Copyright 2006 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.2006
To assess and compare parent and adolescent views on the importance of risk behavior research and the need for parental consent and to identify predictors of views.
Adolescent, general pediatrics clinics.
Adolescents aged 14 to 17 years presenting alone (solo) or with a parent or guardian (paired) and parents or guardians. Of 265 eligible pairs (199 solo adolescents), 134 (93) had useable surveys. The proportion of females in the parent or guardian, paired adolescent, and solo adolescent groups was 92%, 59%, and 75%, respectively; and the proportion of African Americans in these groups was 67%, 69%, and 91%, respectively.
Main Outcome Measures
Views on the importance of and requirement of parental consent for confidential risk behavior research.
Most parents (98%), paired adolescents (99%), and solo adolescents (100%) believed confidential risk behavior research surveys with teenagers were important; greater than 90% of all groups believed surveys should be conducted. The proportion of parents, paired adolescents, and solo adolescents that endorsed requiring parental permission was 84%, 53%, and 19%, respectively. Parents were less likely to endorse requiring permission if they had confidential health concerns as teenagers and were more likely to endorse requiring permission if their child was younger or female, if they believed their child had sex, and if they were a parent vs a guardian. Adolescents were more likely to endorse requiring permission if they were younger, not African American, and had more educated parents. Adolescents presenting with parents were more likely to endorse requiring permission.
Parents and adolescents believed risk behavior research with adolescents was important. Most parents believed parental permission is needed for participation. Compared with parents, fewer adolescents believed parental consent was necessary. It is not clear what effect requirement of parental permission would have on participation or validity. Further research is needed to elucidate views on adolescent risk behavior research.
Adolescent risk behaviors, such as tobacco use, alcohol and other drug use, unprotected sexual intercourse, and weapon carrying, account for significant morbidity and mortality. Research on these risk behaviors, including the perspectives of adolescents, is essential to address these issues and improve adolescent health.1 Information on adolescent risk behaviors has frequently been obtained from confidential research surveys conducted without parental consent. Considerable uncertainty and controversy exist regarding the ethical guidelines and legal statutes that relate to adolescents' involvement in research.2,3 Legal and legislative challenges have led to increased scrutiny of such research.4,5 Including adolescents as research subjects has resulted in debate among parents and schools,6 in research review boards,7 in regulatory and legislative bodies,8- 10 and in the existence of conflicting guidelines.11- 14
Recent legal decisions have also questioned the involvement of minors in research4,15 and the need for parental knowledge regarding their children's participation in the research.6,8,16 This has directly impacted research with adolescents. CN v Ridgewood Board of Education6 challenged schools' ability to conduct confidential risk behavior surveys and led to the cancellation of subsequent surveys.17 This case was framed as an issue of “parents' rights” and, while struck down by the Supreme Court, led to legislative change in New Jersey, limiting adolescent research and increasing requirements for parental consent. Because of difficulty in interpreting federal regulations on adolescent involvement in research, adolescents have been excluded from participation in clinical trials, studies in public health prevention, and other critical research efforts.13,18,19 These surveys are considered minimal-risk research. Recent studies20,21 have shown increased parent and adolescent acceptance of adolescent participation in minimal-risk clinical research. Amid this discussion, there has been little systematic research exploring parent and adolescent perspectives on this important issue of parental consent for risk behavior survey research involving adolescents.
The primary objectives of this study were to assess parent and adolescent (1) perceptions regarding the importance and need for research on adolescent risk behaviors and (2) attitudes regarding a requirement for parental consent for adolescent involvement in confidential risk behavior survey research. Our secondary objectives were to compare parents’/guardians' attitudes and views with those of their adolescents, to compare the views of adolescents receiving care with parents present vs those of adolescents receiving care alone, and to identify predictors of these views.
The study sample was drawn from a population of adolescents and parents or guardians at 2 clinical sites: an urban academic adolescent medicine clinic and a community general pediatrics clinic. Between November 23, 2003, and August 19, 2004, participants were recruited in waiting rooms during a clinic visit. Adolescents were eligible for enrollment if they spoke English and were aged between 14 and 17 years. Adolescent participants were either “solo,” if they presented without a parent or guardian; or “paired,” if they presented with a parent or guardian, both spoke English, and both completed the informed consent process. The selected age range for adolescents was based on prior studies22 suggesting that teenagers of this age had the capacity for informed consent/assent. Persons without the physical or mental capacity to participate (eg, hearing loss in one grandmother) were excluded.
An audiotaped questionnaire was used to improve confidentiality, lessen literacy concerns, and allow participation of multiple pairs at once. This modality has been shown to increase survey completion and accuracy when sensitive data are being gathered.23 The study was approved by The Johns Hopkins University institutional review board and the research review board for the community pediatric clinic.
Survey items were initially drawn from other confidentiality questionnaires on adolescent research.24,25 Focus groups and in-depth interviews with talk-aloud pretesting were conducted with 25 adolescents and 19 parents or guardians to assess relevance and comprehension. The recorded surveys and consent process were also tested to evaluate understanding of specific items, audio quality of the recorded survey, and timing of the enrollment process.
Questions included Likert-scaled, true or false, multiple choice, short fill-in, and yes or no responses. The surveys consisted of questions on demographics, perceptions of the importance of and need for adolescent risk behavior research surveys, perceived need for a requirement for parental permission to participate in this research for various topics (smoking, alcohol use, other drug use, having sex or using birth control, abortion, depression, carrying a weapon, and physical or sexual abuse), and past experience with giving informed consent for relevant research. Predictor variables assessed included relationship to the adolescent (parent vs guardian), parent or guardian and adolescent gender, adolescent age, race/ethnicity, and parental educational level; parent or guardian personal history of a confidential health issue as a teenager and parent or guardian belief that the adolescent had experienced or been involved in risk-taking behaviors and belief that the adolescent has a current confidential health concern; and adolescent report of a current confidential health concern. Race/ethnicity was included as a self-selected demographic variable because prior studies24 have indicated differences in perceptions regarding the confidentiality of risk behavior issues. Research surveys were described as “surveys . . . used to gather information about teens' behaviors and attitudes on sex, drugs and violence. Research surveys help develop programs for teenagers as a group, but NOT for each person answering the survey.” The study focused on confidential survey research, with questions specifying research in which “no names are used.”
Participants were recruited in clinic waiting rooms at each site on randomly chosen clinic days by one of us (R.H.P.) or a trained research assistant. Potential participants were identified via clinic flyers or approached directly by research staff. Written informed consent was obtained from parents or guardians and solo adolescents, and written assent was also obtained from paired adolescents. Each participant then received an audiocassette player containing the recorded survey and a written answer sheet. Participants were instructed on proper survey completion and use of the equipment and completed surveys independently.
Separate color-coded answer sheets were provided to adolescents and parents. Answer sheets contained no questions but only response options, contained no personal identifiers, and were coded to link parent-adolescent pairs. Research staff were available during survey completion to answer questions and ensure that participants worked independently. Each adolescent completing the survey received a $5 gift card, and parents received a brochure on parenting adolescents.
Data were entered into a Microsoft Access 2000 Version 9.0 database (Microsoft Corp, Redmond, Wash). Double data entry was performed, with less than 1.0% data error. For analysis, data were exported to Microsoft Excel, version 9.0 (Microsoft Corp), for use with Stata 8.0 (Stata Corp, College Station, Tex) and SAS statistical software. Responses to questions on perceived need for adolescent risk behavior research, perceived need for a requirement for parental permission, and predictors were tabulated as frequencies. The McNemar test for paired data was used to test the statistical significance of differences in proportions for parent and adolescent attitudes toward research. χ2 Tests were used to test the differences in proportions for solo and paired adolescent attitudes toward research and the associations between predictors and outcome variables on endorsement of parental permission for adolescent involvement in risk behavior research. Outcome variables included endorsement of a requirement for parental permission for adolescent involvement in research generally and on each research topic. An additional variable was generated for endorsement of parental permission for all 8 research topics combined vs endorsement of fewer than 8 topics. Initial logistic regression models were developed using predictors with P≤.10 in bivariate analyses controlling for adolescent age, gender, current confidential health concern, or participation in risk behavior.
Of the 439 parent or guardian–adolescent pairs approached at the 2 sites, 265 were eligible. Of those eligible, 74 were not interested and 19 noted a time constraint; 170 (64.2%) of those eligible gave consent, and there were 134 useable surveys (Figure). Of the 541 solo adolescents approached, 199 were eligible. Of those adolescents, 87 were not interested and 14 noted a time constraint; 94 (47.2%) gave consent. There were 93 useable surveys (Figure). Table 1 presents participant characteristics. Nearly all of the parents or guardians were female, and most of all 3 groups described themselves as African American.
Flow diagram of enrollment for parent-teenager pairs and solo teenagers.
Virtually all parents or guardians and adolescents believed that research surveys involving teenagers aged 14 to 17 years were important; 95% of parents, 94% of paired teenagers, and 92% of solo teenagers believed these surveys should be done (Table 2). When asked if a teenager should “be REQUIRED to have written parent or guardian permission to do a research survey,” in which no names are used, 84% of parents, 53% of paired teenagers, and 19% of solo teenagers said yes. There was little variability in parental views on parental permission by risk behavior topic; 75% of parents or guardians endorsed a requirement for consent for all 8 topics. Adolescents demonstrated more variability in their views, with only 37% of paired teenagers and only 16% of solo teenagers endorsing a requirement for consent for all 8 topics.
There was a significant difference in paired parent or guardian and adolescent responses for all questions on parental permission, with fewer adolescents than parents or guardians endorsing required parental permission (Table 2). Parent-adolescent differences in proportions were greatest for research involving sexual activity or use of birth control, depression, and abortion. Of parents or guardians, 83% and 81%, compared with 63% and 66% of paired adolescents, endorsed a requirement for parental permission for research involving sexual activity and depression, respectively.
There was a significant difference in paired and solo adolescent responses for all questions on parental permission, with fewer solo adolescents than paired adolescents endorsing required parental permission (Table 2).
Participants were asked about prior experience with research consent and assent forms. Of parents or guardians, 53%, and 45% and 49% of paired and solo adolescents, respectively, reported having signed a consent or assent form in the past; and 27% of parents or guardians and 26% and 27% of paired and solo adolescents, respectively, had previously refused to sign a consent form. Of parents or guardians, 31% had previously been asked to sign research consent for their child. Parent refusal to sign consent or history of signing consent for a child was not related to parental attitudes on parental permission.
Table 3 presents bivariate analyses of factors associated with parents' or guardians' views on requirements for parental permission. Compared with parents of female adolescents, parents of male adolescents were less likely to endorse a requirement for parental permission for adolescent involvement in research on alcohol (odds ratio, 0.4 and 0.5, respectively; P<.10). Parents who had had a confidential health concern were not significantly different from parents without history of a confidential health concern with respect to race or ethnicity (P=.62) or level of education (P=.90).
Table 4 presents logistic regression analyses of parents' views regarding requirement for parental permission of adolescent involvement in risk behavior research. Being a parent, compared with being a guardian, was predictive of an increased need for permission for research involving alcohol use. Parents of younger teenagers (aged 14-15 vs 16-17 years) were more likely to endorse a requirement for parental permission for research generally, and parents of female vs male adolescents endorsed parental permission for research involving alcohol and all topics combined. Parents who believed their adolescents had had sex were much more likely to believe permission was needed for research involving smoking, alcohol use, and other drug use.
Table 5 presents the bivariate associations for paired adolescents regarding the outcome variables with significant differences only, noted by race/ethnicity and level of parental education. In the multivariable analyses presented in Table 6, we found that African American race and lower parental education were independently predictive of decreased odds for endorsing a requirement for parental permission for research on general and specific topics, controlling for adolescent gender, age, and history of current confidential health concern.
Table 7 presents bivariate associations for solo adolescents, and Table 8 presents the logistic regression analyses. Because parental education was not known for this group, this variable and race/ethnicity were not included in the logistic regression analyses. Younger adolescents were much more likely to believe permission should be required for research. Current confidential health concern was not predictive of views on permission for solo adolescents.
Information from research on youth risk behavior is important to understand health issues among adolescents and to determine the appropriate distribution of funds for education and health care services for adolescents. As research and controversy have grown, the need to develop guidelines for adolescent research and clinical care has become increasingly apparent. Professional society and legal and ethical guidelines26- 32 have been developed for the inclusion of adolescents in research, and the challenges regarding confidentiality and consent have been discussed. The Society for Adolescent Medicine13(p407) has suggested in its policy statements that adolescents be allowed to participate in minimal-risk research, including confidential surveys, with a waiver of parental consent stating “for research of low risk (eg, confidential or anonymous survey research) . . . informed consent of the adolescent can be substituted for parental permission.” This recommendation has been debated in the research community and within research review boards because of misunderstandings about the legal statutes regarding adolescent consent, fear of litigation, state variability, or a perception that adolescents are not as capable as adults of giving consent.22 To our knowledge, there has been little systematic study of parental and adolescent perspectives on this important issue.
In the present study, we focused on risk behavior survey research involving adolescent minors and found that virtually all parents and adolescents believed that this research is important and should be done. Our research also revealed variability in attitudes regarding the need for parental consent by research topic, substantial differences between adolescents and parents, and substantial differences between adolescents presenting to the clinic with parents and adolescents presenting to the clinic without parents.
Most parents believed parental consent or permission is needed for adolescent participation in risk behavior research. Prior research33 has demonstrated parents' complex and sometimes conflicting responses related to confidentiality in clinical care for adolescents, particularly regarding contraception and reproductive health care. In our study, parents of male adolescents felt less need for permission than parents of female adolescents. However, parents vs other guardians were more likely to believe permission should be required. Parents who believed their adolescent was involved in risk behavior, like “had sex,” were more likely to endorse a requirement for parental consent. This may represent efforts by parents to monitor the behavior of an at-risk adolescent.
Adolescents were less likely than parents to endorse a requirement for parental consent on all topics. There was variability in adolescent attitudes depending on the topic, suggesting recognition of the differential impact of some of the issues. Only 37% and 16% of paired and solo adolescents, respectively, endorsed requirements for parental consent for all topics. Parent-adolescent differences were significant for research involving sexual activity or birth control. This difference in opinion on reproductive health care issues is consistent with the literature and supports research and clinical guidelines that protect adolescent confidentiality regarding these issues.13
The attitudes of teenagers presenting to the clinic were significantly different based on whether they presented to the clinic with (paired) or without (solo) their parent or guardian. Compared with paired teenagers, solo teenagers were less likely to endorse a requirement for parental permission. This may reflect the fact that solo teenagers were more likely to have a confidential health concern because they likely had a chief complaint regarding a medically emancipated condition. Participation of solo teenagers in research is a controversial issue on many institutional review boards.7 This study demonstrates that these solo teenagers had different views from those presenting with parents and underscores the potential bias introduced if they were not included in adolescent research.
For adolescents presenting to clinics with parents, African American race was associated with a decreased perceived need for permission for research after controlling for parental educational level. “Closeness to parent” was included in the survey and was not significantly associated with adolescent race or gender. This finding may be related to a more open relationship between African American adolescents and their parents in discussing these risk behaviors and, therefore, less concern regarding loss of confidentiality with required parental consent. Past studies24,34 have shown that African American adolescents are less concerned with disclosure of risk behaviors than their peers of other ethnicities.
Our study findings on parental and youth attitudes regarding parental permission for research are similar to those of other studies. Like Sikand et al,25 we found that most parents endorsed parental consent for surveys of adolescents involving sexual behavior and found differences in attitudes between parents and adolescents (parents, 83%; paired adolescents, 63%; and solo adolescents, 44%). The proportions of paired parents and adolescents endorsing a requirement for parental permission for research on the various risk behavior research topics were higher in this study than in previous studies.25 This may reflect differences between studies, with this survey's focus on risk behaviors and confidential surveys compared with previous work that addressed a broad array of research topics and methods.
There are limitations to our study. Most parents in our sample were female, most were African American, and all were recruited from clinic settings. Most solo adolescents were African American. These sampling characteristics may affect generalizability. Also, our study was conducted amid a time of much public debate regarding parental consent in pediatric and adolescent health care and pediatric research, including the case of CN v Ridgewood Board of Education6 and the cancelled risk behavior survey in northern Virginia public schools.17 Local and national media, lawmakers, and politicians have addressed these controversies. In addition, the study prompting Grimes v Kennedy Krieger Institute, Inc15 was conducted within the metropolitan area of the study sample. This may have biased results toward greater endorsement requiring parental permission. Another potential limitation may have been respondent confusion regarding clinical research vs confidential survey research. This may have biased results toward greater endorsement of parental permission; however, the study clearly defined survey research and emphasized confidential risk behavior surveys with no names used. Finally, the fact that parent-adolescent pairs were sampled concurrently could affect perceived confidentiality, although headsets were used and safeguards were developed to address this issue.
It is not clear what effect requirement of parental permission for adolescent involvement in risk behavior research would have on external research validity. It is probable that participation in this research would decrease because of logistic issues in contacting parents, parental refusal of participation, and adolescent refusal of participation related to confidentiality concerns. This could increase selection bias and decrease generalizability. In addition, it is possible that the truthfulness of adolescent responses could be influenced by knowledge that their parents are aware of their participation and concerns about disclosure of their responses. This study adds the adolescent and parent perspective on the issue of adolescent risk behavior research importance and parental permission, although further study is needed on the impact of parental consent on research participation and validity.
Parents and adolescents agree that risk behavior research is important and should be done. Policy decisions on whether parental consent should be required for adolescent risk behavior research must consider parental and adolescent views, the impact of required consent on the validity of research, the risk and benefits of the research, and the importance of this research for public health. Clearly, requirement of parental consent for confidential adolescent research would exclude some of the solo adolescents in this study who had different attitudes compared with adolescents presenting with their parents. This study contributes to our understanding of parental and adolescent views that may foster improved communication regarding research and consent. Better communication among parents, researchers, study subjects, and governing bodies is essential to improving adolescent risk behavior research and participation. Agreement on guidelines is needed for research that all agree is important for adolescent health.
Correspondence: Ryan H. Pasternak, MD, MPH, Department of Pediatrics, Louisiana State University, School of Medicine, 1542 Tulane Ave, Campus Box T8-1, New Orleans, LA 70112 (firstname.lastname@example.org).
Accepted for Publication: June 14, 2006.
Author Contributions: Dr Pasternak had full access to all of the data in the study and takes full responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Pasternak, Geller, and Cheng. Acquisition of data: Pasternak and Parrish. Drafting of the manuscript: Pasternak, Geller, and Cheng. Critical revision of the manuscript for important intellectual content: Pasternak, Geller, Parrish, and Cheng. Statistical analysis: Pasternak, Geller, and Cheng. Obtained funding: Pasternak, Geller, and Cheng. Administrative, technical, and material support: Pasternak, Geller, Parrish, and Cheng. Study supervision: Cheng.
Financial Disclosure: None reported.
Funding/Support: This study was supported by grant U48/CCU309674 through the Prevention Research Centers, Centers for Disease Control and Prevention; and by a Primary Care Faculty Development training grant from the Health Resources and Services Administration.
Role of the Sponsor: The funding bodies had no role in data extraction and analyses, in the writing of the manuscript, or in the decision to submit the manuscript for publication.
Disclaimer: The opinions expressed in this article do not necessarily reflect those of the Centers for Disease Control and Prevention or the Health Resources and Services Administration.
Previous Presentation: This study was presented in part as a platform presentation at the annual meeting of the Society for Adolescent Medicine; April 2, 2005; Los Angeles, Calif.
Acknowledgments: We thank Anne Duggan, ScD, Department of Pediatrics, The Johns Hopkins University, for her support and contributions toward this study; Kristen Ehrenberger, now with the University of Illinois School of Medicine, for her dedication to the early development of this project; Diane Markakis, MS, Department of Pediatrics, The Johns Hopkins University, for her assistance with data entry and analysis; Merisa Aranas Pasternak, MPH, Department of Medicine, Tulane University Health Sciences Center, for her critique and for providing a voice for the survey; and the research assistants, for their indispensable and appreciated work.
Pasternak RH, Geller G, Parrish C, Cheng TL. Adolescent and Parent Perceptions on Youth Participation in Risk Behavior Research. Arch Pediatr Adolesc Med. 2006;160(11):1159-1166. doi:10.1001/archpedi.160.11.1159