The impact of parental involvement laws on minors seeking abortion services: a systematic review

Abstract On June 24, 2022, the US Supreme Court overturned the constitutional right to abortion in Dobbs v Jackson Women's Health Organization. Minors are a vulnerable population with a high risk of unintended pregnancy who are likely to be disproportionately affected by abortion restrictions. Examining the impact of historical abortion restrictions in minors may provide insight into the anticipated effects of the Dobbs decision. This study is a systematic review examining the impact of parental involvement (PI) laws on minors seeking abortion services. Our review suggests an association between PI laws and decreased abortion rates. Parental involvement laws also may increase pregnancy and birth rates and out-of-state travel for abortion procedures and lead to later abortions, although effects appear to be heterogenous.


Introduction
On June 24, 2022, the US Supreme Court voted 5 to 4 to overturn the constitutional right to abortion in Dobbs v Jackson Women's Health Organization.This historic decision will be far reaching in its impact, and vulnerable populations will be disproportionately affected.Minors are a vulnerable population with the highest risk of unintended pregnancy. 1These pregnancies are associated with an increased risk of adverse outcomes for both pregnant individuals and children. 2,3This study is a systematic review examining the impact of parental involvement (PI) laws on minors seeking abortion services pre-Dobbs decision.
In 1979, 6 years after the passing of Roe v Wade, the US Supreme Court ruled in Bellotti v Baird that states could require a pregnant, unmarried minor to obtain parental consent prior to obtaining an abortion. 4Over the following 4 decades, 37 states passed various PI laws. 5Parental involvement laws fall into the general categories of need for minors to obtain parental consent to abortion services or need for minors to notify parents of intention to access abortion services.
While minors have legal autonomy to make decisions with regard to contraception and a pregnancy they choose to continue, abortion is a unique aspect of sexual and reproductive health for which they are not allowed independent decisionmaking capacity.In the absence of PI laws, studies suggest that the majority of minors inform a parent or a guardian of their decision to obtain an abortion. 6Minors who choose not to involve their parents commonly cite the desire to preserve the parent-child relationship and maintain autonomy.Others express fear of violence, being forced out of their home, or the inability to involve parents due to logistical barriers. 6,7Since the majority of minors seeking abortion services already involve an adult, PI laws may impact the most vulnerable minors who do not feel that they can safely involve a parent/guardian.
Legal abortions are safe and effective, and limiting access to abortion increases the risk of adverse social and health-related outcomes. 8Parental involvement laws could result in a number of adverse outcomes, including the following: emotional hardship, risk of violence, delays in care with resultant need for a higher risk procedure at a later gestational age, need for out-of-state travel to obtain abortion services, or inability to access abortion services.Our findings may help anticipate consequences of further limitations to abortion access resulting from the Dobbs decision.

Search strategy
This systematic review was performed according to recommendations in the Cochrane Handbook and reported in accordance with the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) 2009 statement. 9 review protocol was developed and registered in the PROSPERO International Prospective Register of Systematic Reviews (registration number CRD42020157863).A systematic literature search strategy was developed with an expert in library sciences (Table S1).We identified English-language publications examining the impact of PI laws on minors seeking abortion services from the passing of Roe v Wade in 1973 to 2021, through searches of PubMed, CINAHL, EMBASE, Web of Science, PsycINFO, and Scopus.
Two authors (A.K. and L.T.) independently screened titles and abstracts for potential inclusion using Covidence (Covidence systematic review software, Veritas Health Innovation, Melbourne, Australia; available at: www.covidence.com).The same reviewers assessed the full texts of the identified articles based on the initial screening.Discrepancies were resolved by a third reviewer (A.T.) (Figure 1).

Inclusion and exclusion criteria
Studies eligible for inclusion were those published in peerreviewed journals that examined the impact of PI laws, defined as parental consent or parental notification, on minors in the United States seeking abortion care from 1973 to 2021.Studies examining judicial bypass were excluded.Studies had to disaggregate outcomes data for minors less than 18 years of age specifically.Case reports, commentaries, and review articles were excluded.

Outcomes
The outcomes of interest included the following: abortion rate, birth rate, pregnancy rate, gestational age at time of abortion, delay in care to receiving abortion services, second-trimester abortion rate, and out-of-state travel to obtain abortion services.

Data extraction
Two authors (A.K. and A.L.) extracted data from eligible studies using a data-extraction form (key elements reported in Table 1).Abstracted data were discussed among the 3 reviewers (A.K, A.L., and A.T.) and disagreements resolved through consensus.
Parental involvement laws were classified as notification laws or consent laws.In some studies, both consent and notification laws were in place, the type of law was not reported, or data were analyzed from multiple states with different types of laws.In this case, the type of law was classified generically as not otherwise specified (NOS).
Risk of bias was assessed using a modified Risk of Bias in Non-randomized Studies of Interventions (ROBINS-I) instrument. 32Our structured instrument provided a framework for the assessment of risk of bias in the following domains: selection bias, information bias, confounding, and reporting bias.Selection bias examined the differences between baseline characteristics of our control and comparison groups.Information bias assessed whether study variables were accurately measured.Reporting bias assessed whether the outcomes measured were prespecified and whether all prespecified outcomes were reported.Confounding bias was assessed based on the characteristics of the control group and whether effect estimates were adjusted for state and demographic variation, time trends, and out-of-state travel.Risk of bias was reported as low, moderate, or high (Figure 2).

Results
Twenty-four studies met the inclusion criteria (Figure 2).Results disaggregated by outcome are discussed below (see also Table 1 with more detailed information).

Abortion rates
Sixteen of the 24 studies 7,10,11,14,15,18,19,[21][22][23][25][26][27][28]30,31 found decreases in abortion rates in states with enactment of PI laws (4 Consent, 3 Notice, 9 NOS). Of these stdies, 7 had a low risk of bias, 4 had a moderate risk of bias, and 5 had a high risk of bias.One low-risk-of-bias study in Illinois found no difference in abortion rates after enactment of a parental notice law.29 Records Health Affairs Scholar, 2023, 1(4), 1-13 Health Affairs Scholar, 2023, 1(4), 1-13 Finally, 2 low-risk-of-bias studies had mixed results.One found no significant change in abortion rates in South Carolina and Tennessee after enactment of PI laws, 16 except for the subpopulation of 16-year-olds who were non-Black in South Carolina who were found to have a decreased probability of accessing abortion.The other study 17 analyzed the effect of Mississippi's 2-parent consent law and South Carolina's 1-parent consent law.In Mississippi, total abortions among minors decreased, whereas in South Carolina, the parental consent law did not have a significant effect on abortion rates of minors.

Birth rates
][28] Of these, 4 were low risk, 1 was moderate risk, and 1 was classified as high risk of bias.Four moderate-risk-of-bias studies found no evidence that PI laws impact birth rates among minors (1 Consent, 1 Notice, 2 NOS). 13,14,20,21Three studies 10,18,30 found that PI laws were associated with decreased birth rates to minors (2 Notice, 1 NOS; risk of bias: 1, low; 1, moderate; 1, high), although in 1 case the observed decline was consistent with long-term population-level trends 30 and less likely as a result of the law.

Pregnancy rates
Three studies found that states with PI laws have significantly lower pregnancy rates among minors (3 NOS).Of these 3 studies, 2 had moderate risk of bias and 1 had a high risk of bias. 10,14,27One study (Consent, low risk of bias) found decreasing pregnancy rates among minors that were likely due to the continuation of a pre-existing trend, rather than as a result of the PI law. 28Three studies found no significant change in the pregnancy rate associated with PI laws (1 Notice, 2 NOS; risk of bias: 3 moderate). 13,21,24

Abortion timing
Five studies 12,17,18,28,30 found an increase in second-trimester abortions among minors associated with PI laws (2 Consent, 3 Notice; risk of bias: 4, low; 1, high).One of these studies is highlighted in greater detail below.Five studies 7,14,15,22,29 found no significant change in the number of second-trimester abortions associated with enactment of PI laws (2 Consent, 3 Notice; risk of bias: 3, low; 1, moderate; 1, high).In 1 of these studies 14 although there was no difference in second-trimester abortions in Minnesota, a parental notification law was associated with a delay in abortion timing past the eighth week of gestation (P < .001) in minors when compared with older women.
One study 17 had mixed results.Although the authors found no association between South Carolina's consent law and the timing of abortions among minors, they did find that Mississippi's consent law resulted in abortions at later gestational ages among minors relative to non-minors (0.6 wk later; P < .05).
Following the implementation of a parental notification law in Texas, 12 the authors found that pregnant minors who were aged 17 years and 8-9 months at the time of conception were more likely to delay obtaining an abortion until turning 18, resulting in abortions performed at statistically and highly clinically significant later gestational ages (18.2 wk vs 8.4 wk).Overall, there was a 3% increase in the proportion of pregnancies terminated in the second trimester due to delaying abortions in these age groups, which represents an increase in the second-trimester abortion rate of 22% (P = .06).

Resident travel out of state
16][17]28 Of these 6 studies, 5 had a low risk of bias and 1 had a moderate risk of bias.
laws result in decreased rates of abortions to minors aged 17 and under.However, there is heterogeneity in this outcome, as well as heterogeneity of effect across the other outcomes analyzed.
In contrast, a narrative review performed in 2009 by Dennis et al 33 that examined the impact of PI laws in 29 published studies found that PI laws resulted in an increase in out-of-state travel by minors to obtain abortion services to states that had less restrictive laws but no difference in abortion rates.The authors also concluded that PI laws had little impact on minors' birth and pregnancy rates.
Parental consent laws are the most restrictive type of PI law, as they require a minor to receive documented consent from either 1 or both parents prior to obtaining an abortion.Parental notification laws require only notification of a parent.While some studies found no difference in outcomes when comparing consent with notification laws, greater impacts were generally seen in states with consent laws.Parental consent laws were associated with decreased abortions in all 5 included studies, while birth rates increased in 2 out of 3 studies that reported on this outcome.
The impact of parental notification laws is less clear.Five out of 6 studies showed a decrease in abortions associated with these notification laws and both studies that reported on delays in accessing abortion care found significant delays.However, the impact of notification laws on the remaining study outcomes suggests either no association or the available studies report conflicting results.Similar to the reasoning above, the lack of consistent association may be explained by the observation that notification laws are less stringent than consent laws in that a minor can, in theory, still obtain an abortion after notifying a parent or guardian.
Multiple studies found that the decreases in abortion rates were offset by minors traveling out of state to obtain abortion services.This out-of-state travel places significant financial and emotional burden on patients seeking care and contributes to widening disparities in care by limiting access to abortion services for the most vulnerable patients.Minors with financial resources and social support may be able to access care, while those without resources may have no alternative but to continue their pregnancy.Some minors may attempt to end the pregnancy on their own through self-managed abortion.Self-managed abortion may be higher among adolescents, 34 although the prevalence of self-managed abortion, especially in individuals under age 18, who are often excluded from population-based surveys, remains unknown. 35ost studies examining the impact of parental consent laws observed an increase in birth rates.This suggests that consent laws do not have a significant impact on sexual activity or contraceptive use.If the goal of PI laws is to decrease sexual activity among minors or decrease pregnancy rates, then other interventions such as comprehensive sexual education and increased access to contraception may be more effective strategies that empower patients, while adhering more closely to the medical ethics principle of patient autonomy.
Given this observed increase in birth rate, it is important to consider the associated increases in pregnancy morbidity and mortality.Legal abortions are safe and effective procedures.Complications from abortion procedures have been documented to occur in less than 1% and not more than 5% of abortions in published studies, with severe complications occurring in no more than a fraction of a percentage of patients. 36In comparison, complication rates in pregnancy and childbirth are much higher.In 2021, 1205 pregnant persons died in the United States due to pregnancy causes. 37It is estimated that 50-100 pregnant persons experience severe morbidity for every single pregnancy-related death. 38Disaggregated morbidity and mortality data for minors are not publicly available.A maternal health crisis already exists in the United States.Increasing birth rates by preventing pregnant individuals from accessing abortion services risks exacerbating already unacceptably high rates of pregnancy morbidity and mortality.
The impact of parental consent laws also shows trends toward later gestational ages at the time of abortion and potential increases in second-trimester abortions.Although abortions are overall low-risk procedures, there is an increased risk of complications with increasing gestational ages. 36From an economic standpoint, procedures performed at later gestational ages are more time intensive, require more resources, and are more expensive.For patient safety and efficient utilization of health care resources, minimizing delays to care should be a priority.

Limitations
A limitation of any systematic review is the quality of included studies.Only 25% of studies included in this review were deemed low risk of bias based on risk-of-bias analysis.The Health Affairs Scholar, 2023, 1(4), 1-13 risk-of-bias analysis was limited by not all elements of the Robins-I tool being applicable to the designs of the included studies.The heterogeneity in the included studies also limited our ability to infer effects of PI laws in different contexts and precluded a meta-analysis.Indeed, effects of PI laws may differ across contexts.Comparison groups in the included studies varied by age, time, and location.The exposure also varied widely, as some PI laws require 2-parent consent with identification requirements or even proof of parenthood, while other laws require notification of 1 parent or allow other relatives or guardians to be notified or provide consent. 5Some studies examine data nationally, while others focus on a region or state.As a result, authors use different techniques to control for confounding across distance and time.
An overall lack of population-based data on abortions also limits the ability of the review to draw conclusions.Abortion data in the United States come from 3 primary sources: the Guttmacher Institute, the Centers for Disease Control and Prevention, and state health departments.Each data source has important limitations.Known abortion data limitations include the following: abortions tallied by the state of abortion occurrence rather than according to the patient's state of residence, which limits evaluation of abortion rates and out-of-state travel; lack of data stratified by age, especially for minors; and underreporting and incompleteness of reporting between states (Table 2).
Finally, the included studies examined data over many years.Since the impact of PI laws is likely greater when there are no neighboring states to which to travel for out-of-state abortion services, the impact of PI laws has likely increased over time as additional states have enacted PI laws.
Strengths of this review include a rigorous study design, adherence to PRISMA guidelines for systematic reviews, and examination of multiple outcomes potentially affected by PI laws, which allows for a more holistic view of downstream health impacts of policy changes.

Conclusion
Our review suggests an association between PI laws and decreased abortion rates.Parental involvement laws may also increase pregnancy rates, birth rates, and out-of-state travel for abortion services, and lead to abortions at later gestational ages.Heterogeneity of studies limits inference.Available evidence suggests that parental consent laws may result in greater impacts than notification laws.However, many questions remain unanswered.Parental involvement laws are still in effect in 36 states and will continue to impact minors.The recent overturning of Roe v Wade paves the way for further restrictions to abortion access and the potential for ever increasing adverse pregnancy outcomes.Additional studies will be needed to examine the impact of extreme restrictions to abortion access on pregnancy morbidity, mortality, and perinatal outcomes.

Table 1 .
Summary of reported outcomes by study.

Table 1 . Continued Study Risk of bias State(s) Law type Time period Number of observations Comparison group Abortion data source Statistical methods Outcomes
Abbreviations: CDC, Centers for Disease Control and Prevention; GA, gestational age; Mod, moderate; NCHS, National Center for Health Statistics; NOS, not otherwise specified; PI, parental involvement.

Table 2 .
Summary of 3 major abortion data sources and limitations.: CDC, Centers for Disease Control and Prevention. Abbreviation