Prenatal Care Initiation Among Pregnant Teens in the United States: An Analysis Over 25 Years

Methods 

Results 

For all of the years studied, there was a consistent relationship between age and the time when prenatal care was initiated. Older adolescents were more likely to start care in the first trimester than younger adolescents, whereas younger adolescents were more likely to start care earlier than preteens. However among all three age groups the percentage initiating prenatal care during the first trimester increased after 1988 (Figure 1). Most of the increase appeared to be reflected in a decrease in the percentage initiating prenatal care during the second trimester, but there were also small decreases in the percentages of adolescents who initiated prenatal care during the third trimester or had no prenatal care at all.

View Large Image Download to PowerPoint

Figure 1. Trends over time for initiation of prenatal care in preteens aged 10–14 years, young adolescents aged 15–16 years, and older adolescents aged 17–19 years. Squares indicate percentage of patients starting prenatal care in the first trimester; diamonds, second trimester; triangles, third trimester; and circles, no prenatal care.

When we looked specifically at girls who received no care or waited until the third trimester to start their care, we found that younger adolescents or preteens were significantly more likely than older adolescents to have delayed care during each year we examined (p < .001) (Table 1). In a logistic regression model adjusted for survey year, race, marital status, education, residence, and prior births and using older adolescents as the reference group, we found that younger adolescents were 1.22 times more likely to have delayed care (95% confidence interval [CI] = 1.21–1.24) and preteens were 1.61 times more likely to have delayed care (95% CI = 1.56–1.65).

Table 1. Comparison of prenatal care initiation in by age group and birth year

	Prenatal care	Adolescents 10–14 years old		Adolescents 15–16 years old		Adolescents 17–19 years old
		N	%	N	%	N	%
	1978 (n)	9934		87,469		421,562
	1st trim	3535	35.6	39,357	45.0	240,531	57.1
	2nd trim	4429	44.6	35,914	41.1	139,448	33.1
	3rd trim	1413	14.2	9020	10.3	31,126	7.4
	None	557	5.6	3178	3.6	10,457	2.5
	1983 (n)	9255		76,572		393,933
	1st trim	3134	33.9	34,513	45.1	220,169	55.9
	2nd trim	4094	44.2	30,817	40.3	130,599	33.2
	3rd trim	1412	15.3	8039	10.5	31,195	7.9
	None	615	6.7	3203	4.2	11,970	3.0
	1988 (n)	10,049		77,452		380,949
	1st trim	3675	36.6	35,722	46.1	209,156	54.9
	2nd trim	4254	42.3	29,844	38.5	125,286	32.9
	3rd trim	1480	14.7	8404	10.9	32,841	8.6
	None	640	9.4	3482	4.5	13,666	3.4
	1993 (n)	11,968		88,414		393,060
	1st trim	5360	44.8	49,411	55.9	248,536	63.2
	2nd trim	4617	38.6	29,414	33.3	111,092	28.3
	3rd trim	1327	11.1	6765	7.7	23,552	6.0
	None	664	5.6	2824	3.2	9880	2.5
	1998 (n)	8978		76,137		386,828
	1st trim	4290	47.8	46,037	60.5	270,857	70.0
	2nd trim	3250	36.2	22,772	29.9	90,021	23.3
	3rd trim	1027	11.4	5269	6.9	18,580	4.8
	None	411	4.6	2059	2.7	7370	1.9
	2003 (n)	6351		57,199		340,588
	1st trim	3077	48.5	35,123	61.4	243,628	71.5
	2nd trim	2268	35.7	16,867	29.5	75,778	22.3
	3rd trim	705	11.1	3813	6.7	15,308	4.5
	None	301	4.7	1396	2.4	5874	1.7

Trim = trimester. For each year, the χ2 distribution is significant at p < .001; for each age group, the χ2 distribution is significant at p < .001.

In examining changes in delayed care between 1978 and 2003, the percentage of adolescents who had delayed prenatal care decreased significantly over time. As Table 2 illustrates, this was true in all three age groups. To adjust for the effects of changes in patient demographics and social factors over time, we performed logistic regression modeling that included the year of delivery along with race, marital status, years of education, residence, and prior births as independent variables. We also constructed separate models for each age group to control for maternal age (Table 3). After adjusting for other social and demographic factors, the likelihood of having delayed prenatal care (compared with 1978) was highest in 1983, then dropped in 1988, and continued downward during and after 1993. This trend was the same for all three age groups. In addition to the year of pregnancy, being unmarried and having less education were associated with an increased likelihood of delayed prenatal care.

Table 2. Percentage of pregnancies with delayed or no prenatal care among adolescents in three age groups

	Year	Adolescents 10–14 years old	Adolescents 15–16 years old	Adolescents 17–19 years old
	1978	19.8%	14.0%	9.9%
	1983	21.9%	14.7%	11.0%
	1988	21.1%	15.4%	12.2%
	1993	16.6%	10.9%	8.5%
	1998	16.0%	9.6%	6.7%
	2003	15.8%	9.1%	6.2%

χ2: p < .001 over time for all three groups.

Table 3. Adjusted relative likelihood (odds ratios and 95% confidence intervals) of having delayed or no prenatal care among adolescents of three age groups

	Variable	Adolescents 10–14 years old	Adolescents 15–16 years old	Adolescents 17–19 years old
	Year
	1978	1(Ref)	1(Ref)	1(Ref)
	1983	1.157(1.060–1.262)	1.040(1.005–1.077)	1.078(1.058–1.098)
	1988	1.073(.983–1.170)	1.011(.976–1.047)	1.088(1.068–1.108)
	1993	.843(.776–.916)	.716(.692–.741)	.737(.724–.751)
	1998	.789(.722–.862)	.619(.597–.604)	.563(.552–.574)
	2003	.747(.677–.824)	.580(.557–.604)	.527(.517–.538)
	Race
	White	1(Ref)	1(Ref)	1(Ref)
	Black	.926(.880–.974)	1.036(1.014–1.059)	1.128(1.114–1.142)
	Other	1.311(1.144–1.503)	1.456(1.380–1.537)	1.538(1.496–1.580)
	Marital status
	Unmarried	1(Ref)	1(Ref)	1(Ref)
	Married	.523(.464–.590)	.569(.551–.587)	.546(.539–.553)
	Education
	No high school	1.166(1.099–1.238)	1.202(1.173–1.231)	1.564(1.535–1.593)
	Some high school	1(Ref)	1(Ref)	1(Ref)
	High school grad	N/Aa	.893(.824–.968)	.782(.773–.791)
	Residence
	Nonmetropolitan	1(Ref)	1(Ref)	1(Ref)
	Metropolitan	.978(.916–1.181)	1.029(1.005–1.053)	1.039(1.026–1.051)
	Prior births
	None	1(Ref)	1(Ref)	1(Ref)
	1–3	1.040(.916–1.181)	1.588(1.545–1.633)	1.679(1.661–1.697)
	≥4	N/Ab	1.109(.620–1.985)	1.992(1.859–2.135)

grad = graduate; Ref = reference.

a There were no individuals in this group. b There were only four individuals in this group, so estimates are not statistically valid.

The effect of race on initiation of prenatal care differed depending upon the age of the mother. Among preteens, whites were more likely than black adolescents to delay prenatal care. However for the intermediate-aged and older adolescents, being a member of a racial/ethnic minority (black or other) was associated with greater likelihoods of delayed care compared with that in white adolescents.

Discussion 

These results demonstrate a trend toward earlier prenatal care initiation in teens of all ages in the U.S. between 1978 and 2003. The largest shift in earlier care and a corresponding reduction in the rates of late or no care occurred between 1988 and 1993, which coincides with liberalization in Medicaid-sponsored funding for pregnant women to obtain care [14], [15]. Although we have no direct evidence that state Medicaid expansion efforts directly resulted in our findings, younger patients are more likely to be uninsured and are likely to benefit the most from the expanded benefit requirements. Our study did not examine other factors that may have contributed to these changes; nonetheless the expanded Medicaid eligibility is likely to have played a major role in spurring these shifts in the initiation of prenatal care.

Our findings for teens are consistent with previous reports showing that the percentage of women in the U.S. who obtained adequate prenatal care increased significantly between 1981 and 1995 [16]. Among teenaged girls, however, most of these gains were achieved through earlier care for women who previously had begun care in the second trimester. Because there is little impact on low–birth-weight rates when delaying care to the second trimester [17], it is uncertain whether this change had any significant impact on pregnancy outcomes because the birth certificate data before 1993 did not include birth outcomes.

Despite improvements, even in 2003 a sizeable proportion of younger adolescents (9.1%) and preteens (15.8%) received either no care or care starting after the second trimester. These delays could reflect the patients’ unfamiliarity with available services or a lack of understanding about what kind of care is expected during pregnancy, along with desires of teenage girls to conceal their pregnancies. The fear in younger girls of having a pregnancy discovered may be a significant factor in the higher likelihood that girls in this age group would delay seeking early care. This possibility is supported by our observation that adolescents who had prior births were more likely to have either delayed or no care, suggesting that the lack of awareness about prenatal services may not be the sole explanation. As the lack or absence of prenatal care is associated with increases in low birth weight, a greater impact on low birth weight would occur if women who did not receive care started their prenatal visits earlier in their pregnancy [18]. This would also have additional benefits because, when comparing babies of similar weight, those born without prenatal care have other unfavorable outcomes [19]. Although the cause of poor pregnancy outcomes is multifactorial and includes both biological determinants and socioeconomic factors, medical care can likely play a role in mitigating some of these factors. Women who delay prenatal care or fail to obtain care are more likely to be in a “social chaos” situation characterized by poverty, unstable relationships, substance abuse, and interfamily violence [20], [21], [22], [23]. All of these can contribute to low birth weight through inadequate nutrition, exposure to infection, and use of drugs (such as tobacco or cocaine) that increase preterm delivery. However a prenatal care program can also identify and address some of these at-risk conditions, resulting in reductions in low–birth-weight deliveries [24]. This study emphasizes the continued need to develop programs specifically aimed at promoting early prenatal care for teens and the other high-risk groups.

This study also provides some insight into which patient populations are at higher risk for not receiving timely prenatal care, and can help public health planners focus on programs that address these populations. For example we observed that girls who had a prior delivery are at higher risk for not obtaining prenatal care on a subsequent pregnancy. Not only is this population less likely to receive early care but, because of the young age of this population, this also suggests that pregnancies have not been spaced optimally. These findings suggest that continued follow-up of pregnant girls may be useful to assist in future pregnancy planning and to identify those who become pregnant so that prenatal care can be encouraged while barriers obstructing early prenatal care (such as child care or transportation) can be addressed.

Another finding worth noting is the differential effect of race on obtaining early prenatal care. For the youngest adolescents, white teens were at higher risk for delaying prenatal care. However for the 15–16-year age group and the older group, adolescents of white race/ethnicity were less likely to delay care. The reasons for this differential effect of race on prenatal care are not readily obvious. It is unclear whether these changes reflect possible differences in social stigmatization of teenage pregnancy in teens of white and black ethnicities, causing younger white girls to attempt to conceal their pregnancy longer, differences in access to care between white and black adolescents as they get older, greater social isolation of minority teens as they become older, or fundamental differences in the at-risk populations between the very young pregnant adolescents and older girls. Additional research in this area would be helpful to ascertain the contributions of these factors and to examine which are amenable to intervention.

The conclusions of this study must be interpreted in light of limitations in the data set and study design. First is the source of our data, i.e., national birth certificate data. By their very nature, these data source only provide information on pregnancies that give birth to viable infants. Consequently changes in the availability or use of abortion, especially if abortions became more common for adolescents who would have not received prenatal care, also can influence these data. In addition the accuracy of birth certificate data is dependent on a variety of nonvalidated sources. Birth certificates capture information about pregnant women from all available sources including patient records. However when prenatal care documentation is not available, patient recall is used and may not be as accurate. Furthermore those asked to recall when prenatal care began may be aware that earlier prenatal care is socially desirable, so when asked to recall when they began care some women may have estimated care to have started earlier than it actually had.

Also we did not examine every year in the 25-year observation period that we selected. It is possible that 1 or more years that we did include in our analysis were aberrations and did not reflect secular trends over the time period. However given the large number of teenagers delivering in any given year (>400,000), it is unlikely that large deviations in practice would occur in any single given year, and we did not believe that additional years between our index years would provide significantly improved information about the trends over time.

Finally, in an ecological study such as this, is it difficult (if not impossible) to ascribe causation when two variables are linked over time. Consequently, although Medicaid policy changes occurred in the same time frame as the shift in prenatal care access, we must be cautious that this association is not viewed as the only explanation for our results. Other changes in adolescent pregnancy patterns, including a reduction in pregnancies among adolescents at high risk or differences in abortion use, could contribute to the changes tha twe observed.

Conclusion 

In summary, this study shows that expanded health care eligibility for younger and poorer women was accompanied by a decrease in the percentage of women receiving delayed prenatal care. This is primarily an effect of women starting care more often in the first trimester and less often in the second trimester. A younger age at delivery, less education, and prior births were all associated with increased likelihood of receiving delayed care.