- Email: [email protected]
Predictors of Repeat Pregnancy in a Program for Pregnant Teens
J Pediatr Adolesc Gynecol (2003) 16:77–81
Original Studies Predictors of Repeat Pregnancy in a Program for Pregnant Teens Mark A. Pfitzner, MD, MPH, Charles Hoff, PhD, and Kathleen McElligott, MD Department of Pediatrics, University of Utah, Salt Lake City, Utah, USA
Abstract. Study Objective: To describe repeat pregnancy among adolescents and to compare those who experienced a repeat pregnancy to those who did not. Design, Setting, Participants: Retrospective case control of all adolescents who entered and exited the Teen Mother & Child Program, a multidisciplinary clinic for pregnant and parenting teens and their children, between 1985 and 2000. Main Outcome Measure: Repeat pregnancy. Results: Over the 16-yr study period, 1838 teens entered and exited the program with the mean time in the program of 1.9 yrs. 194 (10.6%) teens went on to have a second pregnancy, and 4 of those had a third pregnancy. Ninetytwo percent (n ⫽ 175) of these pregnancies resulted in a live birth. The mean interval between delivery of the initial child and delivery of the second child was 21.8 months (range 5.4–53.2). Teens repeating (R) were compared to those who did not repeat (NR). R were younger at entry and older at exit from the program (both P ⬍ 0.0001). R were more likely to be Hispanic (P ⫽ 0.009) or have a partner who was Hispanic (P ⫽ 0.02). R were more likely to have experienced a poor initial pregnancy outcome (miscarriage, stillbirth) (P ⫽ 0.03). R were more likely than NR to be in a stable relationships with the father of the baby (i.e., married or engaged) (P ⫽ 0.03). Self-report of physical and sexual abuse, depression, and substance abuse were common, but did not differ between R and NR. However, R were more likely than NR to self-report suicide gestures/attempts and to have a significant psychiatric history (P ⫽ 0.01, P ⫽ 0.004). Only 24% of R and 26% of NR had completed high school by the time of exit from the program. Conclusions: Ten percent of adolescents served by comprehensive multidisciplinary teen pregnancy program experienced a repeat pregnancy while in the program. Compared to nonrepeaters, adolescents who experienced a repeat pregnancy were younger, were more frequently Hispanic, and were more likely to be in a stable relationship with the
Address reprint requests to: Mark A. Pfitzner, MD, MPH, Department of Pediatrics, 30 N. 1900 East, Rm 1B386, Salt Lake City, UT 84132; E-mail: [email protected]
쑖 2003 North American Society for Pediatric and Adolescent Gynecology Published by Elsevier Science Inc.
baby’s father. Many of the program’s clients have psychosocial factors reported in the literature to be associated with repeat pregnancy. A history of suicide gestures/attempts and a significant psychiatric history were more common in those who had a repeat pregnancy. Disappointingly, only about 25% of the adolescents completed high school by the time they exited the program whether they experienced a repeat pregnancy or not. Although directing interventions (e.g., mental health services, counseling those who miscarried) to teens who appear to be at highest risk for a repeat pregnancy may decrease their risk of repeating, all teens in our program would likely benefit from such services.
Key Words. Adolescent pregnancy—Repeat pregnancy
Introduction Historically, repeat pregnancy rates among adolescents have been high,1 with 30% becoming pregnant in their first postpartum year and another 25% to 50% becoming pregnant in the second year. The consequences of repeat pregnancy are significant for the adolescent and society. Adolescents who experience a repeat pregnancy are more likely to leave school without graduating, to be unemployed, and to be welfare dependent.2 Programs specifically designed to address this have had some limited success.3 Previous studies have found several factors associated with repeat pregnancy among teenagers. These include inconsistent contraceptive use,4 lower educational achievement,5 and a nonsupportive family environment.6 Maynard reported that one-half of teenagers receiving public assistance became pregnant for a second time within 24 months of giving birth to their first child.7 Service programs exist throughout the country to provide care and support to pregnant and parenting 1083-3188/03/$22.00 doi:10.1016/S1083-3188(03)00011-1
78
Pfitzner, Hoff, and McElligott: Predictors of Repeat Teen Pregnancy
teens. One such program is the University of Utah Teen Mother and Child Program (TMCP). TMCP is a multidisciplinary program that has provided care to pregnant and parenting teens since the early 1980’s. TMCP provides comprehensive care directed to both the unique medical and psychosocial issues of a high-risk adolescent population. One objective of the program has been to provide services intended to reduce the risk of a repeat pregnancy while the adolescent is under our care. To better understand the phenomenon of early repeat pregnancy in our population, we designed a study to determine the factors associated with a teenager’s having a second pregnancy while still in our program. Since assisting TMCP clients to complete high school while in our program is a goal, we wanted to determine the effect of a second pregnancy on achieving that positive outcome.
Methods The study group included all clients who entered and exited the program between 1985 and 2000. Teens are eligible for care in the program until they reach 19 years of age or their child(ren) become 2 years old, whichever occurs later. Some voluntarily exit the program while still eligible for services. Since 1985, TMCP has maintained a detailed database containing demographic, psychosocial, and outcome data for the teen and her child. Demographic and medical information comprising the database is obtained from the medical record. In addition, each client has an intake interview with the program’s social worker that focuses on psychosocial issues. Data from both sources is then entered into the database in a standardized manner. We used data associated with the initial pregnancy for comparison between those who experienced a repeat pregnancy (R) and those who did not (NR). Information regarding first pregnancy was not available for 35 individuals who experienced a repeat pregnancy. These teens had either been in the program prior to the database initiation or joined TMCP following a first pregnancy for which they received care at another facility. Typically these teens had been referred to TMCP by one of several community agencies. We included data for the second pregnancy only in the descriptive analysis of the repeat population but not for the comparisons. We compared those who had a repeat pregnancy to those who did not on all the variables in the database using ANOVA or chi-square as appropriate. Statview for Windows (Version 5.0.1, SAS Institute, Inc, Cary, NC, 1992–1998) was used for all calculations. The study was approved by the University of Utah Health Sciences Center’s Institutional Review Board.
Fig. 1. Mean birth interval by year of entry into the program.
Results The results are based on data obtained from all 1838 teens who entered the program between January 1985 and May 2000, and exited the program between February 1988 and December 2000. Mean time in the program was 1.9 ⫾ 0.1 yrs. Forty-six percent (n ⫽ 853) remained in the program through their entire eligibility (i.e., they did not exit prematurely). The 1838 pregnancies resulted in 1734 singleton births and 7 sets of twins; 97 resulted in either a stillbirth or a miscarriage. Of the 1838 teens enrolled in the program, 194 (10.6%) had a subsequent pregnancy and 4 had a third (i.e., second repeat). Ninety-two percent (n = 175) of these second pregnancies resulted in a birth; 2 were twin births. All second repeats were single live births. Overall, the average time between birth or miscarriage of the initial child and birth or miscarriage of the second child was 21.8 ⫾ 1.7 months (range 5.4–53.2) This interval remained constant over the period of the study (Fig. 1). Younger teens who repeated had a longer interval between pregnancies than their older peers (P ⬍ 0.0001) (Fig. 2).
Fig. 2. Linear regression of repeat birth interval by maternal entry age.
Pfitzner, Hoff, and McElligott: Predictors of Repeat Teen Pregnancy
79
Table 1. Analysis of Variance between Repeaters and Nonrepeaters for Select Continuous Variables VARIABLE (range)
REPEATERS mean (SE)
Maternal age at entry (12.6–19.7 yrs) Maternal age at delivery (13.0–19.6 yrs) Maternal age at exit (14.0–22.6 yrs) Paternal age at entry (12–63 yrs) Paternal maternal age difference (⫺3.7–46.6 yrs) Time known father of baby (0–132 mo) Time in program (0–6.5 yrs) Gestational age when prenatal care began (4–40 wks) Infant’s birth weight (530–4734 grams) Last grade completed (0–12) Months out of school (0–70)
15.9 (0.08) 16.3 (0.08) 19.5 (0.10) 18.8 (0.25) 2.9 (0.25) 18.1 (2.36) 3.6 (0.09) 15.5 (0.74) 3128 (50.6) 9.0 (0.17) 5.3 (1.14)
The results of the comparisons between teens who experienced a repeat pregnancy (R) and those who did not (NR) are shown in Tables 1 and 2. R were younger at entry and delivery, spent more time in the program, and were older when exiting. Consistent with their younger age, R averaged 0.5 fewer school grades completed compared to NR. R were more likely to have a significant psychiatric history and to have more frequently contemplated or attempted suicide. R and NR had similarly high rates of reported physical and sexual abuse and depression. R were less likely than NR to have placed their child for adoption but more likely to have experienced an unfavorable outcome with their first pregnancy. R were more likely to be in a committed relationship (i.e., living together or married) and to be Hispanic or to have a Hispanic partner than NR. At the time of exiting the program, R and NR had similar rates of either being currently enrolled or having graduated from high school.
Discussion Teen pregnancy rates are decreasing nationally, but 22% of births to teens are repeat births.8 Although the overall decline is encouraging, the absolute number of adolescent females of childbearing age is increasing by 5% to 10% per year; thus both initial and repeat pregnancies among adolescents continues to remain a significant problem.1 Our program had a repeat pregnancy rate of 10.6% during the 15-yr study period, a rate that is similar to other established teen pregnancy programs providing care to adolescent mothers and one that is lower than the national average.1,6,9 Several psychosocial and lifestyle factors are commonly assumed to be associated with an increased rate of early repeat pregnancy among teenagers. We anticipated that those adolescents in our program who experienced a repeat pregnancy would have higher rates of these when compared to teens who did not have a
NONREPEATERS mean (SE)
ANOVA P value
16.6 (0.03) 17.0 (0.03) 18.4 (0.03) 19.1 (0.09) 2.6 (0.09) 15.7 (0.49) 1.8 (0.03) 15.6 (0.19) 3171 (14.1) 9.5 (0.05) 8.1 (0.40)
⬍0.0001 ⬍0.0001 ⬍0.0001 NS NS NS ⬍0.0001 NS NS 0.005 NS
repeat pregnancy within two years of the first. However, with few exceptions, the rates of these factors were similar among the two groups. Both groups reported high rates of physical and sexual abuse, depression, and alcohol, tobacco, and illicit drug use. Other investigators have reported that sexual abuse during childhood is associated with younger age at first coitus and pregnancy.10 As many as 50% to 60% of pregnant adolescents report physical or sexual abuse.11 We did find that teens who had serious mental health problems including a history of suicidal thought, gestures, or attempts were more likely to have a repeat pregnancy. Other investigators have found that low educational achievement is a strong predictor of repeat pregnancy.12 In our study, those who repeated were enrolled in school or had graduated from school at the same rate as their nonrepeated peers. Disappointingly, only 25% of the entire population had graduated or were enrolled in school when they exited the program. Differences between repeaters and nonrepeaters were found when we examined the teen’s ethnicity and her relationship with the father of the baby. Repeat pregnancies were more common in Latinos and among those who were married or living with the baby’s father. Studies from the 1960’s and 1970’s also reported marriage as being a risk factor for repeat pregnancies.13,14 Teens whose first pregnancy ended in a miscarriage or abortion were more likely to have a second pregnancy. Placement for adoption was less common among repeaters. Repeat pregnancies were spaced nearly 2 yrs apart. Despite changes in contraceptive practices during the study years (e.g., the advent of longer-acting contraceptives like Norplant and Depo-Provera, the social marketing of condom usage, and “safe sex” campaigns with the advent of HIV) birth interval did not change appreciably in our population. Clearly this study has limitations. While the data was collected using standardized collection forms, changes in personnel over the years could have led
80
Pfitzner, Hoff, and McElligott: Predictors of Repeat Teen Pregnancy
Table 2. Chi-Square Analysis Comparing Repeaters to Nonrepeaters
VARIABLE Physical Abuse Yes No Unknown Sexual Abuse Yes No Unknown Depression Yes No Unknown Suicidality Ideation/gesture/attempt None Significant Psychiatric Historya Yes No Unknown Alcohol Useb Experiment/past or current use None Tobacco Useb Experiment/past or current use None Illicit Drug Useb Experiment/past or current use None Parent a Pregnant Teen Yes No Planned Pregnancy Yes No Ambivalent Placed Child for Adoption Yes No School Attendance at Entry Enrolled/graduated Not attending Maternal Ethnicity Non-Hispanic white Hispanic Other Paternal Ethnicity Non-Hispanic white Hispanic Other Relationship at Conception Married/Cohabitating Dating regularly Other Pregnancy Outcome Live birth Miscarriage/Abortion Educational Status at Exit Enrolled/graduated Not attending
χ2 P value
REPEATERS n (%)
NONREPEATERS n (%)
42 (30) 97 (69) 2 (1)
482 (31) 1062 (68) 16 (1)
NS
55 (39) 83 (59) 3 (2)
548 (35) 996 (64) 17 (1)
NS
78 (55) 61 (43) 2 (2)
735 (47) 818 (52) 8 (1)
NS
101 (72) 40 (28)
949 (61) 610 (39)
27 (19) 109 (77) 5 (4)
257 (16) 1290 (83) 12 (1)
45 (74) 16 (26)
689 (72) 272 (28)
NS
39 (64) 22 (36)
590 (61) 372 (39)
NS
36 (59) 25 (41)
546 (57) 413 (43)
NS
15 (10) 140 (90)
142 (9) 1504 (91)
NS
28 (20) 90 (64) 23 (16)
213 (14) 1083 (70) 262 (16)
1 (1) 154 (99)
83 (5) 1565 (95)
0.01
51 (49) 53 (51)
739 (54) 619 (46)
NS
101 (65) 35 (23) 19 (12)
1265 (76) 259 (16) 133 (8)
0.009
80 (52) 50 (32) 25 (16)
1030 (63) 396 (24) 217 (13)
0.02
92 (60) 44 (29) 17 (11)
828 (51) 645 (40) 154 (9)
0.03
145 (94) 10 (6)
1596 (97) 52 (3)
0.03
20 (26) 56 (74)
183 (24) 570 (76)
NS
0.01
0.004
NS
Pfitzner, Hoff, and McElligott: Predictors of Repeat Teen Pregnancy
81
Table 2. (Continued)
VARIABLE Relationship which Father of Baby at Exit Married/cohabitating Dating No relationship/seeing child only Exit Reason Moved/transferred care Not eligible Lost to follow-up Payer Source Insurance Self pay Medicaid a
χ2 P value
REPEATERS n (%)
NONREPEATERS n (%)
74 (48) 10 (7) 68 (45)
679 (42) 202 (13) 730 (45)
0.06
26 (18) 97 (66) 23 (16)
440 (30) 756 (53) 242 (17)
0.002
15 (10) 14 (9) 123 (81)
234 (14) 160 (10) 1224 (76)
NS
Defined as having ever had inpatient or outpatient psychiatric treatment. Variable added to database in 1993.
b
to biases in recording. As much of the psychosocial variables rely on patient self-report, perception by the patient regarding these variables may change over time as society changes its perception of them (e.g., depression and mental health issues). Yet there was not a large degree of variability in these variables over time. One large limitation is our inability to know the outcome of any teen who left the program; 54% left prematurely. Although we have a large number who remained, we do not know if those who left early repeated. This is a well-recognized limitation of programs such as ours as few have had rigorous efficacy studies.15 All teens in this program represent an extremely high-risk population. Repeaters were not adversely affected by their pregnancy when compared to their nonrepeating peers. Perhaps this was due to the longer birth interval seen in this population. However, markers of economic disparities were not examined in this study. Although targeting interventions at those teens at greatest risk would seem advantageous (e.g., by improving mental health services or providing close follow-up to those teens who miscarried or aborted), both repeaters and nonrepeaters in this high-risk population would benefit from those services. Targeting interventions solely to the highest-risk groups in this risky population may only serve to increase repeat pregnancies among those not targeted. Practitioners should realize that pregnancy is but one risk marker for a teen; others may be present. With whom to intervene, how to intervene, when to intervene, and what type of intervention is needed, remain the clinical questions that must be answered. Acknowledgments: The authors wish to thank Dr. Paul Young for his expertise in reviewing the manuscript. This study was presented in part at the Society for Adolescent Medicine meeting, San Diego, California, March 2001.
References 1. Ventura SJ, Curtin SC: Recent trends in teen births in the United States. Stat Bull Metrop Insur Co 1999;2 2. Polit DF, Kahn JR: Early subsequent pregnancy among economically disadvantaged teenage mothers. Am J Public Health 1986; 76(2):167 3. Stevens-Simon C, White MM: Adolescent pregnancy. Pediatr Ann 1991; 20(6):322 4. Gispert M, Brinich P, Wheeler K, Krieger L: Predictors of repeat pregnancies among low-income adolescents. Hospital and Community Psychiatry 1984; 35:719 5. Furstenberg FF, Brooks-Gunn J, Morgan SP: Adolescent mothers and their children in later life. Fam Plann Perspect 1987; 19(4):142 6. Stevens-Simon C, Kelly L, Singer D: Absence of negative attitudes toward childbearing among pregnant teenagers. A risk factor for a rapid repeat pregnancy? Arch Pediatr Adolesc Med 1996; 150(10):1037 7. Maynard R, Rangarajan A: Contraceptive use and repeat pregnancies among welfare-dependent teenage mothers. Fam Plann Perspect 1994; 26:198 8. Moore K: Facts at a glance. Child Trends, 2001 9. O’Sullivan AL, Jacobsen BS: A randomized trial of a health care program for first-time adolescent mothers and their infants. Nurs Res 1992; 41(4):210 10. Fiscella K, Kitzman HJ, Cole RE, Sidora KJ, Olds D: Does child abuse predict adolescent pregnancy. Pediatrics 1998; 101(4):620 11. Alan Guttmacher Institute: Sex and America’s Teenagers. New York, Alan Guttmacher Institute, 1994 12. Rigsby DC, Macones GA, Driscoll DA: Risk factors for rapid repeat pregnancy among adolescent mothers: a review of the literature. J Pediatr Adolesc Gynecol 1998; 11:115 13. De Lissovoy V: Child care by adolescent parents. Child Today 1973; 2(4):22 14. Furstenberg FF, Levine JA, Brooks-Gunn J: The children of teenage mothers: patterns of early childbearing in two generations. Fam Plann Perspect 1990; 22:54 15. Akinbami LJ, Cheng TL, Kornfeld D: A review of teentot programs: comprehensive clinical care for young parents and their children. Adolescence 2001; 36(142):381
Original Studies Predictors of Repeat Pregnancy in a Program for Pregnant Teens Mark A. Pfitzner, MD, MPH, Charles Hoff, PhD, and Kathleen McElligott, MD Department of Pediatrics, University of Utah, Salt Lake City, Utah, USA
Abstract. Study Objective: To describe repeat pregnancy among adolescents and to compare those who experienced a repeat pregnancy to those who did not. Design, Setting, Participants: Retrospective case control of all adolescents who entered and exited the Teen Mother & Child Program, a multidisciplinary clinic for pregnant and parenting teens and their children, between 1985 and 2000. Main Outcome Measure: Repeat pregnancy. Results: Over the 16-yr study period, 1838 teens entered and exited the program with the mean time in the program of 1.9 yrs. 194 (10.6%) teens went on to have a second pregnancy, and 4 of those had a third pregnancy. Ninetytwo percent (n ⫽ 175) of these pregnancies resulted in a live birth. The mean interval between delivery of the initial child and delivery of the second child was 21.8 months (range 5.4–53.2). Teens repeating (R) were compared to those who did not repeat (NR). R were younger at entry and older at exit from the program (both P ⬍ 0.0001). R were more likely to be Hispanic (P ⫽ 0.009) or have a partner who was Hispanic (P ⫽ 0.02). R were more likely to have experienced a poor initial pregnancy outcome (miscarriage, stillbirth) (P ⫽ 0.03). R were more likely than NR to be in a stable relationships with the father of the baby (i.e., married or engaged) (P ⫽ 0.03). Self-report of physical and sexual abuse, depression, and substance abuse were common, but did not differ between R and NR. However, R were more likely than NR to self-report suicide gestures/attempts and to have a significant psychiatric history (P ⫽ 0.01, P ⫽ 0.004). Only 24% of R and 26% of NR had completed high school by the time of exit from the program. Conclusions: Ten percent of adolescents served by comprehensive multidisciplinary teen pregnancy program experienced a repeat pregnancy while in the program. Compared to nonrepeaters, adolescents who experienced a repeat pregnancy were younger, were more frequently Hispanic, and were more likely to be in a stable relationship with the
Address reprint requests to: Mark A. Pfitzner, MD, MPH, Department of Pediatrics, 30 N. 1900 East, Rm 1B386, Salt Lake City, UT 84132; E-mail: [email protected]
쑖 2003 North American Society for Pediatric and Adolescent Gynecology Published by Elsevier Science Inc.
baby’s father. Many of the program’s clients have psychosocial factors reported in the literature to be associated with repeat pregnancy. A history of suicide gestures/attempts and a significant psychiatric history were more common in those who had a repeat pregnancy. Disappointingly, only about 25% of the adolescents completed high school by the time they exited the program whether they experienced a repeat pregnancy or not. Although directing interventions (e.g., mental health services, counseling those who miscarried) to teens who appear to be at highest risk for a repeat pregnancy may decrease their risk of repeating, all teens in our program would likely benefit from such services.
Key Words. Adolescent pregnancy—Repeat pregnancy
Introduction Historically, repeat pregnancy rates among adolescents have been high,1 with 30% becoming pregnant in their first postpartum year and another 25% to 50% becoming pregnant in the second year. The consequences of repeat pregnancy are significant for the adolescent and society. Adolescents who experience a repeat pregnancy are more likely to leave school without graduating, to be unemployed, and to be welfare dependent.2 Programs specifically designed to address this have had some limited success.3 Previous studies have found several factors associated with repeat pregnancy among teenagers. These include inconsistent contraceptive use,4 lower educational achievement,5 and a nonsupportive family environment.6 Maynard reported that one-half of teenagers receiving public assistance became pregnant for a second time within 24 months of giving birth to their first child.7 Service programs exist throughout the country to provide care and support to pregnant and parenting 1083-3188/03/$22.00 doi:10.1016/S1083-3188(03)00011-1
78
Pfitzner, Hoff, and McElligott: Predictors of Repeat Teen Pregnancy
teens. One such program is the University of Utah Teen Mother and Child Program (TMCP). TMCP is a multidisciplinary program that has provided care to pregnant and parenting teens since the early 1980’s. TMCP provides comprehensive care directed to both the unique medical and psychosocial issues of a high-risk adolescent population. One objective of the program has been to provide services intended to reduce the risk of a repeat pregnancy while the adolescent is under our care. To better understand the phenomenon of early repeat pregnancy in our population, we designed a study to determine the factors associated with a teenager’s having a second pregnancy while still in our program. Since assisting TMCP clients to complete high school while in our program is a goal, we wanted to determine the effect of a second pregnancy on achieving that positive outcome.
Methods The study group included all clients who entered and exited the program between 1985 and 2000. Teens are eligible for care in the program until they reach 19 years of age or their child(ren) become 2 years old, whichever occurs later. Some voluntarily exit the program while still eligible for services. Since 1985, TMCP has maintained a detailed database containing demographic, psychosocial, and outcome data for the teen and her child. Demographic and medical information comprising the database is obtained from the medical record. In addition, each client has an intake interview with the program’s social worker that focuses on psychosocial issues. Data from both sources is then entered into the database in a standardized manner. We used data associated with the initial pregnancy for comparison between those who experienced a repeat pregnancy (R) and those who did not (NR). Information regarding first pregnancy was not available for 35 individuals who experienced a repeat pregnancy. These teens had either been in the program prior to the database initiation or joined TMCP following a first pregnancy for which they received care at another facility. Typically these teens had been referred to TMCP by one of several community agencies. We included data for the second pregnancy only in the descriptive analysis of the repeat population but not for the comparisons. We compared those who had a repeat pregnancy to those who did not on all the variables in the database using ANOVA or chi-square as appropriate. Statview for Windows (Version 5.0.1, SAS Institute, Inc, Cary, NC, 1992–1998) was used for all calculations. The study was approved by the University of Utah Health Sciences Center’s Institutional Review Board.
Fig. 1. Mean birth interval by year of entry into the program.
Results The results are based on data obtained from all 1838 teens who entered the program between January 1985 and May 2000, and exited the program between February 1988 and December 2000. Mean time in the program was 1.9 ⫾ 0.1 yrs. Forty-six percent (n ⫽ 853) remained in the program through their entire eligibility (i.e., they did not exit prematurely). The 1838 pregnancies resulted in 1734 singleton births and 7 sets of twins; 97 resulted in either a stillbirth or a miscarriage. Of the 1838 teens enrolled in the program, 194 (10.6%) had a subsequent pregnancy and 4 had a third (i.e., second repeat). Ninety-two percent (n = 175) of these second pregnancies resulted in a birth; 2 were twin births. All second repeats were single live births. Overall, the average time between birth or miscarriage of the initial child and birth or miscarriage of the second child was 21.8 ⫾ 1.7 months (range 5.4–53.2) This interval remained constant over the period of the study (Fig. 1). Younger teens who repeated had a longer interval between pregnancies than their older peers (P ⬍ 0.0001) (Fig. 2).
Fig. 2. Linear regression of repeat birth interval by maternal entry age.
Pfitzner, Hoff, and McElligott: Predictors of Repeat Teen Pregnancy
79
Table 1. Analysis of Variance between Repeaters and Nonrepeaters for Select Continuous Variables VARIABLE (range)
REPEATERS mean (SE)
Maternal age at entry (12.6–19.7 yrs) Maternal age at delivery (13.0–19.6 yrs) Maternal age at exit (14.0–22.6 yrs) Paternal age at entry (12–63 yrs) Paternal maternal age difference (⫺3.7–46.6 yrs) Time known father of baby (0–132 mo) Time in program (0–6.5 yrs) Gestational age when prenatal care began (4–40 wks) Infant’s birth weight (530–4734 grams) Last grade completed (0–12) Months out of school (0–70)
15.9 (0.08) 16.3 (0.08) 19.5 (0.10) 18.8 (0.25) 2.9 (0.25) 18.1 (2.36) 3.6 (0.09) 15.5 (0.74) 3128 (50.6) 9.0 (0.17) 5.3 (1.14)
The results of the comparisons between teens who experienced a repeat pregnancy (R) and those who did not (NR) are shown in Tables 1 and 2. R were younger at entry and delivery, spent more time in the program, and were older when exiting. Consistent with their younger age, R averaged 0.5 fewer school grades completed compared to NR. R were more likely to have a significant psychiatric history and to have more frequently contemplated or attempted suicide. R and NR had similarly high rates of reported physical and sexual abuse and depression. R were less likely than NR to have placed their child for adoption but more likely to have experienced an unfavorable outcome with their first pregnancy. R were more likely to be in a committed relationship (i.e., living together or married) and to be Hispanic or to have a Hispanic partner than NR. At the time of exiting the program, R and NR had similar rates of either being currently enrolled or having graduated from high school.
Discussion Teen pregnancy rates are decreasing nationally, but 22% of births to teens are repeat births.8 Although the overall decline is encouraging, the absolute number of adolescent females of childbearing age is increasing by 5% to 10% per year; thus both initial and repeat pregnancies among adolescents continues to remain a significant problem.1 Our program had a repeat pregnancy rate of 10.6% during the 15-yr study period, a rate that is similar to other established teen pregnancy programs providing care to adolescent mothers and one that is lower than the national average.1,6,9 Several psychosocial and lifestyle factors are commonly assumed to be associated with an increased rate of early repeat pregnancy among teenagers. We anticipated that those adolescents in our program who experienced a repeat pregnancy would have higher rates of these when compared to teens who did not have a
NONREPEATERS mean (SE)
ANOVA P value
16.6 (0.03) 17.0 (0.03) 18.4 (0.03) 19.1 (0.09) 2.6 (0.09) 15.7 (0.49) 1.8 (0.03) 15.6 (0.19) 3171 (14.1) 9.5 (0.05) 8.1 (0.40)
⬍0.0001 ⬍0.0001 ⬍0.0001 NS NS NS ⬍0.0001 NS NS 0.005 NS
repeat pregnancy within two years of the first. However, with few exceptions, the rates of these factors were similar among the two groups. Both groups reported high rates of physical and sexual abuse, depression, and alcohol, tobacco, and illicit drug use. Other investigators have reported that sexual abuse during childhood is associated with younger age at first coitus and pregnancy.10 As many as 50% to 60% of pregnant adolescents report physical or sexual abuse.11 We did find that teens who had serious mental health problems including a history of suicidal thought, gestures, or attempts were more likely to have a repeat pregnancy. Other investigators have found that low educational achievement is a strong predictor of repeat pregnancy.12 In our study, those who repeated were enrolled in school or had graduated from school at the same rate as their nonrepeated peers. Disappointingly, only 25% of the entire population had graduated or were enrolled in school when they exited the program. Differences between repeaters and nonrepeaters were found when we examined the teen’s ethnicity and her relationship with the father of the baby. Repeat pregnancies were more common in Latinos and among those who were married or living with the baby’s father. Studies from the 1960’s and 1970’s also reported marriage as being a risk factor for repeat pregnancies.13,14 Teens whose first pregnancy ended in a miscarriage or abortion were more likely to have a second pregnancy. Placement for adoption was less common among repeaters. Repeat pregnancies were spaced nearly 2 yrs apart. Despite changes in contraceptive practices during the study years (e.g., the advent of longer-acting contraceptives like Norplant and Depo-Provera, the social marketing of condom usage, and “safe sex” campaigns with the advent of HIV) birth interval did not change appreciably in our population. Clearly this study has limitations. While the data was collected using standardized collection forms, changes in personnel over the years could have led
80
Pfitzner, Hoff, and McElligott: Predictors of Repeat Teen Pregnancy
Table 2. Chi-Square Analysis Comparing Repeaters to Nonrepeaters
VARIABLE Physical Abuse Yes No Unknown Sexual Abuse Yes No Unknown Depression Yes No Unknown Suicidality Ideation/gesture/attempt None Significant Psychiatric Historya Yes No Unknown Alcohol Useb Experiment/past or current use None Tobacco Useb Experiment/past or current use None Illicit Drug Useb Experiment/past or current use None Parent a Pregnant Teen Yes No Planned Pregnancy Yes No Ambivalent Placed Child for Adoption Yes No School Attendance at Entry Enrolled/graduated Not attending Maternal Ethnicity Non-Hispanic white Hispanic Other Paternal Ethnicity Non-Hispanic white Hispanic Other Relationship at Conception Married/Cohabitating Dating regularly Other Pregnancy Outcome Live birth Miscarriage/Abortion Educational Status at Exit Enrolled/graduated Not attending
χ2 P value
REPEATERS n (%)
NONREPEATERS n (%)
42 (30) 97 (69) 2 (1)
482 (31) 1062 (68) 16 (1)
NS
55 (39) 83 (59) 3 (2)
548 (35) 996 (64) 17 (1)
NS
78 (55) 61 (43) 2 (2)
735 (47) 818 (52) 8 (1)
NS
101 (72) 40 (28)
949 (61) 610 (39)
27 (19) 109 (77) 5 (4)
257 (16) 1290 (83) 12 (1)
45 (74) 16 (26)
689 (72) 272 (28)
NS
39 (64) 22 (36)
590 (61) 372 (39)
NS
36 (59) 25 (41)
546 (57) 413 (43)
NS
15 (10) 140 (90)
142 (9) 1504 (91)
NS
28 (20) 90 (64) 23 (16)
213 (14) 1083 (70) 262 (16)
1 (1) 154 (99)
83 (5) 1565 (95)
0.01
51 (49) 53 (51)
739 (54) 619 (46)
NS
101 (65) 35 (23) 19 (12)
1265 (76) 259 (16) 133 (8)
0.009
80 (52) 50 (32) 25 (16)
1030 (63) 396 (24) 217 (13)
0.02
92 (60) 44 (29) 17 (11)
828 (51) 645 (40) 154 (9)
0.03
145 (94) 10 (6)
1596 (97) 52 (3)
0.03
20 (26) 56 (74)
183 (24) 570 (76)
NS
0.01
0.004
NS
Pfitzner, Hoff, and McElligott: Predictors of Repeat Teen Pregnancy
81
Table 2. (Continued)
VARIABLE Relationship which Father of Baby at Exit Married/cohabitating Dating No relationship/seeing child only Exit Reason Moved/transferred care Not eligible Lost to follow-up Payer Source Insurance Self pay Medicaid a
χ2 P value
REPEATERS n (%)
NONREPEATERS n (%)
74 (48) 10 (7) 68 (45)
679 (42) 202 (13) 730 (45)
0.06
26 (18) 97 (66) 23 (16)
440 (30) 756 (53) 242 (17)
0.002
15 (10) 14 (9) 123 (81)
234 (14) 160 (10) 1224 (76)
NS
Defined as having ever had inpatient or outpatient psychiatric treatment. Variable added to database in 1993.
b
to biases in recording. As much of the psychosocial variables rely on patient self-report, perception by the patient regarding these variables may change over time as society changes its perception of them (e.g., depression and mental health issues). Yet there was not a large degree of variability in these variables over time. One large limitation is our inability to know the outcome of any teen who left the program; 54% left prematurely. Although we have a large number who remained, we do not know if those who left early repeated. This is a well-recognized limitation of programs such as ours as few have had rigorous efficacy studies.15 All teens in this program represent an extremely high-risk population. Repeaters were not adversely affected by their pregnancy when compared to their nonrepeating peers. Perhaps this was due to the longer birth interval seen in this population. However, markers of economic disparities were not examined in this study. Although targeting interventions at those teens at greatest risk would seem advantageous (e.g., by improving mental health services or providing close follow-up to those teens who miscarried or aborted), both repeaters and nonrepeaters in this high-risk population would benefit from those services. Targeting interventions solely to the highest-risk groups in this risky population may only serve to increase repeat pregnancies among those not targeted. Practitioners should realize that pregnancy is but one risk marker for a teen; others may be present. With whom to intervene, how to intervene, when to intervene, and what type of intervention is needed, remain the clinical questions that must be answered. Acknowledgments: The authors wish to thank Dr. Paul Young for his expertise in reviewing the manuscript. This study was presented in part at the Society for Adolescent Medicine meeting, San Diego, California, March 2001.
References 1. Ventura SJ, Curtin SC: Recent trends in teen births in the United States. Stat Bull Metrop Insur Co 1999;2 2. Polit DF, Kahn JR: Early subsequent pregnancy among economically disadvantaged teenage mothers. Am J Public Health 1986; 76(2):167 3. Stevens-Simon C, White MM: Adolescent pregnancy. Pediatr Ann 1991; 20(6):322 4. Gispert M, Brinich P, Wheeler K, Krieger L: Predictors of repeat pregnancies among low-income adolescents. Hospital and Community Psychiatry 1984; 35:719 5. Furstenberg FF, Brooks-Gunn J, Morgan SP: Adolescent mothers and their children in later life. Fam Plann Perspect 1987; 19(4):142 6. Stevens-Simon C, Kelly L, Singer D: Absence of negative attitudes toward childbearing among pregnant teenagers. A risk factor for a rapid repeat pregnancy? Arch Pediatr Adolesc Med 1996; 150(10):1037 7. Maynard R, Rangarajan A: Contraceptive use and repeat pregnancies among welfare-dependent teenage mothers. Fam Plann Perspect 1994; 26:198 8. Moore K: Facts at a glance. Child Trends, 2001 9. O’Sullivan AL, Jacobsen BS: A randomized trial of a health care program for first-time adolescent mothers and their infants. Nurs Res 1992; 41(4):210 10. Fiscella K, Kitzman HJ, Cole RE, Sidora KJ, Olds D: Does child abuse predict adolescent pregnancy. Pediatrics 1998; 101(4):620 11. Alan Guttmacher Institute: Sex and America’s Teenagers. New York, Alan Guttmacher Institute, 1994 12. Rigsby DC, Macones GA, Driscoll DA: Risk factors for rapid repeat pregnancy among adolescent mothers: a review of the literature. J Pediatr Adolesc Gynecol 1998; 11:115 13. De Lissovoy V: Child care by adolescent parents. Child Today 1973; 2(4):22 14. Furstenberg FF, Levine JA, Brooks-Gunn J: The children of teenage mothers: patterns of early childbearing in two generations. Fam Plann Perspect 1990; 22:54 15. Akinbami LJ, Cheng TL, Kornfeld D: A review of teentot programs: comprehensive clinical care for young parents and their children. Adolescence 2001; 36(142):381