Background. Little is known about availability of services and confidential care for adolescents in primary care practices or how availability among pediatric practices compares to that among other primary care practices. The objective of this study was to assess self-reported availability of services for medically emancipated conditions and confidential care in primary care practices, to compare physician responses to those from office staff who answer appointment lines, and to compare availability in pediatric practices to other primary care practice types.
Methods. We conducted a telephone survey of randomly selected practices from the Washington, DC, metropolitan area in pediatrics (Peds), internal medicine (IM), and family medicine (FM). We asked staff who answer appointment lines about availability of services for medically emancipated conditions and confidential appointments for adolescents. Physicians received the same questions via a mail survey. Responses from office staff and physicians in the same practice were linked for comparison.
Results. Of 434 practices contacted by telephone, 372 (86%) responded. Of the 615 physicians surveyed from these 372 practices, 264 (43%) from 170 practices responded to the mail survey. Peds practices were less likely than FM and IM practices to offer services for medically emancipated conditions and were less likely than FM practices to offer confidential services to adolescents. Office staff and physicians from FM and IM had higher agreement compared with Peds about availability of services for medically emancipated conditions. Agreement between office staff and physicians about provision of confidential appointments to adolescents was low among all practice types. However, having a written office policy on adolescent confidentiality was significantly associated with agreement between office staff and physicians about availability of confidential services.
Conclusions. Care for medically emancipated conditions and confidential services for adolescents are limited among primary care practices, especially among pediatric practices. All primary care practice types had significant disagreement between office staff and physicians about availability of confidential services to adolescents. Adolescents who call appointment lines are likely to receive inaccurate information about confidentiality policies. Establishing written office policies on adolescent confidentiality may help to improve access to confidential care for adolescents.
Adolescents have identified lack of confidentially as a barrier to seeking health care; they are more willing to seek care from and communicate with physicians who assure confidentiality1–3 and may forgo health care to prevent their parents from finding out.3–5 Each state in the United States legally entitles adolescents to consent to treatment for medically emancipated conditions that may include contraception, pregnancy, diagnosis and treatment of sexually transmitted diseases (STDs), human immunodeficiency virus or other reportable diseases, treatment of substance abuse problems, and mental health.6–9 Having the legal right to consent to care is closely related to being guaranteed confidentiality—adolescents who are considered cognitively mature enough to give consent are also granted the right to patient confidentiality.6 However, the availability of confidential care for medically emancipated conditions among primary care practices is not well-characterized.
Although pediatricians recognize adolescents as their patients, studies have found that a high proportion feel uncomfortable with providing services for medically emancipated conditions and/or providing confidential care.10–14 Despite the 1978 recommendations by the Task Force on Pediatric Education to improve training for adolescent health care,15 there is evidence that pediatricians still lack confidence in their ability to address adolescent issues and remain uncommitted to providing comprehensive care to adolescents.10,16 Furthermore, it is unclear how service and confidentiality for adolescents in pediatric practices compares to that in other types of primary care practices.
We conducted a survey in the Washington, DC, metropolitan area to assess availability of services for emancipated conditions to adolescents, the ability of an adolescent to receive care on a confidential basis, and differences among primary care practice types. Because agreement of the entire office staff about confidentiality policies is essential to ensure confidentiality,13 we surveyed both physicians and office staff who answer appointment lines. Obtaining responses from office staff and physicians in the same practice allows comparison of actual physician policy to information that patients receive when they call appointment lines. We asked about services for pelvic examinations, contraception, and STD testing. The jurisdictions in the metropolitan areas included in this study (Virginia, Maryland, and Washington, DC) all have statutes that allow minors to consent for these services.9
Physician offices that provide primary care within a 25-mile radius of Washington, DC, were identified through listings in the Washington Physicians Directory.17 The listings in this private publication are compiled from medical society rosters, the yellow pages, and requests from practitioners for inclusion. Only physicians in primary care specialties were surveyed. A medical practice was defined as any number of practitioners operating at a single site. The list of practices was divided into pediatric and adolescent medicine (Peds; 270 practices), family medicine (FM; 305 practices), and internal medicine (IM; 442 practices), depending on the specialty of the practitioners. Practices with physicians from multiple primary care specialties were excluded because of their small number (17 practices).
The telephone survey was conducted with office staff who answered appointment lines between February and July 1998. To determine the necessary sample size to attain an 80% power in discerning differences among practice types in responses about available services, we performed a pilot telephone survey of a randomly chosen subgroup of 50 practices. The final study sample comprised 481 practices. Each practice was called until a response or refusal was obtained or until it was determined that the practice was ineligible (47 were retired physicians or not a clinical practice). Identifying information was stripped from responses after participation, and each practice was assigned a study number.
For the mail survey, conducted between November 1998 and February 1999, all physicians from practices that completed the telephone survey were mailed questions identical to those in the telephone survey. The mail survey was pilot-tested among 20 physicians at our institution to ensure good comprehension. Three successive mailings were undertaken to maximize response rates. Physician responses were anonymous and identified only by the study number assigned to their practice. Using the study number, we linked responses from the telephone and mail surveys so that the practice became the unit of analysis. A total of 170 practices participated in both the telephone and mail surveys. Analysis was conducted with data from the 137 practices that had at least 1 physician who reported seeing adolescent patients (defined as at least up to 18 years of age) and with no missing office staff or physician responses to questions about service provision. To measure agreement between office staff and physician responses, we scanned all physician responses for each practice to determine whether any physician responded that he or she provided the service. When any physician responded “yes,” the physician response for that practice was categorized as “yes.” We assessed agreement between office staff and physician responses for each practice type by calculating percentage agreement “yes,” percentage agreement “no,” and percentage of discordant responses. To assess differences among practice types in rates of agreement, we used the Breslow-Day test for homogeneity of odds ratios and reported whether the P value was <.05.
Respondent characteristics were determined from survey questions and from information printed in the Washington Physicians Directory (medical school graduation date and board certification status). We calculated the mean number of decades since medical school graduation and the proportion of board-certified physicians for all physicians in each participating practice regardless of whether all physicians in that practice had participated in the mail survey.
We constructed logistic regression models to assess the association between availability of services and practice characteristics. In 5 separate models, availability of service to adolescents (for pelvic examination, contraception, STD testing, confidential contraception, and confidential STD testing) was the dichotomous dependent variable. The following independent variables were included on the basis of findings in previous studies10,12,14,16,18,19 and entered simultaneously: >50% of physicians in the practice were board certified (dichotomous), mean number of decades since physician graduation from medical school, whether the practice typically saw >5 adolescent per week (dichotomous), existence of a specific office policy on adolescent confidentiality (dichotomous), 4 dummy variables combining solo practitioner status and presence of at least 1 female physician in the practice, and 3 dummy variables for primary care practice type. We also used logistic regression models to assess the association between practice characteristics and agreement between office staff and physicians about availability of confidential services. Agreement between office staff and physicians in the same practice about confidential contraception and confidential STD testing was the dichotomous dependent variable (both “yes” or both “no” responses versus discordant responses) in each of 2 models. The independent variables were the same as above. For each model, the Hosmer and Lemeshow goodness-of-fit χ2 test was used and confirmed the null hypothesis that the data fit the model in each case. The Children’s National Medical Center Institutional Review Board approved the study.
The response rate for the telephone survey of office staff was 86%. The response rate for the mail survey of physicians from the 372 practices that responded to the telephone survey was 43%, ranging from 30% among IM physicians to 68% among pediatricians (Table 1). A total of 170 practices had both office staff and physician responses that could be compared. Table 2 shows respondent characteristics for each practice type. For the telephone survey, IM practices were less likely than Peds and FM practices to see adolescent patients (P < .0001). Overall, 88% of respondents were nonphysician office staff. FM practices were significantly more likely to have a medical assistant or physician answer the telephone survey (P = .006). We did not exclude practices for which physicians answered the telephone survey because it was likely that physicians regularly answered appointment lines in those practices. IM practices were more likely to have physicians who graduated from medical school 2 decades ago or less (P = .03). Among physicians who responded to the mail survey, IM physicians were least likely to see adolescent patients (P = .02). Pediatricians and FM physicians were significantly more likely to see greater numbers of adolescent patients per week (P < .0001).
The remainder of the analysis included the 137 practices in which at least 1 physician saw adolescent patients and for which there were no missing responses for questions about provision of services. When asked about services available, office staff and physicians from FM and IM practices were more likely than those from Peds practices to say that pelvic examinations, contraceptive services, and STD testing were available to adolescents (Table 3). For example, 97% of FM physicians and 79% of IM physicians provided pelvic examinations to adolescents compared with 50% of pediatricians. Differences among practice types for both office staff and physicians were statistically significant. When physician and office staff responses from the same practice were compared for all services, 16% to 39% of practices gave discordant responses. The highest level of disagreement was between pediatricians and their office staff about whether STD testing was available to adolescents at their practice.
Next we analyzed confidential services among the 92 practices whose physicians offered services for medically emancipated conditions (Table 4). Although office staff and physicians from IM practices were the least likely to say that confidential services were available, the difference among practice types did not reach statistical significance. Disagreement between office staff and physicians about providing confidential services (Table 4) was substantially higher among all practice types than for provision of services (Table 3). Between 45% and 63% of practices had discordant responses between office staff and physicians about provision of contraceptive and STD services to adolescents without parental knowledge.
When the overall proportion of practices providing confidential services for medically emancipated conditions is compared among practice types, the greatest proportion of FM practices offered such services (data not shown). Compared with 16% of all Peds practices and 6% of IM practices, 29% of FM practices had office staff and physicians who agreed that contraceptive services were available at their practice and that an adolescent could make an appointment for contraception without parental knowledge (P = .02). The differences among practice types for confidential STD testing were not statistically significant: 23% of Peds and 17% of IM practices versus 31% of FM practices had office staff and physicians who agreed that STD testing was available and that an adolescent could make an appointment for testing without parental knowledge (P = .28).
To assess the association between provision of services to adolescents and practice characteristics, we used logistic regression models for each of 5 services asked about in the survey: pelvic examination, contraceptive services, STD testing, confidential contraceptive services, and confidential STD testing (Table 5). The adjusted odds ratios (AORs) for which the 95% confidence interval (CI) excludes the value of 1.0 are shown in bold type. Three factors were associated with providing services for medically emancipated conditions and providing confidential service. First, practice type had the strongest association with providing pelvic examinations, contraception, and STD testing to adolescents. Physicians in FM practices were at significantly higher odds than those in Peds practices to offer all of the services about which we asked and to provide them confidentially. Physicians in IM practices were at significantly higher odds than those in Peds practices to offer pelvic examinations and STD testing and to offer STD testing confidentially. Second, practices that usually saw >5 adolescents per week were at significantly higher odds of providing STD testing and confidential STD testing to adolescents than practices with a lower volume of adolescent patient visits. Having >50% board-certified physicians in a practice was also associated with providing services to adolescents but only significantly with confidential contraceptive services. Third, when solo practitioner status and practitioner gender were considered, group practices with at least 1 female physician (referent group) were at higher odds of providing all 5 services than solo practices with either a male or a female physician and group practices with no female physicians, although the 95% CI of the AOR included 1.0 in some cases.
We also analyzed the association between practice characteristics and agreement between office staff and physicians in the same practice about availability of confidential services. For both models, only having >50% board-certified physicians and having a written office policy on adolescent confidentiality were significantly associated with agreement about confidential services. For agreement about whether adolescents could get contraceptive services without parental knowledge, having >50% board-certified physicians was negatively associated with an AOR of 0.1 (0.02–0.8) and having an office policy positively associated with an AOR of 3.5 (1.2–10.6). For agreement about confidential STD testing for adolescents, having >50% board-certified physicians had an AOR of 0.1 (0.01–0.7) and having an office policy had an AOR of 4.5 (1.5–13.6).
Respondents were also asked about other aspects of providing services that may improve access for adolescents (Table 6). Respondents from IM practices were least likely and those from Peds practices most likely to say that same-day urgent appointments were available. Peds practices had the lowest disagreement between office staff and physicians about availability of same-day appointments. A low percentage of office staff and physicians from all practice types said that the fee required at the time of service for adolescents without health insurance was lower than $50 or that their practice offered a sliding-scale fee schedule based on ability to pay. In general, FM practices seemed to have lower up-front fees for adolescents without insurance, although the differences among practice types was not always statistically significant. There was also a high level of disagreement between office staff and physicians about amount of payment required for uninsured patients and whether a sliding-scale fee schedule was available: discordance rates ranged from 32% to 56%. Finally, when asked whether the practice had a specific office policy on adolescent confidentiality, respondents from Peds practices were most likely to say yes. IM respondents were more likely to say no but had the lowest percentage of discordant responses.
Table 7 shows reasons given by pediatricians for not providing pelvic examinations or contraceptive services to adolescents. The most common reasons for not providing pelvic examinations were lack of equipment and expertise, low patient demand for this service, and inadequate staffing. Pediatricians who did not provide contraceptive services were most likely to respond that they did not provide pelvic examinations (and therefore could not provide the medical surveillance necessary for contraceptive services), that they lacked expertise, and that patient demand for contraception was low.
Providing confidential care to adolescents is challenging. The Society for Adolescent Medicine6 has called for health providers to inform patients and families about the requirements of confidential care and to strike the difficult balance between respecting an adolescent’s wishes about sharing information and involving responsible adults when necessary. The tumultuous nature of adolescence and the varied stages of autonomy among adolescents pose challenges. In addition, practitioners face barriers such as limited time for office visits; lack of training in adolescent issues; difficulties in keeping billing and medical records confidential; and private, public and political debates about confidential health care for adolescents.6
Our results show that care for medically emancipated conditions for adolescents is not universally available in primary care practices in metropolitan Washington, DC, and that confidential care is even less accessible to adolescents. This is especially true in Peds practices, a finding supported by previous studies.14,20 The most common reasons given by pediatricians in our study for not providing pelvic examinations and contraceptive services are lack of expertise and equipment and low patient demand. Given the high rates of sexual activity, pregnancy, and STDs among adolescents,21 the perception of low patient demand is probably attributable to the fact that adolescent patients perceive that pediatricians are unable or unwilling to provide the needed services and seek these services elsewhere.20,22 It may be that pediatric training in gynecologic skills remains insufficient, that pediatricians have less opportunity to use these skills in practice, or both. Few pediatricians cited being uncomfortable with the family negotiations involved or being reluctant to see adolescent patients as found in previous work.12 Because pediatricians are positioned to guide patients and families through the sensitive issues that arise in adolescence, they should enter practice well-prepared to exercise the spectrum of skills necessary for this role. Past work has shown that training in adolescent medicine during and after residency is associated with physicians seeing more adolescent patients and that continuing medical education was associated with increased provision of services for medically emancipated conditions.14 Our study reinforces the need to provide both adolescent medicine training in pediatric residency and continuing medical education in treating adolescents.
The multivariate analysis revealed that primary care practice type had the strongest association with provision of services and with provision of confidential care. Other work has shown that a high percentage of FM physicians deliver reproductive preventive services: >70% reported asking adolescent patients about contraceptive use, condom use, and sexual relationships and regularly discussing confidentiality with their adolescent patients.18 We also found that seeing >5 adolescents per week was associated with providing STD services and providing confidential STD testing. Belonging to a group practice of male practitioners or being a solo practitioner (regardless of gender of practitioner) was negatively associated with providing services and providing confidential services. This pattern may indicate that other barriers exist to increasing access to care for adolescents. For example, solo practitioners may face a larger opportunity cost than group practitioners of spending nonreimbursed time to counsel parents and patients about providing confidential care to adolescents. Practitioner gender may pose a barrier for female patients who may be apprehensive about being examined by a male physician, or, as shown in previous studies,10,14,16 male physicians may feel less confident in providing pelvic examinations and contraceptive counseling. The mean number of decades since medical school graduation for all physicians in a practice had no significant association with provision of services, although our measure was crude. Another concerning finding is the large degree of disagreement between office staff who answer appointment lines and physicians in the same practice about the availability of confidential care for medically emancipated conditions to adolescents. Other studies have shown that a substantial proportion of adolescents wish to be seen without their parent’s knowledge.3–5 Information that they receive about confidentiality when trying to schedule an appointment may be a deciding factor in whether to seek health care. To ensure that adolescents receive needed care, office staff should be fully informed of services available and the office policy on adolescent confidentiality13 and also have specific referral information if the services that adolescent patients seek are not available at the practice. Our results suggest that having a written office policy on adolescent confidentiality is an important way to ensure consistency within a practice. Ideally, such a policy should establish that providers inform adolescents and their parents about the process of developing a provider-patient relationship as the adolescent gains ability to make independent decisions. Providers should also discuss the conditions under which information will be shared with others, such as suspected physical or sexual abuse or when the adolescent poses a severe risk of harm to her- or himself or to others. The adolescent’s wishes on how the information will be shared should be respected.6 Ongoing education about laws and regulations about adolescent confidentiality should be included in the policy.
The responses to our questions about fees required up front and sliding-scale fee schedules suggest that payment may pose additional barriers both to obtaining care and to keeping the care received confidential. Adolescents who seek care without insurance or outside their parent’s insurance plan may face high payments up front; uninsured adolescents report missing care because of high cost.4 Although statutes may authorize adolescents to consent to their own care, provisions are rarely made for making services financially accessible for adolescents. Agreements about payment for services among practitioners and adolescents and their parents are important in increasing access to confidential care.6
The generalizability of our study is limited by a number of factors. First, the survey was limited to the Washington, DC, metropolitan area. There may be large regional variation in availability of confidential services based on state law, the supply of primary care providers, availability of alternative sites for health care, and local attitudes toward providing confidential services to adolescents. Second, the response rate of physicians to the mail survey was low, particularly for FM and IM physicians. Also, because of limited physician participation, sample sizes for each practice type were low for analyses of confidential services. The power of our study is therefore limited. Third, a large health maintenance organization in the Washington, DC, area declined to respond to our survey. Thus, a large provider of primary care was not represented in the study. We cannot compare the sample population to the physician population of the area as a whole because we did not collect information on nonresponding physicians. However, among the 86% of sampled practices that responded to the telephone survey, we compared characteristics of those with responding and nonresponding physicians. Practices with responding (n = 170) versus nonresponding (n = 202) physicians had the same percentage of physicians with 2 or fewer decades since graduation (38%). However, practices with nonresponding physicians were less likely to see adolescent patients (73% vs 84%) and less likely to have >50% of its physicians board certified (73% vs 85%). These differences may have biased our findings, but we surmise that access for adolescents in the practices with nonresponding physicians may have been even more restricted than in practices with responding physicians. Fourth, because clinics such as school-based health clinics and family planning clinics are not included in the Washington Physicians Directory, it is not possible to use the survey results to assess overall availability of confidential care in the Washington, DC, area. Finally, another limitation is possible bias as a result of social desirability in survey response. This bias would likely result in overestimation of service availability and may account for some of the discordance between the telephone and mail surveys.
It has long been clear that clinical services are an important component in health-promotion and disease-prevention efforts required to address STDs, unintended pregnancies, and other health problems among adolescents.23 Confidentiality is essential to providing such services to adolescents.6 Ongoing changes may be required in pediatric residency training and continuing medical education to improve health care for adolescents and to ensure that pediatricians feel confident in treating adolescents. Pediatric practices that are unable or unwilling to provide these services must be able to screen for health risk behaviors and to offer adolescents a referral to a specific practice or specialist who is proficient in providing the needed care. Establishing an office policy and ensuring that all office staff are knowledgeable about the policy may help to ensure greater access for adolescents to these services.
We thank Kanti Patel, PhD, for statistical assistance and Larry D’Angelo, MD, MPH, Julia Rhodes, PhD, and Ken Schoendorf, MD, MPH, for critical review.
- Received May 7, 2002.
- Accepted September 4, 2002.
- Address correspondence to Lara J. Akinbami, National Center for Health Statistics, 6525 Belcrest Rd, Rm 790, Hyattsville, MD 20782. E-mail:
At the time this work was started, the authors were affiliated with Children’s National Medical Center, George Washington University, Washington, DC.
- English A. Treating adolescents: legal and ethical considerations. Adolesc Med.1990;74 :1097– 1112
- ↵Boonstra H, Nash E. Minors and the right to consent to health care. The Guttmacher Report on Public Policy, August 2000. Available at: www.agi-usa.org/pubs/ib_minors_00.pdf
- ↵American Academy of Pediatrics, Task Force on Pediatric Education. The Future of Pediatric Education. Evanston, IL: American Academy of Pediatrics; 1978
- ↵Chastain DO, Sanders JM, DuRant RH. Recommended changes in pediatric education: the impact on pediatrician involvement in health care delivery to adolescents. Pediatrics.1988;82(suppl) :469– 476
- ↵The Washington Physicians Directory. Silver Spring, MD: National Directories, Inc; 1997
- ↵MacKay AP, Fingerhut LA, Duran CR. Adolescent Health Chartbook. Health, United States, 2000. Hyattsville, MD: National Center for Health Statistics; 2000
- ↵United States Congress, Office of Technology Assessment. Adolescent Health: Summary and Policy Options, I. Washington, DC: US Government Printing Office; 1991. Publ. No. OTA-H-468
- Copyright © 2003 by the American Academy of Pediatrics