Unintended Childbearing, Maternal Beliefs, and Delay of Prenatal Care

A primary year 2000 national objective in the United States is to increase to 90 percent the rate of early prenatal care (1). In 1995, about 81 percent of pregnant women initiated care during the first trimester, continuing the upward trend in the 1990s (2). Despite this progress, however, significant racial and ethnic group differences persist. The Institute of Medicine organized the numerous correlates of prenatal care use into three broad categories, including sociodemographics, health care system characteristics, and maternal beliefs and knowledge (3). With birth certificates providing a ready source of data, numerous studies have examined sociodemographic variables. These studies consistently showed that low education, unmarried status, young age, minority status, and high parity were associated with delayed care (4,5).

Although less often studied, several health care system and maternal knowledge and belief variables were reported to be associated with poor prenatal care use, including lack of insurance (6,7), noncosmopolitan social networks (8–10), unintended childbearing (11,12), transportation and child care problems (3), and the belief that prenatal care was unimportant or ineffective (13). Multivariate analyses simultaneously considering a range of prenatal care determinants usually find that no insurance is the strongest predictor of delayed care (3).

The Institute of Medicine specified that two of these determinants, unintended childbearing and the belief that care is unimportant, deserved further study. Unintended childbearing is an increasingly important issue for several reasons. First, the number of unintended births has increased during the 1980s and 1990s following a decline in the 1970s (14,15). Second, with delayed marriage and childbearing on the rise, more women are experiencing lengthy periods during which they do not wish to become pregnant (16). Third, as managed care expands, family planning and other programs to prevent unintended pregnancy are threatened. Studies documenting the poor social and health outcomes of unintended childbearing may have an important role in debates concerning essential managed care services. Fourth, as evidence grows showing the benefits of health education and preventive health care before conception, it becomes increasingly important that pregnancies be planned (17).

About 40 percent of infants born in the United States are unintended, of which about one-fourth are unwanted and three-fourths are mistimed (i.e., occur sooner than desired) (12,18,19). Unintended childbearing has been linked to insufficient prenatal care (11,12,20,21), alcohol and tobacco use during pregnancy (11), neonatal mortality (22), family violence (23,24), delayed child development (25), and low educational attainment (26).

Women who believe prenatal care is unimportant or ineffective may delay seeking care or not comply with the prescribed sequence of prenatal visits. At least three studies have linked beliefs about the importance and effectiveness of prenatal care to its use. These studies indicated that women who viewed care negatively were 1.3 to 3.5 times more likely to obtain insufficient care (13,27,28).

The present study examined the effect of unintended childbearing and maternal beliefs on delay of prenatal care, controlling for sociodemographic and health care system variables. It was anticipated that women with an unintended birth and who believed care was unimportant would have significantly higher rates of delayed care.

Methods

Results

From the total sample of 2032 women, 1175 were either “other” race or from households above 200 percent of poverty, leaving 857 subjects. Higher income women were dropped from the analysis because the rate of delayed prenatal care was greater among low-income women. Item-specific nonresponse rates ranged from 0.4 to 6.8 percent. Therefore, of the 857 women, 132 (17 percent) were excluded because of missing data on one or more variables, yielding 725 subjects available for analysis. Those excluded due to item nonresponse were more likely to be minority women, have less than a high school education, be unmarried, and believe care to be unimportant.

About one-half of this low-income subsample delayed initiation of prenatal care past the first trimester (48.4%), of whom 20 percent were teens, 39 percent were unmarried, and 49.4 percent had less than a high school education. About one-half the sample were Hispanic (49.7%), one-third were white (32.1%), and the remainder black. Sixty-seven percent had a previous live birth. About one-half reported having no third party insurance, and one-fourth each private insurance or Medicaid coverage.

Table 1 presents the cross-tabulations between delayed care and the full set of independent variables. Significant differences existed between the delayed care group and the first trimester group on each of the sociodemographic variables except previous live births. Teenagers, unmarried women, women with less than a high school education, and black women delayed prenatal care more frequently. In addition, women with Medicaid and no insurance delayed prenatal care more frequently than women with private insurance, women who believed that care was unimportant delayed care more frequently than women who believed that care was important, and women who reported an unintended birth delayed care more frequently than women who reported an intended birth.

Table 1. . Delayed Prenatal Care by Maternal Characteristics

In the logistic regression analysis, entry of both blocks of predictor variables resulted in a significant improvement in the log likelihood χ² over the constant for the first block (χ²= 83.84 df= 6, p<0.0001), and over the previous 5-variable model for the second 3-variable block (χ²= 31.1, df= 4, p<0.0001). The 8 independent variables were relatively uncorrelated, with phi or Pearson r coefficients ranging from 0.15 to 0.22.

Table 2 presents the results of the multiple logistic regression analysis, with both crude and adjusted odds ratios (OR) displayed, as well as 95% confidence intervals. After controlling for sociodemographics, the three remaining variables significantly predicted delayed care. Women without private insurance or who believed that care was unimportant were more than twice as likely to delay prenatal care (adjusted OR = 2.39 and 2.04, respectively). Women whose infant was unintended were also significantly more likely to have delayed care (adjusted OR = 1.57).

Table 2. . Results of Multiple Logistic Regression Analysis with Delayed Prenatal Care as the Dependent Variable * p<0.05.† p<0.01.‡ p<e1>0.001. § Reference group' White. ∥ Reference group' Private insurance.

Discussion and Conclusions

Relationships among the sociodemographic variables and delayed care were consistent with findings from previous studies. Women without private insurance and with high-risk demographics were more likely to delay care (3,4,6). Most importantly, significant odds of delayed care were found for unintended childbearing and for the belief that prenatal care was unimportant. The coefficient for the belief variable was larger than that for each sociodemographic variable and was about equal to the coefficient for health insurance. The coefficient for unintended childbearing exceeded that for parity and closely approached the size of the coefficients for the remaining covariates. This suggests that efforts to communicate the value of prenatal care and to prevent unintended births are important directions for intervention and policy.

Women who believed that prenatal care was unimportant may have considered it ineffective, had an unsatisfactory experience with prenatal care during a previous pregnancy, or assumed self-care was sufficient. If no troublesome symptoms existed, they may have correctly considered pregnancy to be a normal and healthy process, but incorrectly concluded that this obviates the necessity of preventive care. Lack of knowledge about the diagnostic and educational content of prenatal care also may have led to the belief that it was unimportant. Moreover, in this low-income sample the life stresses and problem circumstances inherent to poverty may have overwhelmed any priority attached to health care.

In studies in Detroit, New York City, and Missouri, beliefs about the importance of prenatal care were found to be associated with prenatal care use. In Detroit, Poland et al reported that beliefs about the importance of prenatal care accounted for 5 percent of the variance in explaining adequacy of care (29). In New York City, Kalmuss and Fennelly observed that women who reported that they “didn't need care if feeling fine” were 1.3 times more likely to delay care (13). Sable et al reported that women who perceived care as not very necessary were 3.5 times more likely to have inadequate care (28). Thus, the current results fell in the middle of the range of effect sizes from earlier studies.

Unintended childbearing may affect prenatal care through several mechanisms. Women with unplanned infants may not have recognized their pregnancy until later in gestation, or may have put off seeking care while deciding whether or not to continue with the pregnancy. In addition, unintendedness may lead to denial of pregnancy or engender ambivalence, which in turn, could result in placing low priority on seeking timely care.

A relationship between unintended childbearing and prenatal care use was reported in several other studies, including six states participating in the Centers for Disease Control and Prevention's Pregnancy Risk Assessment Monitoring System (PRAMS). Similar to this study, odds ratios ranged from 1.4 to 2.2. For example, in Oklahoma, DePersio et al reported that mothers of unintended births were 1.5 to 2 times more likely to delay care (11). Using PRAMS data, Adams et al showed that almost 70 percent of mothers with intended births initiated care in the first trimester, whereas only about 25 percent of mothers with unintended births did so (12). Using the Kessner index as the dependent variable, a Missouri study found unintended births were 2.2 times more likely to involve inadequate prenatal care (28). A Maine study found unintended births significantly related to the number of prenatal care visits when adjusted for length of gestation (21). In Michigan, women with an unintended birth were 1.4 times more likely to have obtained inadequate care (29).

Several limitations of the present work should be noted. First, although the postpartum hospital-based design reduced recall bias and provided a high response rate and a broad sample, it probably missed women who left the hospital early, as well as those unable to read or write. Second, the data are self-reported. Since the operationalization of unintended childbearing in this study differed from some others, it was not feasible to classify unintended births into the more refined categories of unwanted and mistimed. Although it is of value to distinguish these, differences in odds ratios on delayed care for unwanted and mistimed births, although somewhat larger for unwanted births, in general have been relatively minor (12). Third, data were collected after birth, and therefore causality should not be inferred. Postpartum views may differ from prenatal views. Even if some women in the sample considered their pregnancy unintended at the time it was recognized, they might be less likely to report this postpartum. Similarly, views about prenatal care may become more favorable after the birth of a healthy infant. Fourth, given higher rates of delayed care among low-income populations, the analyses focused on women below 200 percent of the poverty level, limiting external validity. When analyses were conducted with the full sample regardless of income, however, similar results were obtained; adjusted odds ratios and confidence intervals for unintendedness and prenatal care importance beliefs were 1.70 (1.23, 2.36) and 2.49 (1.46, 4.28), respectively. Fifth, although a relatively large set of covariates was included, additional unmeasured factors may influence delay of care. For example, barriers to prenatal care not measured in the current study include long waiting times for initial appointments, an inadequate number of providers, and complex application procedures.

The results suggest a need to expand community-based and preconception education programs that communicate the value of prenatal care, as well as family planning and other programs to reduce unintended childbearing. Social networks influence beliefs about prenatal care and prenatal behavior (8–10), and therefore messages directed at women of childbearing age as well as family, friends, and co-workers seem essential. In addition, the results suggest that by taking efforts to prevent unintended childbearing, such as family planning programs, delay of prenatal care could be reduced. These interventions should be conducted together with efforts to overcome financial and structural barriers to achieve progress toward national prenatal care objectives.