water, eat more food, and breathe more air than adults. Children’s immune systems are less developed and less able to handle toxins, and the development of their organs can be disrupted by toxic exposures.” (Children’s Defense Fund, 2000: 40). As a result, children’s need for regular, high-quality health care is great.

The promise of preventative health care lies in its potential to protect children against health and developmental problems. Preventative health care increases the likelihood of timely diagnosis and treatment of acute and chronic health conditions (American Academy of Pediatrics, 1995; Perrin, Guyer, and Lawrence, 1992). For instance, while middle-ear infections can be effectively treated with proper medical care, late or no treatment can lead to persistent problems. The effective management of chronic health conditions, in turn, requires regular monitoring by health care professionals. As an example, the medical management of asthma may reduce the risk of life– threatening attacks that require expensive trips to the emergency room (Brown et al., 1999). Moreover, morbidity and mortality risks in adulthood are associated with illness and poverty in childhood (e.g., Blackwell, Hawyard, and Crimmins, 2001; Wadsworth, 1997). Thus, preventative health services for children not only have the potential to improve health in childhood, but they also may enable children to live longer and healthier lives.

In addition to its role in the diagnosis and treatment of health problems, preventative health care is important for monitoring the developmental progress of children as they age (Perrin, Guyer, and Lawrence, 1992). Among children, poor health may be manifested in failure to achieve developmental milestones at the expected pace. Children who receive preventative care are likely to have a regular health care provider, which improves continuity of care and ensures the availability of baseline measures of health and development before the onset of illness or other conditions (Perrin, Guyer, and Lawrence, 1992; Wagner, Herdman, and Alberts, 1989).

THE ROLES OF POVERTY AND INSURANCE

Approximately 11 percent of families in the United States live below the poverty line (del Pinal and Singer, 1997). However, the rate of poverty is substantially higher for some racial and ethnic groups. About 36 percent of Puerto Rican families lived in poverty in 1995-a percentage six times higher than that for non-Latino whites (6 percent), and substantially higher than that for blacks (26 percent). Puerto Rican families also have the highest poverty rate among all Latino groups: 28 percent of Mexican-American and 16 percent of Cuban-American families lived in poverty in 1995. Among Puerto Ricans, the poverty rate is especially high for some family types. Fully 64 percent of Puerto Ricans living in female-headed families are poor (del Pinal and Singer, 1997).

The extremely high rate of poverty among Puerto Ricans is troublesome, given that poor children underutilize health services (Cunningham and Hahn, 1994; St. Peter, Newacheck, and Halfon, 1992; Wood et al., 1990). As previously noted, one important factor in the relative low health care utilization of the poor is a lack of insurance coverage (Cunningham and Hahn, 1994; Perrin, Guyer, and Lawrence, 1992). Unemployment and employment in “bad” jobs without medical benefits severely limit parents’ ability to secure private medical insurance. Further, those employed in jobs with poor pay and few benefits often fall into the “near poor” category, and thus fail to qualify for government insurance programs. Children who lack health insurance-public or private-are less likely to have a regular source of care and to receive preventative health care visits. They are also less likely to be in good or excellent health (Gavin and Bencio, 1996; Monheit and Cunningham, 1992; Stoddard, St. Peter, and Newacheck, 1992; Trevino et al., 1991).

Government insurance programs such as Medicaid are intended to serve as a safety net for low-income children. All state Medicaid programs are required to provide comprehensive preventative health care services for children, in addition to treatment for both acute and chronic health conditions (Devaney, Ellwood, and Love, 1997). However, a substantial number of poor children who qualify for Medicaid are not enrolled (one-third of uninsured children in 1994) due to a lack of parental awareness about which children are eligible, difficult application procedures, and the continuing stigma of Medicaid as a welfare program (Devaney, Ellwood, and Love, 1997). In addition, some parents see no immediate need to enroll their children) in Medicaid, a fact which significantly diminishes the preventative role of the program.

In general, the success of the Medicaid program is mixed. While children on Medicaid are roughly equivalent to children with private insurance in having a usual source of care and accessing preventative health services, children on Medicaid are disadvantaged with respect to the location of care. They are less likely to receive care in a private doctor’s office, and are more likely to receive care in hospital clinics, neighborhood clinics, and emergency rooms. This situation decreases continuity of care, which makes it more difficult to ensure that children receive preventative services (e.g., immunizations) and that their developmental progress is monitored (Devaney, Ellwood, and Love, 1997; Gavin and Bencio, 1996; Gavin et al., 1997; Love, 1997; Monheit and Cunningham, 1992; St. Peter, Newacheck, and Halfon, 1992).

PUERTO RICAN CHILDREN

Until recently, studies of health care utilization among Latinos typically combined the various Latino subgroups into a single pan-ethnic category. However, researchers increasingly recognize the heterogeneity of the Latino population. Although Puerto Ricans, Mexicans, Cubans, and other Latinos share many commonalities, the groups differ greatly in significant respects, including educational attainment, income, and family structure (Bean and Tienda, 1987). Thus, scholars concerned with the Latino population argue that processes and outcomes should be examined separately by national-origin group. Findings from research on one group are not necessarily generalizable to another.

Puerto Ricans are an important Latino subgroup to study because of their size and their socioeconomic disadvantage. Puerto Ricans are the second-largest Latino group, with a population of approximately 3.1 million in the United States in 1996 (del Pinal and Singer, 1997). In addition, due to high fertility rates and patterns of migration, children comprise a large part of the mainland Puerto Rican population. In 1996, 24 percent of non-Latino whites were under the age of 18, compared to 38 percent of mainland Puerto Ricans (del Pinal and Singer, 1997).

Because Puerto Rican children are more economically disadvantaged than other Latino children, they have an especially high risk of health problems in childhood and later life. At the same time, Puerto Ricans have the highest health insurance coverage of Latinos, due in part to their U.S. citizenship status.1 According to 1989 CPS data, 20 percent of Puerto Ricans were uninsured, compared to over 33 percent of MexicanAmericans and 25 percent of Cuban Americans. Comparable figures for nonLatino blacks and non-Latino whites, respectively, are 20 percent and 10 percent (Trevino et al., 1991).

Despite their advantage with respect to insurance coverage, Puerto Ricans face several financial and nonfinancial barriers to preventative health care. Out-of-pocket costs associated with health care (e.g., child care, transportation, time from work) can be a significant burden to those who are poor. In addition, health care utilization may be constrained by a lack of knowledge about the medical system, especially among recent migrants to the United States. Language barriers and a scarcity of Latino physicians may also discourage parents from obtaining routine preventative health care checkups for children (Council on Scientific Affairs, 1991; Vega and Amaro, 1994).

These various factors may contribute to the relatively high risks of health problems faced by Puerto Rican children. For example, Puerto Ricans have higher rates of low birth weight and infant mortality than other Latino groups and non-Latino whites (e.g., Landale, Oropesa, and Gorman, 1999). While less information is available on Puerto Rican children’s health status after the first year of life, some evidence indicates that they experience relatively poor cognitive development in early childhood and have elevated rates of serious and chronic illnesses, such as asthma and anemia (McCarton, Brooks-Gunn, and Tonascia, 1994; Mendoza, 1994; Mendoza and Dixon, 1999).

DATA AND METHODS

This analysis utilizes data from the Puerto Rican Maternal and Infant Health Study (PRMIHS). The purpose of this study was to assess maternal and infant health outcomes among Puerto Ricans in the United States and Puerto Rico. Birth and death certificates for 1994 and 1995 from six states in the United States (Connecticut, Florida, Massachusetts, New Jersey, New York, and Pennsylvania) as well as the Commonwealth of Puerto Rico provided the sampling frame for this survey.2 These areas were selected because of the high number of births to Puerto Rican women each year. In 1994 and 1995, 72 percent of all births to Puerto Rican women on the mainland occurred in the included states. Infants in the United States were eligible for inclusion in the birth sample if the Hispanic ethnicity of the mother was designated as Puerto Rican on the birth certificate. Infants who died before their first birthday were eligible for inclusion in the U.S. death sample if they were identified as Puerto Rican on the death certificate or if their mother was classified as Puerto Rican on the birth certificate. To allow for analysis of the determinants of low birth weight, infants weighing less than 2,500 grams at birth were oversampled. In addition, the full population of infant deaths was drawn into the study to ensure that there were enough infant death cases for analyses of the determinants of infant mortality.

In-person interviews were conducted with 2,763 mothers of infants sampled from the birth (1,946) and death (817) records. The interviews took place between August, 1995, and September, 1997, when nearly all of the children were between the ages of one and three. The questionnaire was available in both English and Spanish, and all study interviewers were bilingual.

When weights are applied, the birth sample is representative of 1994-1995 births to Puerto Rican women residing in the study area. The present analysis is restricted to the 649 mainland children from the birth sample who were under 24 months of age at interview and whose mothers reported ever living with their child.3,4 The reason for the age restriction is discussed below.

Cases with missing values are not excluded from the analysis to avoid incorrect inferences from the rejection of cases that are not missing completely–at– random. Instead, Bayesian procedures for the multiple imputation of missing data (see Schafer, 1997) were employed using the NORM software package (http://stat. psu.edu/~jls/misoftwa.html). Five imputations were made to generate values for missing data. Each of the five datasets was then analyzed with SUDAAN to generate correct parameter estimates and standard errors, given the complex sampling design of the PRMIHS (see Shah, Barnwell, and Bieler, 1997).5 The results were then combined to yield estimates, standard errors, and test statistics that reflect the uncertainty about missing data (Schafer, 1997). Additionally, all analyses are weighted using the final PRMIHS birth sample weight.

MEASURES

Well-Baby Care Barriers and Adequacy: Two measures are utilized to examine access to and adequacy of well– baby care. We first examine maternal reports of the barriers encountered while attempting to obtain well-baby care for the sampled infant. Specifically, women were asked if any of the following made it difficult to get routine well-baby care: (a) problems getting enough money or insurance to pay for visits; (b) problems getting appointments; (c) problems with transportation to the clinic or doctor’s office; and (d) problems with child care. A summary measure was created where 1 = any barriers to well-baby care, and 0 = no barriers to well-baby care.6

The second dependent variable indicates whether or not children received an inadequate number of well-baby visits. Inadequate in this analysis is defined according to the American Academy of Pediatrics (AAP) recommended number of well-child visits during infancy and early childhood (American Academy of Pediatrics, 1995).7 Specifically, the AAP provides a schedule of the recommended preventative health care visits for children, by age. For example, by six months of age children should have had four visits; by 12 months, six visits; by two years, nine visits. Our measure of inadequacy is constructed by comparing the child’s age at interview with responses to the following question, which was asked after women were questioned about whether or not their child was ever taken for well-baby care when he/she was not sick or injured: “How many times has your baby gone to the doctor for well– baby care? 1-2 times, 3-4 times, 5-6 times, or 7 or more times?” If children of a given age had received the recommended number of well-baby visits they received a score of 0, which indicates adequate well-baby care. If they had not achieved the number of visits recommended for their age they were assigned a score of 1, which indicates inadequate well-baby care. While past research has tended to use imprecise measures of preventative health care (e.g., “usual source of care,” “number of doctor visits in last year”), we improve upon these measures by restricting our focus to preventative care and addressing both the number and timing of visits.

However, it must be noted that our measure of inadequate well-baby care is imprecise due to the manner in which the question was asked in the PRMIHS. First, while the PRMIHS asked about the total number of well-baby visits the child had obtained by interview, the exact timing of these visits is unknown. We assume that if a child received the recommended number of visits for his/her current age, the care received was adequate. Second, the categorical nature of the possible responses is not ideal. The AAP recommends that children have six well-baby visits by 12 months of age; seven visits by 15 months; eight visits by 18 months; and nine visits by 24 months of age. While the age distribution of children in the PRMIHS allows a clear categorization of adequate or inadequate care for children who received 1-2 or 3-4 visits, this is not the case for certain categories of children who had 5-6 or 7+ visits. We assume that children who were less than 15 months of age and had 5-6 visits received adequate care. Because a substantial number of children in the 7+ category probably had more than 7 visits, children greater than 15 months and less than 24 months of age who had 7+ visits were also assigned a score of adequate. Children 24 months of age and older were removed from the analysis because they could not be coded with reasonable certainty.

Income and Insurance: Given the emphasis in prior studies on the role of poverty in access to health care and the health status of children, we construct a measure of the ratio of income to needs. The PRMIHS asked women to select a category representing their total combined yearly household income around the time the baby was born. Responses were recoded to the category midpoints. Next, an income-to-needs ratio was created by dividing the total household income by census-defined poverty income thresholds (which vary by household composition and year). The result is a continuous measure that represents each respondent’s household income as a proportion of the income level that defines the poverty line. For descriptive purposes, Table I categorizes respondents as in deep poverty (

Women also were asked to indicate how their baby’s medical care was paid for: (a) Medicaid; (b) other government assistance; (c) private health insurance; (d) own cash income or savings; (e) husband’s or partner’s cash income or savings; and (f) parent’s or relative’s cash income or savings. Respondents were to select multiple sources if more than one source was used. Given our interest in the distinction between private, public, and no insurance for accessing the health care system, responses were coded into the following categories: private insurance, no government assistance (reference category); any government assistance; and no private or government insurance.

Socioeconomic Status: Since poverty and insurance status are related to other aspects of socioeconomic status, three additional variables are controlled. First, maternal education is included as a continuous measure of years of schooling. Second, because it may be more difficult for employed women to take their child for well-baby visits, we include a dummy variable that contrasts mothers who were not employed during the first year of the child’s life (1) and mothers who were employed (0). Lastly, we include a dummy variable measuring WIC program participation during the first year after birth (1 = enrolled in WIC, 0 = not enrolled in WIC). We include WIC participation because this program provides medical care and nutritional support to children. All children under 5 years of age who live in households below 185 percent of the poverty threshold are eligible for the program.

Maternal Demographic Characteristics: Several additional characteristics of the mother at the birth of her child are included. First, we include a dummy variable for nativity status (1 = U.S. born, 0 = Puerto Rico born). Our rationale for including this measure is twofold. First, women born in Puerto Rico may be less knowledgeable about the medical system in the United States, resulting in more barriers and less adequate well-baby care for children. Second, differences in health have been noted previously among children born to foreign- and native-born Latina women (Guendelman et al., 1990; Landale, Oropesa, and Gorman, 1999; Markides and Coreil, 1986).

In addition, a lack of proficiency in English has been noted as a barrier to health care access among the Latino population (Council on Scientific Affairs, 1991). We measure self-reported English proficiency as a dummy variable where 1 = proficient in English, and 0 = not proficient in English. Respondents are given a value of 1 if the interview was conducted mostly or completely in English. If the interview was conducted in Spanish, respondents were asked: “How well do you speak English? Very well, well, not well, or not at all?” Respondent’s who answered “very well” or “well” are also assigned a value of 1, while those who answered “not well” or “not at all” are assigned a value of 0.

In addition, we include continuous measures of age at birth and the number of children previously born to the mother. A categorical measure describes union status at birth: legally married (reference category), cohabiting, and not married or living with the child’s father.

Child Characteristics: We include measures of several child characteristics at birth and interview, since the characteristics of a particular child are likely to influence parents’ motivation for seeking regular well-baby visits, as well as health outcomes in early childhood. Along with gender, we include a continuous measure of the child’s age (in months) at interview. We also control for health status at birth because children with poor perinatal health may need more medical care than other children, given their increased risk of health and developmental problems (Gorman, 2000; Hack, Klein, and Taylor, 1995). Our measure of health at birth combines information on birth weight and gestation. We subdivide low-birthweight children (those weighing less than 2,500 grams, or 5.5 lbs.) into two categories: preterm (

Stress and Social Support: Our final set of measures addresses the influence of maternal stress and social support on barriers to and adequacy of well-baby care. Women who experience high levels of stress may be less likely to keep up with well-baby visits, while support from friends and family may enable women to make the recommended number of visits. Direct and indirect measures are available in the PRMIHS. With regard to stress, women were asked to report the level of stress they felt during the first year after the birth of their child, where 0 = no stress and 4 = a very great deal of stress. Women also were asked to describe the amount of emotional support or advice they received during the first year, where 1 = a lot less than needed, and 5 = a lot more than needed. Further, each woman identified all members of her household at the time the child was born. From this roster we construct a dummy measure, where 1 = any extended family members living in the household at birth, and 0 = no extended family in household.

RESULTS

DESCRIPTIVE STATISTICS

The first column of Table I contains weighted means and percentages for the two outcome measures (any barriers to well-baby care; inadequate care) and each predictor. Columns 2 through 5 provide equivalent information by the income–to– needs ratio at the time of the child’s birth. Although weighted percentages for categories of the income-to-needs ratio are not shown in the table, it is important to note that 54 percent of the children in our sample lived below the poverty line at the time of the survey (29 percent in deep poverty and 24 percent in poverty). An additional 16 percent were near poor, and 30 percent were not poor.

Table 1 indicates that the relationships between the income-to-needs ratio and the outcome measures are not significant, despite percentages that suggest that: 1) women who are not impoverished are the least likely to report barriers and 2) inadequate well-baby care increases with the income-to-needs ratio. As will be shown below, the non-significant result for barriers stems from the categorization of the income-to-needs ratio.

While the majority of children receive Medicaid or some other type of governmental assistance (64 percent), coverage type varies significantly by income–to– needs level. Compared to children living in deep poverty, those who are near poor or not poor are significantly more likely to have private insurance coverage for medical expenses, and less likely to receive any type of governmental assistance. Indeed, 89 percent of children living in deep poverty have part or all of their medical expenses paid for by governmental insurance, with only 3 percent lacking both private and public coverage. At the same time, the results of Table 1 reveal the unmet need of children in households with an income-to-needs level just above the federal poverty line. The percentage of children lacking health insurance is over six times higher for children living in near-poor households (19 percent) as for children living in deep poverty (3 percent). The former figure is also substantially higher than those for poor and not poor children (10 percent and 6 percent, respectively).

As expected, several other economic and demographic characteristics of the mother show significant variation by the ratio of income to needs. Compared to children living in deep poverty, those who are near poor or not poor have more welleducated mothers who were less likely to be out of the labor market during the first year after the birth and less likely to be enrolled in the WIC program. In addition, children in higher-income homes have significantly fewer siblings, are more likely to be born to married parents, and are more likely to have a mother who is proficient in English.

ANY BARRIERS TO ROUTINE WELL-BABY CARE

Table 2 provides results from logistic regression models of any barriers to well– baby care. The first column presents odds ratios that summarize the bivariate relationships between barriers and our key independent variables, the income–to– needs ratio and insurance source. The parameter estimates indicate that there is a significant association between the income-to-needs ratio and whether or not any barriers to well-baby care are reported by the mother. As expected, as the income-to-needs ratio of the household increases, the odds of women experiencing barriers to well-baby care decline. Insurance source is also associated with barriers to well-baby care. Relative to mothers of children with private insurance, the odds of experiencing any barners to care were over 4 times as high for mothers whose children had governmental insurance, and over 9 times as high for mothers whose children had no insurance coverage.

The remaining models in Table 2 demonstrate the extent to which the relationship between the income-to-needs ratio and barriers to care can be explained by the other measured factors. Model 1 includes variables measuring socioeconomic status, maternal demographic characteristics, and child characteristics. Inclusion of these variables strengthens the relationship between the income-to-needs ratio and barriers slightly. Model 2 adds insurance source. Once insurance source is included in the model, the relationship between the income-to-needs ratio and barriers is non-significant. This finding is important because it indicates that income influences barriers to well-baby care via a family’s ability to acquire medical insurance for their child.

Model 3 is the full additive model, which differs from Model 2 in that controls for maternal stress and social support are added. Although inclusion of these variables does not diminish the association between insurance status and barriers, maternal stress is strongly related to barriers to well-baby care. As stress levels increase, so do the odds that women will report at least one circumstance that made it difficult to get routine well-baby care. Apart from insurance status and stress, only two predictors are significant in Model 3. The odds of reporting at least one barrier increase with the number of prior births. Furthermore, women who were not employed during the first year after birth are significantly less likely to report that they experienced barriers than women who were employed during that time.

Using Model 3 as the baseline model, we also explored whether the influence of income on barriers to care varies by other key variables of interest. We tested interactions between the income-to-needs ratio and insurance status, WIC participation, maternal socioeconomic and demographic characteristics, stress, social support, and birth weight. Among the terms tested, one is significant and improves the model fit: the income-to-needs ratio and insurance source.9

Model 4 includes this interaction. For those with private insurance, the income– to-needs ratio has a marginally significant negative relationship with barriers to care (odds ratio = .51). However, for those with government insurance, the relationship between the income-to-needs ratio and barriers is weakly positive (odds ratio = .51*2.42 = 1.23) and non-significant; for those with no insurance, the relationship is negative (odds ratio = .51*1.59 = .81) and non-significant.10 Thus, those with both private health insurance and adequate income report the fewest barriers to well-baby care. The lack of a relationship between the income-to-needs ratio and barriers for those with government insurance and those with no insurance may result from the compressed income distribution for these groups. Very few individuals with government insurance or no insurance are at the upper end of the income-to-needs distribution, and thus they may have insufficient income to overcome the barriers to well-baby care that they face.

ADEQUACY OF WELL-BABY CARE RECEIVED

Overall, the results in Table 2 show that the income-to-needs ratio is related to barriers to well-baby care, and that a key intervening factor is insurance status. The next question we explore is whether income and insurance are associated with the adequacy of care received. Additionally, we consider whether barriers to routine well-baby care increase the risk of receiving an inadequate number of visits. Table 3 follows the same model-building sequence as Table 2, except that we include an additional additive model (Model 4), which adds the measure of whether any barriers to well-baby care were reported.

The income-to-needs ratio is unrelated to receiving inadequate care in the bivariate and multivariate models. Although this is unexpected, given the emphasis in prior studies on the role of income in children’s access to regular health care, it appears to be related to the fact that Puerto Rican children in both the upper and lower ends of the income distribution are adequately insured. Across all of the models, there is no difference between children with public and private health insurance in the adequacy of care received. Medicaid has apparently been successful in allowing low-income Puerto Rican children to maintain a regular schedule of well-child visits, at least in the first two years of life. However, children with no insurance are much more likely than children with private insurance to receive an inadequate number of well-baby visits, as is indicated by odds ratios greater than 3 in Models 2 through 4.

Model 4 is the full additive model. In addition to insurance status, maternal nativity is a significant predictor in the model. The odds of receiving inadequate well-baby care are less than half as high for children of native-born mothers as for children of island-born mothers. While no difference in reported barriers was found in Table 2, native-born women may possess qualities, such as increased knowledge of the health care system, that significantly reduce the odds of inadequate care for their children. English proficiency is marginally significant (p

A set of interaction terms that is identical to that outlined in the models of any barriers was tested in the analysis of inadequate care (using Model 4 as baseline), except that the interaction between barriers and the income-to-needs ratio was added. Among the interactions, two significant terms were found: the income–to– needs ratio and insurance source, and the income-to-needs ratio and birth weight.11 Model 5 reveals that, while income is not important to well-baby care for those with private or public insurance, it is very important for children who lack any type of health insurance coverage (odds ratio = 1.06*.14 = .15). Children with no insurance face a very high risk of receiving an inadequate number of well-baby visits for their age at the lowest income–to– needs level. As income levels increase, the odds of inadequate care are substantially reduced for those with no insurance.

The interaction between the income– to-needs ratio and birth weight reveals that income is more strongly related to well-baby care for infants who were lowbirth-weight and preterm than for fullterm infants. For the former group, the odds ratio for the income-to-needs ratio is .60 (1.06*.57), while for the latter it is not significantly different from one (1.06). This finding may reflect the relatively high expenses associated with raising a low-birth-weight and preterm child (absolutely, as well as indirectly via other mechanisms such as decreased time at work).

SUMMARY AND CONCLUSIONS

The goal of this investigation was to examine the implications of income and health insurance for barriers to well-baby care and adequacy of well-baby care among mainland Puerto Ricans. Puerto Ricans are of special interest because they are the second largest and most disadvantaged Latino subgroup in the United States. Puerto Rican children have an elevated risk of inadequate well-baby care because of both financial (e.g., poverty, lack of insurance) and nonfinancial (e.g., migrant status, language) problems.

How important is adequate income for preventative health care among Puerto Rican children? While past research led us to expect a strong, direct relationship, our findings revealed a process that is somewhat more complex. Additive models show that the income-to-needs ratio is related to reported barriers to care, and a key intervening factor is insurance status. However, multiplicative models reveal that the relationship between income and barriers to well-baby care varies by insurance status. There is no relationship for those with government insurance or no insurance. However, the income-to-needs ratio is related negatively to barriers among those with private health insurance. Thus, those with both private health insurance and adequate income report the fewest barriers to well-baby care. Having private insurance increases parents’ flexibility in selecting a provider, and monetary resources can be used to find solutions to problems with transportation, child care, and other considerations.

There is also an interaction between income and insurance status in the models of the adequacy of the care received. In contrast to the findings for barriers, income is most important for obtaining adequate well-baby care among those with no health insurance-that is, those who typically must pay for care with their own resources. It comes as no surprise that among those with no health insurance, those with greater income are more likely to obtain adequate care.

A key research and policy question is whether government insurance (primarily Medicaid in this sample) is successful in providing poor children access to preventative health care. Our findings show that Puerto Rican mothers whose children have government insurance report more barriers to well-baby care than Puerto Rican mothers whose children have private insurance. However, children with public insurance are as likely as those with private insurance to receive an adequate number of well-baby visits. It is primarily those children who are uninsured who have an elevated risk of inadequate well– baby care. As illustrated in Table 1, children in near poor families are the most likely to lack health insurance, and their mothers are the most likely to report experiencing barriers to routine well-baby care. Overall, this suggests that attention should be devoted to the issue of how children in families with incomes that are slightly above the poverty line (the “near– poor”) can be provided with greater access to adequate medical care.

In contrast to most prior studies, we found few significant relationships between the socioeconomic and demographic characteristics of mothers and the adequacy of their children’s health care. Several factors may contribute to the difference between our findings and those of other studies. First, the children in our study were very young at the time of the interview. Young children may have greater access to publicly-funded health care than older children because of the emphasis in the United States on ensuring access to preventative health care during the first years of life. Moreover, Puerto Ricans have greater access to many publicly-funded programs than other Latino groups, given that even those who were born in Puerto Rico are U.S. citizens. The young age of the children in our sample, coupled with the widespread availability of publicly-funded health care for young children, may have attenuated the influence of maternal characteristics on the health care received by the children in our study.

In addition, our measure of the adequacy of care differs from those used in some prior studies. As previously mentioned, past research has tended to use measures such as “usual source of care,” “ever saw a physician,” and “number of doctor visits in last year,” often without isolating preventative visits from those in response to illness or injury (e.g., Brown et al., 1999; Wood et al., 1990; Trevino et al., 1991). Alternatively, well-baby visits are examined, but the timing of care is ignored (e.g., Cunningham and Hahn, 1994).12 We improve upon these measures by restricting our focus to preventative care and addressing both the number and timing of visits.13

Overall, our research indicates that the public safety net protects Puerto Rican children against inadequate well-baby care. A high proportion of impoverished Puerto Rican children are covered by Medicaid, and are therefore able to access preventative health care. These findings reinforce the importance of data collection efforts and analyses that focus on specific Latino groups. Puerto Ricans are unique among Latinos in their unrestricted access to public benefits. Their health care utilization patterns-especially among the poor-reflect their distinct position due to Puerto Rico’s commonwealth status.

Comparable research on the number and timing of preventative health care visits for children is needed for other Latino groups. The Latino population in the United States is projected to grow considerably in the coming decades. Thus, understanding how Puerto Rican, Mexican, Cuban, and other children from Central and South America access the health care system in the face of poverty will become increasingly important to efforts to improve the health and well-being of all children in the United States.

ACKNOWLEDGEMENTS

The research reported in this paper was supported by the National Institute of Child Health and Human Development, the Maternal and Child Health Bureau, and the Centers for Disease Control. Support services were provided by the Population Research Institute, The Pennsylvania State University, and the Carolina Population Center, The University of North Carolina at Chapel Hill. This manuscript was written while the first author was funded by a National Institute of Health, National Research Service Award (T32 HD07168-22) from the National Institute of Child and Health and Human Development.

Direct all correspondence to Bridget K. Gorman, Carolina Population Center, University Square CB# 8120, 123 West Franklin St., Chapel Hill, NC 27516-2524. Email: bkgorman@unc.edu.

1Because Puerto Rico is a commonwealth of the United States, both island and mainland Puerto Ricans are U.S. citizens. Thus, migrants from Puerto Rico to the U.S. mainland are eligible for public benefits.

2New York state is divided into two separate vital statistics reporting areas: New York City and the remainder of the state. While permission to conduct the

study was received from New York City, it was not obtained from the state of New York. New York cases are therefore restricted to births and deaths occurring in New York City.

3The decision to limit the current analysis to mainland residents was made because of the substantial differences in the structure of the health care system in the United States and Puerto Rico. While a comparative

analysis would be useful, we think it would be most profitably undertaken as a second step after a clear understanding of the relationships between income, insurance and well-baby care in each setting is established.

4The birth sample from the PRMIHS contained 1,946 cases (including children who were initially selected into the birth sample, but who later died). We limited the birth sample to the 1,256 mainland residents. Elimination of 83 cases in which the mother reported that she never lived with the child (a necessary restriction, as the questions pertaining to well-baby care were only asked of women who reported ever living with their child) left 1,173 cases. Finally, limiting the analysis to children who were under 2 years of age at interview removed an additional 524 children from the analysis, resulting in a final sample size of 649 cases.

5Failure to account for the complex sampling de

sign in the statistical analysis would result in underestimated standard errors of point estimates, leading to an increase in false-positive tests of treatment effects (Brogan, 1998).

6Approximately 3.5 percent of women reported having problems getting enough money or insurance to pay for a visit; 5.8 percent mentioned problems with transportation to the clinic or doctor’s office; 3.3 percent mentioned problems getting appointments; and 3.5 percent mentioned problems with child care. A dichotomous measure of “any barriers” was used in the analysis rather than an additive index of the number of barriers because the additive index was highly skewed.

7This schedule of visits “represents a consensus by

the Committee on Practice and Ambulatory Medicine in consultation with national committees and sections of the American Academy of Pediatrics.” (American Academy of Pediatrics, 1995: 374).

8We chose to keep the income-to-needs ratio continuous only after preliminary analyses in which it was in

eluded in categorical form. Although several different cut-points for “near poor” and “not poor” were explored, the findings were similar for the categorical and continuous measures of the income-to-needs ratio.

9Because we use the NORM software package to analyze multiple imputed data sets (see methods section), the traditional likelihood ratio test for model fit is not available. Instead, we compute model fit by utilizing the multiparameter inference test in NORM, which tests the null hypothesis that a group of coefficients (e.g., a set of interaction terms) are simultaneously zero against the alternative hypothesis that at least one coefficient is not zero. See http://www.stat.psu.edu/-jls/ misoftwa.html for more information.

10We ran Model 3 separately for children with private, governmental, and no insurance. The income-to

needs ratio was significantly related to barriers only for children with private insurance; the relationship was non-significant for children with governmental or no insurance.

11We could not obtain a reliable estimate of the in

teraction between barriers to well-baby care and insurance source due to problems of small cell size. See footnote #9 for an explanation of model-fit testing procedures.

12An exception is a study by St. Peter, Newacheck, and Halfon (1992). Using data from the 1988 National Health Interview Study (NHIS) on Child Health, they construct a measure of whether or not care was received in a timely manner, relative to the schedule set by the American Academy of Pediatrics. However, the NHIS provided information on the timing of only the last doctor visit, so their measure of timely preventative health care is an approximation. While they did document that poorer children were less likely to receive preventative health care at the appropriate intervals, they did not stratify their analysis by the age of the child at interview or Hispanic ethnic group.

13In addition, unlike many studies with complex sampling designs, we utilize a statistical program (SUDAAN) that appropriately adjusts the standard errors reported for our models. It is noteworthy that when our models of inadequate care are run using SAS in

REFERENCES

AMERICAN ACADEMY OF PEDIATRICS. 1995. Recommendations for preventive pediatrics health care. Pediatrics 96 (2):374.

BEAN, FRANK, and MARTA TIENDA. 1987. The hispanic population of the United States. New York: Russell Sage.

BLACKWELL, DEBRA L., MARK D. HAYWARD, and EILEEN M. CRiMMINS. 2001. Vol. 52:1269-84. Does childhood health affect chronic morbidity in later life? Social Science and Medicine.

BROGAN, DONNA J. 1998. Pitfalls of using standard statistical software packages for sample survey data. In Encyclopedia of biostatistics, edited by Peter Armitage and Theodore Colton. Publisher: Chichester; New York: J. Wiley, Vol. 5, pp. 4167 174.

BROOKS-GUNN, JEANNE, GREG J. DUNCAN, and NANCY MARITATO. 1997. Poor families, poor outcomes: the well-being of children and youth. Chapter 1 in Consequences of growing up poor, edited by Greg J. Duncan and Jeanne BrooksGunn. New York: Russell Sage Foundation.

BROWN, E. RICHARD, ROBERTA WYN, HONGJIAN YU, ABEL VALENZUELA, and LIANE DONG. 1999. Access to health insurance and health care for children in immigrant families. In Children of immigrants: health, adjustment, and public assistance, edited by Donald J. Hernandez. Committee on the Health and Adjustment of Immigrant Children and Families, Board on Children, Youth, and Families. Washington, DC: National Academy Press.

CHILDREN’S DEFENSE FUND. 2000. The state of America’s children. Boston: Beacon Press.

COUNCIL ON SCIENTIFIC AFFAIRS. 1991. Hispanic health in the United States. Journal of the American Medical Association 265 (2):248-252.

CUNNINGHAM, PETER J., and BETH A. HAHN. 1994. The changing American family: implications for children’s health insurance coverage and the use of ambulatory care services. The Future of Children: Critical Health Issues for Children and Youth 4 (3):24-43.

DEL FINAL, JORGE, and AUDREY SINGER. 1997. Generations of diversity: Latinos in the United States. Population Bulletin 52 (3). Washington DC: Population Reference Bureau.

DEVANEY, BARBARA L., MARILYN R. ELLWOOD, and JOHN M. LOVE. 1997. Programs that mitigate the effects of poverty on children. The Future of Children: Children and Poverty 7 (2):88-112.

GAVIN, NORMA I., ELICIA J. HERZ, KATHLEEN E. ADAMS, ANITA J. CHAWLA, MARILYN ELLWOOD, IAN HILL, BETH ZIMMERMAN, and JEFFREY WASSERMAN. 1997. Comparative study of the use of EPSDT and other preventative and curative health care services by children enrolled in Medicaid: final project synthesis report. Research Triangle Institute, Research Triangle Park, NC.

GAVIN, NORMA I., and DEO S. BENCIO. 1996. Impact of Medicaid on children’s health service use and expenditures: 1987 National Medical Expenditure Survey. NTIS: Springfield, VA.

GORMAN, BRIDGET K. 2000. Early child development among mainland and island Puerto Ricans: the roles of birth outcomes, socioeconomic status, and social support. Ph.D. Diss.

GUENDELMAN, SYLVIA, JEFFREY B. GOULD, MARK HUDES, and BRENDA ESKANAZI. 1990. Generational differences in perinatal health among the Mexican American population: findings from HHANES, 1982-1984. American Journal of Public Health 80 (Suppl):61-65.

HACK, MAUREEN, NANCY K. KLEIN, and H. GERRY TAYLOR. 1995. Long-term developmental outcomes of low birth weight infants.The Future of Children: Low Birth Weight 5 (1):176-196.

LANDALE, NANCY S., and DANIEL T. LIGHTER. 1997. Geography and the etiology of poverty among Latino children. Social Science Quarterly 78 (4):874-894.

LANDALE, NANCY S., R. S. OROPESA, and BRIDGET K. GORMAN. 1999. Immigration and infant health: birth outcomes of immigrant and native women. In Children of immigrants: health, adjustment, and public assistance, edited by Donald J. Hernandez. Committee on the Health and Adjustment of Immigrant Children and Families, Board on Children, Youth, and Families. Washington, DC: National Academy Press.

MARKIDES, KYRAKOS S., and JEANNINE COREIL. 1986. The health of Hispanics in the southwestem United States: an epidemiological paradox. Public Health Reports 101:253-265.

MCCARTON, CECELIA M., JEANNE BROOKS-GUNN, and JAMES TONASCIA. 1994. The cognitive, behavioral, and health status of mainland Puerto Rican children in the infant health and development program. Chapter 10 in Puerto Rican women and children: issues in health, growth, and development, edited by Gontran Lamberty and Cynthia Garcia Coll. Plenum Press: New York.

MENDOZA, FERNANDO S. 1994. The health of Latino children in the United States. The Future of Children: Critical Health Issues for Children and Youth 4 (3):43-72.

MENDOZA, FERNANDO S., and LORI BETH DIXON. 1999. The health and nutritional status of immigrant Hispanic children: analyses of the Hispanic health and nutritional examination survey. In Children of immigrants: health, adjustment, and public assistance, edited by Donald J. Hernandez. Committee on the Health and Adjustment of Immigrant Children and Families, Board on Children, Youth, and Families. Washington, DC: National Academy Press.

MENDOZA, FERNANDO S., STEPHANIE J. VENTURE, R. BURCIAGA VALDEZ, RICHARD O. CASTILLO,

LAURA ESCOTO SALDIVAR, KATHERINE BAISDEN, and REYNALDO MARTORELL. 1991. Selected measures of health status for MexicanAmerican, mainland Puerto Rican, and CubanAmerican children. Journal of the American Medical Association 265 (2):227-232.

MONHEIT, ALAN C., and PETER J. CUNNINGHAM. 1992. Children without health insurance. The Future of Children 2 (2):154-170.

OROPESA, R. S., and NANCY S. LANDALE. 1997. Immigrant legacies: ethnicity, generation, and children’s familial and economic lives. Social Science Quarterly 78 (2):399-416.

PERRIN, JAMES, BERNARD GUYER, and JEAN M. LAWRENCE. 1992. Health care services for children and adolescents. The Future of Children: U.S. Health Care for Children 2 (3):58-77.

SCHAFER, JOSEPH L. 1997. Analysis of incomplete multivariate data. London: Chapman and Hall.

SHAH, B. V., B. G. BARNWELL, and G. S. BIELER. 1997. SUDAAN user’s manual, release 7.5. Research Triangle Park, NC: Research Triangle Institute.

STODDARD, JEFFREY J., ROBERT F. ST. PETER, and PAUL W. NEWACHECK. 1994. Health insurance status and ambulatory care for children. New England Journal of Medicine 330:1421-1425.

ST. PETER, ROBERT F., PAUL W. NEWACHECK, and NEAL HALFON. 1992. Access to care for poor children: separate and unequal? Journal of

the American Medical Association 267 (20): 2760-2764.

TREVINO, FERNANDO M., EUGENE MOYER, BURCIAGA VALDEZ, and CHRISTINE A. STROUP-BENHAM. 1991. Health insurance coverage and utilization of health services by Mexican Americans, mainland Puerto Ricans, and Cuban Americans. Journal of the American Medical Association 265 (2):233-237.

VEGA, WILLIAM A., and HORTENSIA AMARO. 1994. Latino outlook: good health, uncertain prognosis. Annual Review of Public Health 15:39-67.

WADSWORTH, M. E. J. 1997. Health inequalities in the life course perspective. Social Science and Medicine 44 (6):859-869.

WAGNER, J., R. HERDMAN, and D. ALBERTs. 1989. Well-child care: how much is enough? Health Affairs 8 (3):147-157.

WOLFE, BARBARA L. 1995. Economic issues of health care. In Escape from poverty: what makes a difference for children?, edited by P. Lindsay Chase-Lansdale and Jeanne BrooksGunn, pp. 170-185. Cambridge: Cambridge University Press.

WOOD, DAVID L., RODNEY A. HAYWARD, CHRISTOPHER R. COREY, HOWARD E. FREEMAN, and MARTIN F. SHAPIRo. 1990. Access to medical care for children and adolescents in the United States. Pediatrics 86 (5):666-673.

Bridget K. Gorman, Nancy S. Landale, and R. S. Oropesa*

Carolina Population Center, University of North Carolina, Chapel Hill; Department of Sociology and Population Research Institute, The Pennsylvania State University; Department of Sociology and Population Research Institute, The Pennsylvania State University

*Direct correspondence to Bridget K. Gorman, Carolina Population Center, University Square CB# 8120, 123 West Franklin St., Chapel Hill, NC 27516-2524.

Email: bkgorman@unc.edu.

Copyright Society for the Study of Social Biology Spring 2001

Provided by ProQuest Information and Learning Company. All rights Reserved