- Email: [email protected]
Differences in Pregnancy Desire Among Pregnant Female Adolescents at a State-Funded Family Planning Clinic
Differences in Pregnancy Desire Among Pregnant Female Adolescents at a State-Funded Family Planning Clinic Elizabeth J. Heavey, CNM, PhD, Kirsten B. Moysich, PhD, Andrew Hyland, PhD, Charlotte M. Druschel, MD, PhD, and Michael W. Sill, PhD We wished to examine variables associated with pregnancy desire among pregnant adolescents from low socioeconomic backgrounds. This study analyzed 335 charts at a state-funded family planning clinic. Participants were adolescents who had a positive pregnancy test at the clinic on the day of the survey. Logistic regression was utilized to determine differences in pregnancy desire. We found that Hispanic teens were more than twice as likely to desire pregnancy as African American teens (adjusted odds ratio [AOR], 2.22; 95% confidence interval [CI], 1.22–3.65), and adolescents who were not in school were almost twice as likely as those who were in school full-time to desire pregnancy (AOR, 1.83; 95% CI, 1.08 –3.09). Hispanic teens who were not in school were 12 times more likely to desire pregnancy than African American teens who were in school full-time (odds ratio [OR], 11.47; 95% CI, 3.68 –35.75). Adolescent pregnancy desire is significantly associated with educational status and racial background. Developing culturally appropriate interventions to encourage continued education and asking about community and familial norms are essential steps in addressing this issue. J Midwifery Womens Health 2008;53:130 –137 © 2008 by the American College of Nurse-Midwives. keywords: adolescent pregnancy, education, pregnancy desire, race, teen pregnancy
INTRODUCTION Despite some reported improvement over the last decade, the rates of adolescent pregnancy in the United States remain high compared to other developed nations.1,2 Although the age of onset of sexual activity and sexual activity patterns do not differ significantly among most developed nations, there are significant differences in the acceptability and availability of information about adolescent sexuality, along with the differences in social expectations regarding pregnancy.3–5 According to the National Center for Health Statistics, the highest recorded teen pregnancy rates in the United States were in 1990, at 11.6% for women ages 15–19.6 These numbers dropped between 1990 and the year 2000.6 Even with this improvement, 8.4% of women ages 15–19 were pregnant in the year 2000.6 This translates into 852,000 teen pregnancies, which resulted in more than 478,000 live births to adolescents in this single year alone.6 Efforts to promote abstinence and encourage effective contraception usage appear to have had some effect, but the Healthy People 2010 goal of decreasing adolescent pregnancy rates to 4.6% has not been reached. Further, the female adolescent population will grow by approximately 10% within the next 5 years.7 Even with stable or declining rates of adolescent pregnancy, this may translate into a large increase in the number of actual births.7 The public health costs associated with teen pregnancy are significant and the social implications are staggering. The demographics of teen pregnancy have shifted in the
Address correspondence to Elizabeth Heavey, CNM, PhD, State University of New York College at Brockport, School of Nursing, 350 New Campus Dr., Brockport, NY 14420-2988. E-mail: [email protected]
130 © 2008 by the American College of Nurse-Midwives Issued by Elsevier Inc.
last 40 years. Today, pregnant adolescents are much more likely to be unmarried, to continue to have subsequent pregnancies into their twenties while not married, and have a greater likelihood to need extended welfare support than in decades past.5 The National Center for Health Statistics reports that in 2002, two-thirds of all adolescent births were paid for by the government, while only 14% of those births to women over the age of 30 required this financial assistance.6 Further, the economic and social burden is not felt equally among all people. There are significant racial differences in teen pregnancy and birth rates.8 In the year 2000, white teens had a pregnancy rate of 5.7%, while Hispanic teens had a rate of 13.2% and African American teens were at 15.1%.6 Adolescent birth rates in the year 2004 were highest among Hispanic adolescents at 8.2%, while African American adolescents have a birth rate of 6.3% and white adolescents have a rate of 2.7%.6 In addition, one-quarter of adolescent births are to teens who have already given birth.7 While some decline in pregnancy rates is noted among all teens, the rates among minority teens continue to be more than double those of white teens. In fact, the overall decline in adolescent pregnancy and birth rates since the 1980s is the least among Hispanic adolescents.6 Many explanations have been offered as to why this phenomenon occurs. Contraceptive decision-making may be impacted by racial identity.9 Lower socioeconomic status of minority women, combined with limited educational and employment opportunities, may impact the ability or desire to use contraception and thereby increase pregnancy rates.9 Unfortunately, it is very difficult to separate poverty effects from racial effects because of a high degree of association between these covariates. Volume 53, No. 2, March/April 2008 1526-9523/08/$34.00 • doi:10.1016/j.jmwh.2007.09.005
Poverty has a direct impact on the type of childhood experiences an adolescent may have had and becomes both a social and financial condition.10 These years are filled with significantly more work and family responsibilities and may be so chaotic that mere survival is the primary goal.10 The extensive effects of poverty may impair adolescent coping, diminish self-esteem, and foster adolescent sexual relationships, thereby impacting adolescent pregnancy rates.11 There are many studies which show a correlation between social and economic disadvantage and adolescent pregnancy.12–18 Some reports suggest that adolescents may see pregnancy as a good alternative when faced with a lack of opportunity and direction.7,19 –22 Some researchers have argued that early pregnancy and childbearing are an adaptive response to the cumulative disadvantage associated with the combination of poverty and minority status.12,23–28 One study examining African American, Hispanic, and white women found that each group had the largest number of first births during the time period when the infant mortality rate was the lowest.23 The combination of infant health advantage and additional familial support available for impoverished teens who become parents may improve the ability of these teens to minimize long-term risks associated with delaying pregnancy until after the age of 20.23 Teen pregnancy may actually be a rational decision directly associated with the limited nature of the options which are available to this group of adolescents.23 The current paradigm for sexual decision-making and teen pregnancy interventions frequently assumes that teens are getting pregnant by mistake. Educational interventions are structured around this belief, and adolescents are approached with this assumption by health care providers. Frequently, the lack of future opportunities, familial and peers’ perceptions of pregnancy, and limited alternatives are not taken into account. This may be why many adolescent pregnancy prevention programs have not been effective in decreasing pregnancy rates or birth rates.29 –35 If a significant number of adolescents desire pregnancy or are ambivalent about pregnancy, the current approach may continue to fail. This study examined
Elizabeth J. Heavey, CNM, PhD, is an epidemiologist specializing in maternal/child health and an assistant professor in the School of Nursing at SUNY Brockport. She is also completing a post-doctoral fellowship in Adolescent Medicine at the University of Rochester. Kirsten B. Moysich, PhD, Roswell Park Cancer Institute, Department of Epidemiology. Andrew Hyland, PhD, Roswell Park Cancer Institute, Department of Health Behavior. Charlotte M. Druschel, MD, PhD, New York State Department of Health, SUNY at Albany. Michael W. Sill, PhD, Roswell Park Cancer Institute, Gynecology Oncology Group.
Journal of Midwifery & Women’s Health • www.jmwh.org
the association between educational status, racial background, and adolescent pregnancy desire. METHODS Participants were all female adolescents between the ages of 12 and 19 who had a positive pregnancy test at a family planning clinic on the day of their appointment. The data were originally collected from 1999 –2004. Racial identity was self-reported by participants. Adolescents had to be 19 years of age or younger and below state income limits in order to participate in this particular family planning program. Income requirements were met if the adolescents already qualified for public health insurance or had no health insurance at the time of the visit. There were a small number of adolescents who had private health insurance and were still included in the program after meeting with a social worker at the clinic who determined their income eligibility according to the state guidelines. Those who were not eligible for participation in the state-funded program were not included in this study. The census tract in which the clinic is located has 43% of families living below the poverty line and 25% of families relying on public assistance.36 Within the five major zip code areas served by the clinic, 60%–75% of the births were covered by Medicaid, and teen pregnancy rates were approximately 20% in 2000.37 All participants were deemed eligible for free contraception available on site at the time of the visit or through public health insurance coverage. The information utilized was originally part of the recordkeeping for the New York State Family Planning Program and is the same data collected at all the reproductive health care sites participating in this program throughout the state. The information collected included questions regarding age, sex, monthly income, family size, educational level completed, student status (enrolled full time, part time, or not enrolled), number of pregnancies, number of births, and additional sources of health care. Additional clinical information was collected regarding contraception use, services required, and services provided. Demographic and interview information on the chart was originally completed by a registered nurse at the clinic during a private intake session with the patient. The nurse had completed a brief training session and all questions and counseling were available in both English and Spanish. Age was categorized as 14 years and younger, 15–17 years, and 18 –19 years because of noted clinical differences among these populations, differences cited in developmental literature, and to match the categories utilized by the U.S. Bureau of the Census.38,39 Race was self-identified by participants from a list including white, black, American Indian, Alaskan Native, Asian, or other. In addition, participants were asked if they would identify themselves as 131
Hispanic. Because the number of previous births and the number of previous pregnancies were highly correlated, this analysis utilized only the variable reflecting the number of previous pregnancies. Patients were asked about their desire for pregnancy during the same intake session in which they were told they were pregnant. Pregnancy desire was measured by the adolescent’s response to the question, “Was your pregnancy desired now, desired sooner, desired later, not desired, or unsure?” Patients who indicated their current pregnancy was desired now or desired sooner are categorized as having intended pregnancies, while those which were desired later or not desired are classified as having unintended pregnancies. Those who indicated they were unsure were classified as ambivalent. The clinical information was then completed by the provider who examined the patient. Permission to analyze the data was obtained from the director of the state program and the medical director at the clinic. The data obtained from the forms were de-identified before use and provided no opportunity to track repeat visits. The review of the records for the current study was approved by the institutional review board at the State University of New York College at Brockport. All statistical analyses were done utilizing the SAS program (version 9.0; SAS Institute, Cary, NC). Associations between demographic variables and pregnancy desire were examined utilizing the 2 test for categorical data and t tests for continuous variables. The unadjusted odds ratio was calculated for each independent variable and pregnancy desire, and those which were significant were included in the logistic regression to examine adjusted odds ratios. Those variables which had a P value of ⬍.05 were included in the final model. We hypothesized a priori that a significant number of adolescents at the clinic wanted to be pregnant now or sooner. In order to utilize data from teenagers who expressed ambivalent desire for pregnancy, our first model grouped the ambivalent teens with those who desired pregnancy. Our second model then eliminated the ambivalent teens and only those who clearly desired pregnancy or clearly did not desire pregnancy were compared. Lastly, the ambivalent teens were grouped with those who did not desire pregnancy. Logistical regression analysis was utilized to examine the groups of teenagers defined by pregnancy desire which was the dependent variable. These models assume independence in visits, because we had no means to determine the proportion of repeat visits. The sample was stratified by race and student status, because we hypothesized a priori that pregnancy desire would differ in these subgroups. RESULTS We retrospectively reviewed 349 charts. Ten charts were excluded because of incomplete data. We also excluded 132
four charts from adolescents not identified in the three main racial groups. The final analysis was limited to 335 charts. The sample population was largely composed of teens from minority racial groups, and more than half of the population received public insurance, with more than 44% uninsured. Sixty-two percent of the sample reported using no contraception, and 32% were no longer in school. Fifty-four percent of the adolescents were pregnant for the first time, while 33% were for the second time and more than 12% were pregnant two or more times previously. The mean age was 17.4 years. Sample population characteristics are summarized in Table 1. Adolescents responded to a question regarding pregnancy desire with 92 indicating a desired pregnancy (27%), 124 being ambivalent about their pregnancy (37%), and 119 reporting an undesired pregnancy (36%). The first analysis grouped those who were ambivalent with those who desired pregnancy, and found that compared to African American adolescents, Hispanic teenagers were twice as likely to desire or be ambivalent about pregnancy (Table 2). There was no significant difference between white adolescents and African American adolescents in terms of pregnancy desire/ambivalence (Table 2). In addition, adolescents who were not in school were almost twice as likely to desire or be ambivalent about pregnancy as those who were in school full time (Table 2). The Hispanic adolescents who were not in school were six times as likely to either desire or be ambivalent about pregnancy as the African American adolescents who were in school full time. After stratifying by race, the Hispanic adolescents who were not in school were three times as likely to desire or be ambivalent about pregnancy as those who were in school full time. There were no significant relationships by student status found for white or African American adolescents. We also examined the impact of race after stratifying the adolescents into either full-time student status, parttime status, or not in school, and found significant relationships only for those who were not in school. Among adolescents not in school, Hispanic teens are four times as likely to desire pregnancy compared to African American teens, and there was no significant difference between white teens and African American teens. In addition, interaction terms for student status and number of previous pregnancies, number of previous pregnancies and age category, and student status and race were tested and were not significant. The analysis was repeated, eliminating the ambivalent group and comparing only those who desired pregnancy and those who did not desire pregnancy. The observed associations were still significant and increased in magnitude (Table 3). Hispanic clients were two and a half times as likely to desire pregnancy compared to the African American teenagers, and white adolescents were not significantly more likely to desire pregnancy than African American adolescents (Table 3). The young Volume 53, No. 2, March/April 2008
Table 1. Characteristics of 335 Teenagers with a Positive Pregnancy Test (1999 –2004) Variable
Desired Pregnancy [n ⴝ 92; n (%)]
Ambivalent† [n ⴝ 124; n (%)]
Undesired Pregnancy [n ⴝ 119, n (%)]
0 (0) 41 (44.6) 51 (55.4)
4 (3.2) 57 (46.0) 63 (50.8)
7 (5.9) 48 (40.3) 64 (53.8)
8 (8.7) 47 (51.1) 37 (40.2)
9 (7.3) 78 (62.9) 37 (29.9)
9 (7.6) 87 (73.1) 23 (19.3)
41 (44.6) 5 (5.4) 46 (50.0)
66 (53.2) 2 (1.6) 56 (45.2)
72 (60.5) 3 (2.5) 44 (37.0)
49 (53.3) 1 (1.1) 42 (45.7)
83 (67.5) 3 (2.4) 37 (30.1)
88 (74.0) 4 (3.4) 27 (22.7)
52 (56.5) 26 (28.3) 7 (7.6) 7 (7.6)
73 (58.9) 37 (29.8) 10 (8.1) 4 (3.2)
59 (49.6) 46 (38.7) 11 (9.2) 3 (2.5)
Age Category (Years) ⱕ14 years 15–17 years 18–19 years Racial Background White African-American Hispanic Type of Insurance Public Private None Student Status Full-time Part-time None Previous Pregnancies 0 1 2 ⱖ3
P* .19
.02
.12
.01
.32
*Significance tested using 2 test. † Missing student status data for 1 participant.
women who were not in school were more than twice as likely to desire pregnancy (Table 3). The Hispanic adolescents who were not in school were almost 12 times as likely to desire pregnancy when compared to the African American adolescents who were in school fulltime. After stratification by student status, the Hispanic adolescents who were not in school were eight times as likely as those who were in school to desire pregnancy. There were no significant relationships found for the white or African American adolescents. We again stratified by student status into part-time, not in school, and full-time, and only found significant relationships among adolescents who were not in school. Hispanic teens who were not in school were 17 times as likely to desire pregnancy compared to African Ameri-
can teens who were not enrolled in school. There was no significant difference between white teens and African American teens, but there was a significant association with age. For adolescents who were not in school, each additional year of age was associated with a twofold increase in desire for pregnancy. We conducted further sensitivity analyses by combining the ambivalent group with the adolescents who did not desire pregnancy and compared this combined group to those who did desire pregnancy. Again, the Hispanic adolescents were almost twice as likely to desire pregnancy compared to the African American adolescents (Table 4). The pregnancy desire among white adolescents was not significantly different compared to African American adolescents (Table 4). Adolescents who were
Table 2. Association Between Pregnancy Desire and Sample Characteristics in Teenagers with a Positive Pregnancy Test* (N ⫽ 335) Variable
Desired and Ambivalent about Pregnancy [n ⴝ 216†; n (%)]
Undesired Pregnancy [n ⴝ 119; n (%)]
OR (95% CI)
Adjusted OR (95% CI)
Hispanic White African American Part-time student Not in school Full-time student
74 (24.9) 17 (9.0) 125 (66.1) 4 (1.9) 79 (36.7) 132 (61.4)
23 (19.3) 9 (7.6) 87 (73.1) 4 (3.4) 27 (22.7) 88 (73.9)
2.26 (1.31–3.88) 1.33 (0.56–2.31) Reference group 0.67 (0.16–2.73) 1.95 (1.17–3.26) Reference group
2.11 (1.22–3.65) 1.11 (0.46–2.67) Reference group 0.67 (0.16–2.80) 1.83 (1.08–3.09) Reference group
CI ⫽ confidence interval; OR ⫽ odds ratio. *Number of previous pregnancies, age, type of health insurance, and other sources of healthcare were tested and were not significant. †
Missing student status data for 1 participant.
Journal of Midwifery & Women’s Health • www.jmwh.org
133
Table 3. Association Between Pregnancy Desire and Sample Characteristics in Teenagers with a Positive Pregnancy Test* (N ⫽ 211) Variable
Desired Pregnancy [n ⴝ 92; n (%)]
Undesired Pregnancy [n ⴝ 119; n (%)]
OR (95% CI)
Adjusted OR (95% CI)
Hispanic White African American Part-time student Not in school Full-time student
37 (40.2) 8 (8.7) 47 (51.1) 1 (1.1) 42 (45.7) 49 (53.3)
23 (19.3) 9 (7.6) 87 (73.1) 4 (3.4) 27 (22.7) 88 (73.9)
2.98 (1.59–5.59) 1.65 (0.60–4.55) Reference group 0.45 (0.05–4.13) 2.80 (1.54–5.07) Reference group
2.54 (1.33–4.87) 1.34 (0.47–3.83) Reference group 0.46 (0.05–4.36) 2.42 (1.31–4.48) Reference group
CI ⫽ confidence interval; OR ⫽ odds ratio. *Number of previous pregnancies, age, type of health insurance, and other sources of healthcare were tested and were not significant.
not in school were twice as likely to desire pregnancy as those who were in school (Table 4). After stratification by race and student status, there were no significant relationships found among the variables for any of the racial groups. DISCUSSION This study clearly shows racial and educational differences in pregnancy desire among low-income adolescent women. However, several limitations need to be addressed in the interpretation of these findings. We lacked a measure of the strength of cultural identity and had no way to further divide categories of racial/cultural background. There may be distinct differences among these broad racial categories which we were unable to detect given the available data. The sample was a convenience sample drawn from the clients utilizing a community-based obstetric/gynecology clinic. The sample is drawn largely from the community surrounding the clinic. Thus the findings may not be generalizable to this population or to populations that differ significantly from this one, including populations with higher socioeconomic status and improved access to future economic or educational opportunities. Because the study relied on previously collected data, we were unable to include all relevant variables, including some of the psychological predictors, family struc-
ture, drug use, and partner’s perceptions of contraception and pregnancy. Previous studies have reported an association between pregnancy and lack of self-esteem, drug use, and partners who desire pregnancy.40 Some of these factors may be confounding variables or even the actual variable of interest. For example, student status may simply be a reflection of the adolescent’s self-esteem, in which case our study would be reporting an association between pregnancy desire and a confounding variable, not the actual variable with which there is an association. Because of the privacy limitations imposed by the clinic administration, it was not possible to track those individuals who had repeat measures in the dataset. The lack of independence for some of the observations may limit the accuracy of the inferences drawn from these data. The nature of the data meant relying on adolescents’ self-reports for behavioral variables. Self-reports of sensitive information are frequently inaccurate, and underestimate the occurrence of socially sanctioned behavior or experiences. Underestimates of pregnancy desire are expected, and more accurate data would be expected to increase the associations found in this study. Alternatively, pregnancy desire may be the same across all groups, but the differences in ethnic norms may be related to differences in the acceptance and acknowledgement of pregnancy desire.
Table 4. Association Between Pregnancy Desire and Sample Characteristics in Teenagers with a Positive Pregnancy Test* (N ⫽ 335) Variable
Desired Pregnancy [n ⴝ 92; n (%)]
Ambivalent and Undesired Pregnancy [n ⴝ 243†; n (%)]
OR (95% CI)
Adjusted OR (95% CI)
Hispanic White African American Part-time student Not in school Full-time student
37 (40.2) 8 (8.7) 47 (51.1) 1 (1.1) 42 (45.7) 49 (53.3)
60 (24.7) 18 (7.4) 165 (67.9) 7 (2.9) 64 (26.4) 171 (70.7)
2.21 (1.30–3.74) 1.54 (0.63–3.76) Reference group 0.50 (0.06–4.15) 2.290 (1.39–3.79) Reference group
1.97 (1.15–3.35) 1.24 (0.49–3.10) Reference group 0.50 (0.06–4.24) 2.14 (1.28–3.58) Reference group
CI ⫽ confidence interval; OR ⫽ odds ratio. *Number of previous pregnancies, age, type of health insurance, and other sources of healthcare were tested and were not significant. †
Missing student status data for one participant.
134
Volume 53, No. 2, March/April 2008
Despite these limitations, there are several noteworthy strengths associated with the current study. Because socioeconomic status is a significant contributor to adolescent pregnancy desire, any study on this topic must control for socioeconomic variability in order to examine other contributing variables.41 In this study, the combination of income requirements for participation in this program and the fact that the majority of the participants resided in the high-poverty zip codes resulted in a fairly well controlled sample of individuals with limited socioeconomic means, living in communities with limited socioeconomic resources, which minimized residual confounding from socioeconomic status. In addition, participants were asked about their desire for pregnancy at the same visit as their pregnancy test. When asked about pregnancy desire after their pregnancy, adult women are more likely to indicate that pregnancy occurred at the right time, even if this was not true at the time of pregnancy confirmation.42 However, there is no research indicating how an adolescent’s perspective might change, and we postulated that the reverse may be true of female adolescents. The reality of childbirth, loss of peer connections and romantic relationships, responsibilities of parenthood, and the constant needs of an infant may make an adolescent more likely to look back and indicate a pregnancy was not desired when it actually was at the time of conception. Our study avoided a retrospective bias which could have affected the desire for pregnancy negatively or positively, and provided a better measure of pregnancy desire. This study includes several points of interest which require further research. Hispanic teenagers are much more likely to desire pregnancy, which is a significant risk factor for adolescent pregnancy. Gender roles within Hispanic households may encourage pregnancy as a way for women to find fulfillment.43 Early and frequent pregnancies may be seen in a positive light.43 In addition, female independence is not a traditional value, and continued education may be seen as incompatible with pregnancy and childbearing.43 Lack of educational enrollment is also significantly associated with pregnancy desire, and when both risk factors are present, there is a much higher likelihood that an adolescent will want to become pregnant. These findings suggest that racial norms and educational achievement have a large impact on adolescent pregnancy risk. Although there are many reports of Hispanic values, there are no previous studies which address how they may impact adolescent pregnancy desire. Even among adult women, Hispanic women are frequently not considered as a separate demographic population and disparities in reproductive health care are not well documented. Most of the previous studies on adolescent pregnancy desire have focused exclusively on African American teens, differences between white and African Journal of Midwifery & Women’s Health • www.jmwh.org
American teens, or have not examined the role that race may play.14,19,38,44 – 46 These findings suggest a need to develop pregnancy prevention interventions specifically designed to address adolescents of all races who desire pregnancy. Such interventions have been sorely neglected in the literature. It may be that this population is amenable to at least delaying pregnancy and this may be an improvement over very early childbearing. CONCLUSION The clinical implications of this study include the need for adolescent family planning programs to include an assessment of pregnancy desire. Because peers and community have such a significant impact in the teen years, clinicians need to explore how teens feel about pregnancy within this context. It is important to consider community social structure and familial norms and include this information in the subsequent plan of care. The single screening question used in this study may help identify adolescents who desire pregnancy. Simply asking about pregnancy desire allows a clinician to identify adolescents who may be at higher risk, including those who are ambivalent. Behaviors associated with ambivalence and desiring pregnancy are the same and only the group of women who do not want to be pregnant behave differently clinically.14,22,38 Ignoring the feelings of teenagers regarding pregnancy might result in missing an opportunity to address a key factor in pregnancy risk. This study identified adolescent pregnancy desire and pregnancy ambivalence as a much more common occurrence than many health care professionals believe. This issue is more complex than a lack of information or a lack of economic resources. Although contraception education and free services are helpful for those wishing to avoid pregnancy, they may not be effective interventions for young women who do not wish to avoid pregnancy. For adolescents identified as desiring pregnancy, further interventions are needed to address the complex reasons why. The solutions in this case may not be as simple as a straightforward talk about contraception, but they need to be identified before the adolescent pregnancy goals of Healthy People 2010 can become a reality. Supported by the Mark Diamond Research Fund, The Graduate Student Association at SUNY Buffalo, and Roswell Park Cancer Institute.
REFERENCES 1. Singh S, Wulf D, Samara R, Cuca YP. Gender differences in the timing of first intercourse. Int Fam Plann Perspect 2000;26: 21– 8. 2. Moore M. Adolescent pregnancy rates in three European countries. J Obstet Gynecol Neonatal Nurs 2000;29:355– 62.
135
3. Feldmann J, Middleman A. Adolescent sexuality and sexual behavior. Curr Opin Obstet Gynecol 2002;14:489 –93.
quences of teen childbearing reconsidered. Q J Econ 1992; 107:1187–1214.
4. Kartoz C. New options for teen pregnancy prevention. Am J Matern Child Nurs 2004;29:30 –5.
25. Geronimus A. Commentary: Weathering Chicago. Int J Epidemiol 2003;32:90 –1.
5. Coley R, Chase-Lansdale P. Adolescent pregnancy and parenthood. Am Psychol 1998;53:152– 66.
26. Satin A, Leveno K, Sherman L, Reedy N, Lowe T, McIntire D. Maternal youth and pregnancy outcomes. Am J Obstet Gynecol 1994;171:184 –7.
6. National Center for Health Statistics Web site. Available from: www.cdc.gov [Accessed March 17, 2006]. 7. Klein J. Adolescent pregnancy: Current trends and issues. Am Acad Pediatr 2005;116:281– 6. 8. Elster A, Jarosik J, VanGeest J, Fleming M. Racial and ethnic disparities in health care for adolescents: A systematic review of the literature. Arch Pediatr Adolesc Med 2003;157:867– 74. 9. Raine T, Harper C, Paukku M, Darney P. Race, adolescent contraceptive choice, and pregnancy at presentation to a family planning clinic. Obstet Gynecol 2002;99:241–7. 10. Smith L. Psychotherapy, classism and the poor. Am Psychol 2005;60:687–96. 11. Culhane J, Elo I. Neighborhood context and reproductive health. Am J Obstet Gynecol 2005;192:S22–9. 12. Lu M, Halfon N. Racial and ethnic disparities in birth outcomes. Matern Child Health J 2003;7:13–30. 13. Jewell D, Tacchi J, Donovan J. Teenage pregnancy: Whose problem is it? Fam Pract 2000;17:522– 8. 14. Zabin L, Astone N, Emerson M. Do adolescents want babies? J Res Adolesc 1993;3:67– 86. 15. SmithBattle L. The vulnerabilities of teenage mothers. Adv Nurs Sci 2000;23:29 – 40. 16. Duncan G, Hoffman S. Welfare benefits, economic opportunities, and out-of-wedlock births among black teenage girls. Demography 1990;4:519 –35. 17. Brewster K, Billy J, Grady W. Social context and adolescent behavior. Social Forces 1993;71:713– 40. 18. Cherlin A, Kiernan K, Chase-Lansdale P. Parental divorce in childhood and demographic outcomes in young adulthood. Demography 1995;32:1–10. 19. Montgomery K. Planned adolescent pregnancy: What they want. J Pediatr Health Care 2002;16:282–9. 20. Hellerstedt W, Fee R, McNeely C, Sieving R, Shew M, Resnick M. Pregnancy feelings among adolescents awaiting pregnancy test results. Public Health Rep 2001;116:180 –93.
27. Rich-Edwards J, Buka S, Brennan R, Earls F. Diverging associations of maternal age with low birthweight for black and white mothers. Int J Epidemiol 2003;32:83–90. 28. Rauh V, Andrews H, Garfinkel R. The contribution of maternal age to racial disparities in birthweight. Am J Public Health 2001;91:1815–24. 29. DiCenso A, Guyatt G, Willan A, Griffith L. Interventions to reduce unintended pregnancies among adolescents. Br Med J 2002;324:1–17. 30. Maynard R. Teenage childbearing and welfare reform. Child Youth Serv Rev 1995;17:309 –32. 31. Bennett S, Assefi N. School-based teenage pregnancy prevention programs: A systematic review of randomized controlled trials. J Adolesc Health 2005;36:72– 81. 32. Kirby D. Emerging answers: Research findings on programs to reduce teen pregnancy. Washington, DC: National Campaign to Prevent Teen Pregnancy, 2001. 33. Robin L, Dittus P, Whitaker D, Crosby R, Ethier K, Mezoff J, et al. Behavioral interventions to reduce incidence of HIV, STD and pregnancy among adolescents: A decade in review. J Adolesc Health 2004;34:3–26. 34. Guyatt G, DiCenso A, Farewell V, Willan A, Grifith L. Randomized trials versus observational studies in adolescent pregnancy prevention. J Clin Epidemiol 2000;53:167–74. 35. Nicoletti A. Perspectives on pediatric and adolescent gynecology from the allied health professional. Teens, tattoos, and body piercings. J Pediatr Adolesc Gynecol 2004;17:215– 6. 36. Johnson W. State of the City Mayorial Speech. March 1, 2004. Available from: www.cityofrochester.gov [Accessed February 23, 2005]. 37. New York State Health Department. As reported in 2004 Grant Application, Jordan Health Center, Rochester, NY. 38. Lindemann C, Scott W. Wanted and unwanted pregnancy in early adolescence. J Early Adolesc 1981;1:185–93. 39. Davis A. The role of hormonal contraceptives. Am J Obstet Gynecol 1994;170(5 Suppl):1581–5.
21. Davies S, DiClemente R, Wingood G, Harrington K, Crosby R, Sionean C. Pregnancy desire among disadvantaged African American adolescent females. Am J Health Behav 2003;27:55– 62.
40. Modrcin-Talbott M, Pullen L, Zandstra K, Ehrenberger H, Muenchen B. A study of self-esteem among well adolescents. Issues Compr Pediatr Nurs 1998;21:229 – 41.
22. Crosby R, DiClemente R, Wingood G, Harrington K, Davies S, Hook E, et al. Psychosocial predictors of pregnancy among low-income African-American adolescent females. J Pediatr Adolesc Gynecol 2002;15:293–9.
41. Young T, Turner J, Denny G, Young M. Examining external and internal poverty as antecedents of teen pregnancy. Am J Health Behav 2004;28:361–73.
23. Geronimus A. The weathering hypothesis and the health of African-American women and infants. Ethn Dis 1992;2:207–21.
42. Sable M. Pregnancy intentions may not be a useful measure for research on maternal and child health outcomes. Fam Plann Perspect 1999;31:249 –50.
24. Geronimus A, Korenman S. The socioeconomic conse-
43. Russell S, Lee F. Practitioners’ perspectives on effective
136
Volume 53, No. 2, March/April 2008
practices for Hispanic teenage pregnancy prevention. Perspect Sexual Reprod Health 2004;36:1–17.
Harrington K, et al. Predictors of inconsistent contraceptive use among adolescent girls. J Adolesc Health 2006;39:43–9.
44. Bloom K. Pregnancy wantedness in adolescents presenting for pregnancy testing. Am J Matern Child Nurs 1999;24:296 –300.
46. Davis S, DiClemente R, Wingood G, Person S, Crosby R, Harrington K, et al. Relationship characteristics and sexual practices of African American adolescent girls who desire pregnancy. Health Educ Behav 2004;31:85S–96S.
45. Davies S, DiClemente R, Wingood G, Person S, Dix E,
ACCESS THE JOURNAL OF MIDWIFERY & WOMEN’S HEALTH ONLINE ARTICLES Articles published in the Journal of Midwifery & Women’s Health are available online to ACNM members and JMWH subscribers. To access your online subscription for the first time: 1. Type in www.jmwh.com/claim 2. Go to the “New User” box and enter your last name and email address 3. Click on “Sign in” 4. At the registration screen, select a user name and password and fill in requested details 5. Read the registered user agreement at the end of the form and click box 6. Click on “Register to continue” 7. The next screen thanks you for registering. Select the “Continue” box. 8. Go to the box that says “My society membership includes this journal.” 9. Enter your member number in the box under the heading “ACNM member number or Subscriber number.” Your member number can be found either on the shipping label of your print version of JMWH or on your ACNM membership card. If you are unable to determine your member number, e-mail [email protected] with your complete name, city and state. Include “JMWH online” in the e-mail subject line. You will receive a response within 24 hours. You may also call 240-485-1825 to get your membership number. If you already have established online access and are having difficulty, contact: Elsevier Help Desk Tel: (800) 654-2452 (Toll Free US & Canada) Tel: (407) 345-4000 (Outside US & Canada) Fax: (407) 363-9661 E-mail: [email protected]
Journal of Midwifery & Women’s Health • www.jmwh.org
137
INTRODUCTION Despite some reported improvement over the last decade, the rates of adolescent pregnancy in the United States remain high compared to other developed nations.1,2 Although the age of onset of sexual activity and sexual activity patterns do not differ significantly among most developed nations, there are significant differences in the acceptability and availability of information about adolescent sexuality, along with the differences in social expectations regarding pregnancy.3–5 According to the National Center for Health Statistics, the highest recorded teen pregnancy rates in the United States were in 1990, at 11.6% for women ages 15–19.6 These numbers dropped between 1990 and the year 2000.6 Even with this improvement, 8.4% of women ages 15–19 were pregnant in the year 2000.6 This translates into 852,000 teen pregnancies, which resulted in more than 478,000 live births to adolescents in this single year alone.6 Efforts to promote abstinence and encourage effective contraception usage appear to have had some effect, but the Healthy People 2010 goal of decreasing adolescent pregnancy rates to 4.6% has not been reached. Further, the female adolescent population will grow by approximately 10% within the next 5 years.7 Even with stable or declining rates of adolescent pregnancy, this may translate into a large increase in the number of actual births.7 The public health costs associated with teen pregnancy are significant and the social implications are staggering. The demographics of teen pregnancy have shifted in the
Address correspondence to Elizabeth Heavey, CNM, PhD, State University of New York College at Brockport, School of Nursing, 350 New Campus Dr., Brockport, NY 14420-2988. E-mail: [email protected]
130 © 2008 by the American College of Nurse-Midwives Issued by Elsevier Inc.
last 40 years. Today, pregnant adolescents are much more likely to be unmarried, to continue to have subsequent pregnancies into their twenties while not married, and have a greater likelihood to need extended welfare support than in decades past.5 The National Center for Health Statistics reports that in 2002, two-thirds of all adolescent births were paid for by the government, while only 14% of those births to women over the age of 30 required this financial assistance.6 Further, the economic and social burden is not felt equally among all people. There are significant racial differences in teen pregnancy and birth rates.8 In the year 2000, white teens had a pregnancy rate of 5.7%, while Hispanic teens had a rate of 13.2% and African American teens were at 15.1%.6 Adolescent birth rates in the year 2004 were highest among Hispanic adolescents at 8.2%, while African American adolescents have a birth rate of 6.3% and white adolescents have a rate of 2.7%.6 In addition, one-quarter of adolescent births are to teens who have already given birth.7 While some decline in pregnancy rates is noted among all teens, the rates among minority teens continue to be more than double those of white teens. In fact, the overall decline in adolescent pregnancy and birth rates since the 1980s is the least among Hispanic adolescents.6 Many explanations have been offered as to why this phenomenon occurs. Contraceptive decision-making may be impacted by racial identity.9 Lower socioeconomic status of minority women, combined with limited educational and employment opportunities, may impact the ability or desire to use contraception and thereby increase pregnancy rates.9 Unfortunately, it is very difficult to separate poverty effects from racial effects because of a high degree of association between these covariates. Volume 53, No. 2, March/April 2008 1526-9523/08/$34.00 • doi:10.1016/j.jmwh.2007.09.005
Poverty has a direct impact on the type of childhood experiences an adolescent may have had and becomes both a social and financial condition.10 These years are filled with significantly more work and family responsibilities and may be so chaotic that mere survival is the primary goal.10 The extensive effects of poverty may impair adolescent coping, diminish self-esteem, and foster adolescent sexual relationships, thereby impacting adolescent pregnancy rates.11 There are many studies which show a correlation between social and economic disadvantage and adolescent pregnancy.12–18 Some reports suggest that adolescents may see pregnancy as a good alternative when faced with a lack of opportunity and direction.7,19 –22 Some researchers have argued that early pregnancy and childbearing are an adaptive response to the cumulative disadvantage associated with the combination of poverty and minority status.12,23–28 One study examining African American, Hispanic, and white women found that each group had the largest number of first births during the time period when the infant mortality rate was the lowest.23 The combination of infant health advantage and additional familial support available for impoverished teens who become parents may improve the ability of these teens to minimize long-term risks associated with delaying pregnancy until after the age of 20.23 Teen pregnancy may actually be a rational decision directly associated with the limited nature of the options which are available to this group of adolescents.23 The current paradigm for sexual decision-making and teen pregnancy interventions frequently assumes that teens are getting pregnant by mistake. Educational interventions are structured around this belief, and adolescents are approached with this assumption by health care providers. Frequently, the lack of future opportunities, familial and peers’ perceptions of pregnancy, and limited alternatives are not taken into account. This may be why many adolescent pregnancy prevention programs have not been effective in decreasing pregnancy rates or birth rates.29 –35 If a significant number of adolescents desire pregnancy or are ambivalent about pregnancy, the current approach may continue to fail. This study examined
Elizabeth J. Heavey, CNM, PhD, is an epidemiologist specializing in maternal/child health and an assistant professor in the School of Nursing at SUNY Brockport. She is also completing a post-doctoral fellowship in Adolescent Medicine at the University of Rochester. Kirsten B. Moysich, PhD, Roswell Park Cancer Institute, Department of Epidemiology. Andrew Hyland, PhD, Roswell Park Cancer Institute, Department of Health Behavior. Charlotte M. Druschel, MD, PhD, New York State Department of Health, SUNY at Albany. Michael W. Sill, PhD, Roswell Park Cancer Institute, Gynecology Oncology Group.
Journal of Midwifery & Women’s Health • www.jmwh.org
the association between educational status, racial background, and adolescent pregnancy desire. METHODS Participants were all female adolescents between the ages of 12 and 19 who had a positive pregnancy test at a family planning clinic on the day of their appointment. The data were originally collected from 1999 –2004. Racial identity was self-reported by participants. Adolescents had to be 19 years of age or younger and below state income limits in order to participate in this particular family planning program. Income requirements were met if the adolescents already qualified for public health insurance or had no health insurance at the time of the visit. There were a small number of adolescents who had private health insurance and were still included in the program after meeting with a social worker at the clinic who determined their income eligibility according to the state guidelines. Those who were not eligible for participation in the state-funded program were not included in this study. The census tract in which the clinic is located has 43% of families living below the poverty line and 25% of families relying on public assistance.36 Within the five major zip code areas served by the clinic, 60%–75% of the births were covered by Medicaid, and teen pregnancy rates were approximately 20% in 2000.37 All participants were deemed eligible for free contraception available on site at the time of the visit or through public health insurance coverage. The information utilized was originally part of the recordkeeping for the New York State Family Planning Program and is the same data collected at all the reproductive health care sites participating in this program throughout the state. The information collected included questions regarding age, sex, monthly income, family size, educational level completed, student status (enrolled full time, part time, or not enrolled), number of pregnancies, number of births, and additional sources of health care. Additional clinical information was collected regarding contraception use, services required, and services provided. Demographic and interview information on the chart was originally completed by a registered nurse at the clinic during a private intake session with the patient. The nurse had completed a brief training session and all questions and counseling were available in both English and Spanish. Age was categorized as 14 years and younger, 15–17 years, and 18 –19 years because of noted clinical differences among these populations, differences cited in developmental literature, and to match the categories utilized by the U.S. Bureau of the Census.38,39 Race was self-identified by participants from a list including white, black, American Indian, Alaskan Native, Asian, or other. In addition, participants were asked if they would identify themselves as 131
Hispanic. Because the number of previous births and the number of previous pregnancies were highly correlated, this analysis utilized only the variable reflecting the number of previous pregnancies. Patients were asked about their desire for pregnancy during the same intake session in which they were told they were pregnant. Pregnancy desire was measured by the adolescent’s response to the question, “Was your pregnancy desired now, desired sooner, desired later, not desired, or unsure?” Patients who indicated their current pregnancy was desired now or desired sooner are categorized as having intended pregnancies, while those which were desired later or not desired are classified as having unintended pregnancies. Those who indicated they were unsure were classified as ambivalent. The clinical information was then completed by the provider who examined the patient. Permission to analyze the data was obtained from the director of the state program and the medical director at the clinic. The data obtained from the forms were de-identified before use and provided no opportunity to track repeat visits. The review of the records for the current study was approved by the institutional review board at the State University of New York College at Brockport. All statistical analyses were done utilizing the SAS program (version 9.0; SAS Institute, Cary, NC). Associations between demographic variables and pregnancy desire were examined utilizing the 2 test for categorical data and t tests for continuous variables. The unadjusted odds ratio was calculated for each independent variable and pregnancy desire, and those which were significant were included in the logistic regression to examine adjusted odds ratios. Those variables which had a P value of ⬍.05 were included in the final model. We hypothesized a priori that a significant number of adolescents at the clinic wanted to be pregnant now or sooner. In order to utilize data from teenagers who expressed ambivalent desire for pregnancy, our first model grouped the ambivalent teens with those who desired pregnancy. Our second model then eliminated the ambivalent teens and only those who clearly desired pregnancy or clearly did not desire pregnancy were compared. Lastly, the ambivalent teens were grouped with those who did not desire pregnancy. Logistical regression analysis was utilized to examine the groups of teenagers defined by pregnancy desire which was the dependent variable. These models assume independence in visits, because we had no means to determine the proportion of repeat visits. The sample was stratified by race and student status, because we hypothesized a priori that pregnancy desire would differ in these subgroups. RESULTS We retrospectively reviewed 349 charts. Ten charts were excluded because of incomplete data. We also excluded 132
four charts from adolescents not identified in the three main racial groups. The final analysis was limited to 335 charts. The sample population was largely composed of teens from minority racial groups, and more than half of the population received public insurance, with more than 44% uninsured. Sixty-two percent of the sample reported using no contraception, and 32% were no longer in school. Fifty-four percent of the adolescents were pregnant for the first time, while 33% were for the second time and more than 12% were pregnant two or more times previously. The mean age was 17.4 years. Sample population characteristics are summarized in Table 1. Adolescents responded to a question regarding pregnancy desire with 92 indicating a desired pregnancy (27%), 124 being ambivalent about their pregnancy (37%), and 119 reporting an undesired pregnancy (36%). The first analysis grouped those who were ambivalent with those who desired pregnancy, and found that compared to African American adolescents, Hispanic teenagers were twice as likely to desire or be ambivalent about pregnancy (Table 2). There was no significant difference between white adolescents and African American adolescents in terms of pregnancy desire/ambivalence (Table 2). In addition, adolescents who were not in school were almost twice as likely to desire or be ambivalent about pregnancy as those who were in school full time (Table 2). The Hispanic adolescents who were not in school were six times as likely to either desire or be ambivalent about pregnancy as the African American adolescents who were in school full time. After stratifying by race, the Hispanic adolescents who were not in school were three times as likely to desire or be ambivalent about pregnancy as those who were in school full time. There were no significant relationships by student status found for white or African American adolescents. We also examined the impact of race after stratifying the adolescents into either full-time student status, parttime status, or not in school, and found significant relationships only for those who were not in school. Among adolescents not in school, Hispanic teens are four times as likely to desire pregnancy compared to African American teens, and there was no significant difference between white teens and African American teens. In addition, interaction terms for student status and number of previous pregnancies, number of previous pregnancies and age category, and student status and race were tested and were not significant. The analysis was repeated, eliminating the ambivalent group and comparing only those who desired pregnancy and those who did not desire pregnancy. The observed associations were still significant and increased in magnitude (Table 3). Hispanic clients were two and a half times as likely to desire pregnancy compared to the African American teenagers, and white adolescents were not significantly more likely to desire pregnancy than African American adolescents (Table 3). The young Volume 53, No. 2, March/April 2008
Table 1. Characteristics of 335 Teenagers with a Positive Pregnancy Test (1999 –2004) Variable
Desired Pregnancy [n ⴝ 92; n (%)]
Ambivalent† [n ⴝ 124; n (%)]
Undesired Pregnancy [n ⴝ 119, n (%)]
0 (0) 41 (44.6) 51 (55.4)
4 (3.2) 57 (46.0) 63 (50.8)
7 (5.9) 48 (40.3) 64 (53.8)
8 (8.7) 47 (51.1) 37 (40.2)
9 (7.3) 78 (62.9) 37 (29.9)
9 (7.6) 87 (73.1) 23 (19.3)
41 (44.6) 5 (5.4) 46 (50.0)
66 (53.2) 2 (1.6) 56 (45.2)
72 (60.5) 3 (2.5) 44 (37.0)
49 (53.3) 1 (1.1) 42 (45.7)
83 (67.5) 3 (2.4) 37 (30.1)
88 (74.0) 4 (3.4) 27 (22.7)
52 (56.5) 26 (28.3) 7 (7.6) 7 (7.6)
73 (58.9) 37 (29.8) 10 (8.1) 4 (3.2)
59 (49.6) 46 (38.7) 11 (9.2) 3 (2.5)
Age Category (Years) ⱕ14 years 15–17 years 18–19 years Racial Background White African-American Hispanic Type of Insurance Public Private None Student Status Full-time Part-time None Previous Pregnancies 0 1 2 ⱖ3
P* .19
.02
.12
.01
.32
*Significance tested using 2 test. † Missing student status data for 1 participant.
women who were not in school were more than twice as likely to desire pregnancy (Table 3). The Hispanic adolescents who were not in school were almost 12 times as likely to desire pregnancy when compared to the African American adolescents who were in school fulltime. After stratification by student status, the Hispanic adolescents who were not in school were eight times as likely as those who were in school to desire pregnancy. There were no significant relationships found for the white or African American adolescents. We again stratified by student status into part-time, not in school, and full-time, and only found significant relationships among adolescents who were not in school. Hispanic teens who were not in school were 17 times as likely to desire pregnancy compared to African Ameri-
can teens who were not enrolled in school. There was no significant difference between white teens and African American teens, but there was a significant association with age. For adolescents who were not in school, each additional year of age was associated with a twofold increase in desire for pregnancy. We conducted further sensitivity analyses by combining the ambivalent group with the adolescents who did not desire pregnancy and compared this combined group to those who did desire pregnancy. Again, the Hispanic adolescents were almost twice as likely to desire pregnancy compared to the African American adolescents (Table 4). The pregnancy desire among white adolescents was not significantly different compared to African American adolescents (Table 4). Adolescents who were
Table 2. Association Between Pregnancy Desire and Sample Characteristics in Teenagers with a Positive Pregnancy Test* (N ⫽ 335) Variable
Desired and Ambivalent about Pregnancy [n ⴝ 216†; n (%)]
Undesired Pregnancy [n ⴝ 119; n (%)]
OR (95% CI)
Adjusted OR (95% CI)
Hispanic White African American Part-time student Not in school Full-time student
74 (24.9) 17 (9.0) 125 (66.1) 4 (1.9) 79 (36.7) 132 (61.4)
23 (19.3) 9 (7.6) 87 (73.1) 4 (3.4) 27 (22.7) 88 (73.9)
2.26 (1.31–3.88) 1.33 (0.56–2.31) Reference group 0.67 (0.16–2.73) 1.95 (1.17–3.26) Reference group
2.11 (1.22–3.65) 1.11 (0.46–2.67) Reference group 0.67 (0.16–2.80) 1.83 (1.08–3.09) Reference group
CI ⫽ confidence interval; OR ⫽ odds ratio. *Number of previous pregnancies, age, type of health insurance, and other sources of healthcare were tested and were not significant. †
Missing student status data for 1 participant.
Journal of Midwifery & Women’s Health • www.jmwh.org
133
Table 3. Association Between Pregnancy Desire and Sample Characteristics in Teenagers with a Positive Pregnancy Test* (N ⫽ 211) Variable
Desired Pregnancy [n ⴝ 92; n (%)]
Undesired Pregnancy [n ⴝ 119; n (%)]
OR (95% CI)
Adjusted OR (95% CI)
Hispanic White African American Part-time student Not in school Full-time student
37 (40.2) 8 (8.7) 47 (51.1) 1 (1.1) 42 (45.7) 49 (53.3)
23 (19.3) 9 (7.6) 87 (73.1) 4 (3.4) 27 (22.7) 88 (73.9)
2.98 (1.59–5.59) 1.65 (0.60–4.55) Reference group 0.45 (0.05–4.13) 2.80 (1.54–5.07) Reference group
2.54 (1.33–4.87) 1.34 (0.47–3.83) Reference group 0.46 (0.05–4.36) 2.42 (1.31–4.48) Reference group
CI ⫽ confidence interval; OR ⫽ odds ratio. *Number of previous pregnancies, age, type of health insurance, and other sources of healthcare were tested and were not significant.
not in school were twice as likely to desire pregnancy as those who were in school (Table 4). After stratification by race and student status, there were no significant relationships found among the variables for any of the racial groups. DISCUSSION This study clearly shows racial and educational differences in pregnancy desire among low-income adolescent women. However, several limitations need to be addressed in the interpretation of these findings. We lacked a measure of the strength of cultural identity and had no way to further divide categories of racial/cultural background. There may be distinct differences among these broad racial categories which we were unable to detect given the available data. The sample was a convenience sample drawn from the clients utilizing a community-based obstetric/gynecology clinic. The sample is drawn largely from the community surrounding the clinic. Thus the findings may not be generalizable to this population or to populations that differ significantly from this one, including populations with higher socioeconomic status and improved access to future economic or educational opportunities. Because the study relied on previously collected data, we were unable to include all relevant variables, including some of the psychological predictors, family struc-
ture, drug use, and partner’s perceptions of contraception and pregnancy. Previous studies have reported an association between pregnancy and lack of self-esteem, drug use, and partners who desire pregnancy.40 Some of these factors may be confounding variables or even the actual variable of interest. For example, student status may simply be a reflection of the adolescent’s self-esteem, in which case our study would be reporting an association between pregnancy desire and a confounding variable, not the actual variable with which there is an association. Because of the privacy limitations imposed by the clinic administration, it was not possible to track those individuals who had repeat measures in the dataset. The lack of independence for some of the observations may limit the accuracy of the inferences drawn from these data. The nature of the data meant relying on adolescents’ self-reports for behavioral variables. Self-reports of sensitive information are frequently inaccurate, and underestimate the occurrence of socially sanctioned behavior or experiences. Underestimates of pregnancy desire are expected, and more accurate data would be expected to increase the associations found in this study. Alternatively, pregnancy desire may be the same across all groups, but the differences in ethnic norms may be related to differences in the acceptance and acknowledgement of pregnancy desire.
Table 4. Association Between Pregnancy Desire and Sample Characteristics in Teenagers with a Positive Pregnancy Test* (N ⫽ 335) Variable
Desired Pregnancy [n ⴝ 92; n (%)]
Ambivalent and Undesired Pregnancy [n ⴝ 243†; n (%)]
OR (95% CI)
Adjusted OR (95% CI)
Hispanic White African American Part-time student Not in school Full-time student
37 (40.2) 8 (8.7) 47 (51.1) 1 (1.1) 42 (45.7) 49 (53.3)
60 (24.7) 18 (7.4) 165 (67.9) 7 (2.9) 64 (26.4) 171 (70.7)
2.21 (1.30–3.74) 1.54 (0.63–3.76) Reference group 0.50 (0.06–4.15) 2.290 (1.39–3.79) Reference group
1.97 (1.15–3.35) 1.24 (0.49–3.10) Reference group 0.50 (0.06–4.24) 2.14 (1.28–3.58) Reference group
CI ⫽ confidence interval; OR ⫽ odds ratio. *Number of previous pregnancies, age, type of health insurance, and other sources of healthcare were tested and were not significant. †
Missing student status data for one participant.
134
Volume 53, No. 2, March/April 2008
Despite these limitations, there are several noteworthy strengths associated with the current study. Because socioeconomic status is a significant contributor to adolescent pregnancy desire, any study on this topic must control for socioeconomic variability in order to examine other contributing variables.41 In this study, the combination of income requirements for participation in this program and the fact that the majority of the participants resided in the high-poverty zip codes resulted in a fairly well controlled sample of individuals with limited socioeconomic means, living in communities with limited socioeconomic resources, which minimized residual confounding from socioeconomic status. In addition, participants were asked about their desire for pregnancy at the same visit as their pregnancy test. When asked about pregnancy desire after their pregnancy, adult women are more likely to indicate that pregnancy occurred at the right time, even if this was not true at the time of pregnancy confirmation.42 However, there is no research indicating how an adolescent’s perspective might change, and we postulated that the reverse may be true of female adolescents. The reality of childbirth, loss of peer connections and romantic relationships, responsibilities of parenthood, and the constant needs of an infant may make an adolescent more likely to look back and indicate a pregnancy was not desired when it actually was at the time of conception. Our study avoided a retrospective bias which could have affected the desire for pregnancy negatively or positively, and provided a better measure of pregnancy desire. This study includes several points of interest which require further research. Hispanic teenagers are much more likely to desire pregnancy, which is a significant risk factor for adolescent pregnancy. Gender roles within Hispanic households may encourage pregnancy as a way for women to find fulfillment.43 Early and frequent pregnancies may be seen in a positive light.43 In addition, female independence is not a traditional value, and continued education may be seen as incompatible with pregnancy and childbearing.43 Lack of educational enrollment is also significantly associated with pregnancy desire, and when both risk factors are present, there is a much higher likelihood that an adolescent will want to become pregnant. These findings suggest that racial norms and educational achievement have a large impact on adolescent pregnancy risk. Although there are many reports of Hispanic values, there are no previous studies which address how they may impact adolescent pregnancy desire. Even among adult women, Hispanic women are frequently not considered as a separate demographic population and disparities in reproductive health care are not well documented. Most of the previous studies on adolescent pregnancy desire have focused exclusively on African American teens, differences between white and African Journal of Midwifery & Women’s Health • www.jmwh.org
American teens, or have not examined the role that race may play.14,19,38,44 – 46 These findings suggest a need to develop pregnancy prevention interventions specifically designed to address adolescents of all races who desire pregnancy. Such interventions have been sorely neglected in the literature. It may be that this population is amenable to at least delaying pregnancy and this may be an improvement over very early childbearing. CONCLUSION The clinical implications of this study include the need for adolescent family planning programs to include an assessment of pregnancy desire. Because peers and community have such a significant impact in the teen years, clinicians need to explore how teens feel about pregnancy within this context. It is important to consider community social structure and familial norms and include this information in the subsequent plan of care. The single screening question used in this study may help identify adolescents who desire pregnancy. Simply asking about pregnancy desire allows a clinician to identify adolescents who may be at higher risk, including those who are ambivalent. Behaviors associated with ambivalence and desiring pregnancy are the same and only the group of women who do not want to be pregnant behave differently clinically.14,22,38 Ignoring the feelings of teenagers regarding pregnancy might result in missing an opportunity to address a key factor in pregnancy risk. This study identified adolescent pregnancy desire and pregnancy ambivalence as a much more common occurrence than many health care professionals believe. This issue is more complex than a lack of information or a lack of economic resources. Although contraception education and free services are helpful for those wishing to avoid pregnancy, they may not be effective interventions for young women who do not wish to avoid pregnancy. For adolescents identified as desiring pregnancy, further interventions are needed to address the complex reasons why. The solutions in this case may not be as simple as a straightforward talk about contraception, but they need to be identified before the adolescent pregnancy goals of Healthy People 2010 can become a reality. Supported by the Mark Diamond Research Fund, The Graduate Student Association at SUNY Buffalo, and Roswell Park Cancer Institute.
REFERENCES 1. Singh S, Wulf D, Samara R, Cuca YP. Gender differences in the timing of first intercourse. Int Fam Plann Perspect 2000;26: 21– 8. 2. Moore M. Adolescent pregnancy rates in three European countries. J Obstet Gynecol Neonatal Nurs 2000;29:355– 62.
135
3. Feldmann J, Middleman A. Adolescent sexuality and sexual behavior. Curr Opin Obstet Gynecol 2002;14:489 –93.
quences of teen childbearing reconsidered. Q J Econ 1992; 107:1187–1214.
4. Kartoz C. New options for teen pregnancy prevention. Am J Matern Child Nurs 2004;29:30 –5.
25. Geronimus A. Commentary: Weathering Chicago. Int J Epidemiol 2003;32:90 –1.
5. Coley R, Chase-Lansdale P. Adolescent pregnancy and parenthood. Am Psychol 1998;53:152– 66.
26. Satin A, Leveno K, Sherman L, Reedy N, Lowe T, McIntire D. Maternal youth and pregnancy outcomes. Am J Obstet Gynecol 1994;171:184 –7.
6. National Center for Health Statistics Web site. Available from: www.cdc.gov [Accessed March 17, 2006]. 7. Klein J. Adolescent pregnancy: Current trends and issues. Am Acad Pediatr 2005;116:281– 6. 8. Elster A, Jarosik J, VanGeest J, Fleming M. Racial and ethnic disparities in health care for adolescents: A systematic review of the literature. Arch Pediatr Adolesc Med 2003;157:867– 74. 9. Raine T, Harper C, Paukku M, Darney P. Race, adolescent contraceptive choice, and pregnancy at presentation to a family planning clinic. Obstet Gynecol 2002;99:241–7. 10. Smith L. Psychotherapy, classism and the poor. Am Psychol 2005;60:687–96. 11. Culhane J, Elo I. Neighborhood context and reproductive health. Am J Obstet Gynecol 2005;192:S22–9. 12. Lu M, Halfon N. Racial and ethnic disparities in birth outcomes. Matern Child Health J 2003;7:13–30. 13. Jewell D, Tacchi J, Donovan J. Teenage pregnancy: Whose problem is it? Fam Pract 2000;17:522– 8. 14. Zabin L, Astone N, Emerson M. Do adolescents want babies? J Res Adolesc 1993;3:67– 86. 15. SmithBattle L. The vulnerabilities of teenage mothers. Adv Nurs Sci 2000;23:29 – 40. 16. Duncan G, Hoffman S. Welfare benefits, economic opportunities, and out-of-wedlock births among black teenage girls. Demography 1990;4:519 –35. 17. Brewster K, Billy J, Grady W. Social context and adolescent behavior. Social Forces 1993;71:713– 40. 18. Cherlin A, Kiernan K, Chase-Lansdale P. Parental divorce in childhood and demographic outcomes in young adulthood. Demography 1995;32:1–10. 19. Montgomery K. Planned adolescent pregnancy: What they want. J Pediatr Health Care 2002;16:282–9. 20. Hellerstedt W, Fee R, McNeely C, Sieving R, Shew M, Resnick M. Pregnancy feelings among adolescents awaiting pregnancy test results. Public Health Rep 2001;116:180 –93.
27. Rich-Edwards J, Buka S, Brennan R, Earls F. Diverging associations of maternal age with low birthweight for black and white mothers. Int J Epidemiol 2003;32:83–90. 28. Rauh V, Andrews H, Garfinkel R. The contribution of maternal age to racial disparities in birthweight. Am J Public Health 2001;91:1815–24. 29. DiCenso A, Guyatt G, Willan A, Griffith L. Interventions to reduce unintended pregnancies among adolescents. Br Med J 2002;324:1–17. 30. Maynard R. Teenage childbearing and welfare reform. Child Youth Serv Rev 1995;17:309 –32. 31. Bennett S, Assefi N. School-based teenage pregnancy prevention programs: A systematic review of randomized controlled trials. J Adolesc Health 2005;36:72– 81. 32. Kirby D. Emerging answers: Research findings on programs to reduce teen pregnancy. Washington, DC: National Campaign to Prevent Teen Pregnancy, 2001. 33. Robin L, Dittus P, Whitaker D, Crosby R, Ethier K, Mezoff J, et al. Behavioral interventions to reduce incidence of HIV, STD and pregnancy among adolescents: A decade in review. J Adolesc Health 2004;34:3–26. 34. Guyatt G, DiCenso A, Farewell V, Willan A, Grifith L. Randomized trials versus observational studies in adolescent pregnancy prevention. J Clin Epidemiol 2000;53:167–74. 35. Nicoletti A. Perspectives on pediatric and adolescent gynecology from the allied health professional. Teens, tattoos, and body piercings. J Pediatr Adolesc Gynecol 2004;17:215– 6. 36. Johnson W. State of the City Mayorial Speech. March 1, 2004. Available from: www.cityofrochester.gov [Accessed February 23, 2005]. 37. New York State Health Department. As reported in 2004 Grant Application, Jordan Health Center, Rochester, NY. 38. Lindemann C, Scott W. Wanted and unwanted pregnancy in early adolescence. J Early Adolesc 1981;1:185–93. 39. Davis A. The role of hormonal contraceptives. Am J Obstet Gynecol 1994;170(5 Suppl):1581–5.
21. Davies S, DiClemente R, Wingood G, Harrington K, Crosby R, Sionean C. Pregnancy desire among disadvantaged African American adolescent females. Am J Health Behav 2003;27:55– 62.
40. Modrcin-Talbott M, Pullen L, Zandstra K, Ehrenberger H, Muenchen B. A study of self-esteem among well adolescents. Issues Compr Pediatr Nurs 1998;21:229 – 41.
22. Crosby R, DiClemente R, Wingood G, Harrington K, Davies S, Hook E, et al. Psychosocial predictors of pregnancy among low-income African-American adolescent females. J Pediatr Adolesc Gynecol 2002;15:293–9.
41. Young T, Turner J, Denny G, Young M. Examining external and internal poverty as antecedents of teen pregnancy. Am J Health Behav 2004;28:361–73.
23. Geronimus A. The weathering hypothesis and the health of African-American women and infants. Ethn Dis 1992;2:207–21.
42. Sable M. Pregnancy intentions may not be a useful measure for research on maternal and child health outcomes. Fam Plann Perspect 1999;31:249 –50.
24. Geronimus A, Korenman S. The socioeconomic conse-
43. Russell S, Lee F. Practitioners’ perspectives on effective
136
Volume 53, No. 2, March/April 2008
practices for Hispanic teenage pregnancy prevention. Perspect Sexual Reprod Health 2004;36:1–17.
Harrington K, et al. Predictors of inconsistent contraceptive use among adolescent girls. J Adolesc Health 2006;39:43–9.
44. Bloom K. Pregnancy wantedness in adolescents presenting for pregnancy testing. Am J Matern Child Nurs 1999;24:296 –300.
46. Davis S, DiClemente R, Wingood G, Person S, Crosby R, Harrington K, et al. Relationship characteristics and sexual practices of African American adolescent girls who desire pregnancy. Health Educ Behav 2004;31:85S–96S.
45. Davies S, DiClemente R, Wingood G, Person S, Dix E,
ACCESS THE JOURNAL OF MIDWIFERY & WOMEN’S HEALTH ONLINE ARTICLES Articles published in the Journal of Midwifery & Women’s Health are available online to ACNM members and JMWH subscribers. To access your online subscription for the first time: 1. Type in www.jmwh.com/claim 2. Go to the “New User” box and enter your last name and email address 3. Click on “Sign in” 4. At the registration screen, select a user name and password and fill in requested details 5. Read the registered user agreement at the end of the form and click box 6. Click on “Register to continue” 7. The next screen thanks you for registering. Select the “Continue” box. 8. Go to the box that says “My society membership includes this journal.” 9. Enter your member number in the box under the heading “ACNM member number or Subscriber number.” Your member number can be found either on the shipping label of your print version of JMWH or on your ACNM membership card. If you are unable to determine your member number, e-mail [email protected] with your complete name, city and state. Include “JMWH online” in the e-mail subject line. You will receive a response within 24 hours. You may also call 240-485-1825 to get your membership number. If you already have established online access and are having difficulty, contact: Elsevier Help Desk Tel: (800) 654-2452 (Toll Free US & Canada) Tel: (407) 345-4000 (Outside US & Canada) Fax: (407) 363-9661 E-mail: [email protected]
Journal of Midwifery & Women’s Health • www.jmwh.org
137