Objective. To examine the prevalence and quality of alcohol prevention services delivered to adolescents in the United States.
Methods. A national, stratified random sample of pediatricians and family practitioners was drawn from the American Medical Association Masterfile. The response rate was 63%, and the final sample size was 1842 physicians. Quality of services delivered was assessed by 2 scales: quality of screening (percentage of patients screened and depth of screening questions used) and quality of education (level of effectiveness of educational methods used and frequency of use).
Results. Although most physicians reported providing some degree of alcohol prevention services, their efforts were typically inconsistent, not in enough depth, and they failed to incorporate the most effective educational methods. Reported rates of universal screening and counseling were low, and younger adolescents were less likely to receive services. Physicians’ beliefs about their alcohol management skills and perceptions of resource availability were the most consistent correlates of higher quality service.
Conclusions. Prospective studies that elucidate the conditions under which individual physicians do and do not screen, as well as future efforts to educate physicians about the most effective brief intervention approaches, seem warranted.
- preventive health services
- physician’s practice patterns
- family medicine
- alcohol screening
- brief office counseling
Alcohol is the most commonly used drug among adolescents1,2 and is responsible for more mortality and morbidity in this age group than are all other drugs combined.3 Use typically begins early, with peak initiation during grades 7 through 9. By the 12th grade, 80% of high school seniors report having used alcohol, 62% report having gotten drunk, and 31% report heavy episodic use.1 Among adolescents who drink, 38% to 62% report having had problems related to their drinking,4 such as interference with work, emotional and psychological health problems, the development of tolerance, and the inability to reduce use.
Based on the demonstrated effectiveness of brief interventions for alcohol use in adults5 and other behavioral risk factors in adolescents,6 clinical practice guidelines recommend that primary care providers screen adolescents for alcohol use and provide preventive education and counseling.7–12 Alcohol use is considered an important pediatric concern by the American Academy of Pediatrics, who recommend that pediatricians provide alcohol screening and counseling to all adolescents as well as children in the upper elementary grades.7,8
Studies on the delivery of alcohol prevention practices to adolescents have typically found that physicians, including pediatricians, provide these services at lower than recommended levels,13–17 even for higher-risk patients such as pregnant adolescents18 and those with chronic illness.19 But the generalizability of these studies is limited, with 2 exceptions13,17—they have involved relatively small, regional samples.
There is also a need to examine the quality of the preventive health services delivered.20 Physicians often fail to use the most effective counseling and intervention methods,21,22 an important indicator of quality. However, none of the studies above provide information about the content of the educational interventions conducted.
The present study addresses these important gaps in our current understanding and represents the most comprehensive investigation conducted to date on physicians’ delivery of alcohol preventive services to adolescents. Using a national sample of primary care physicians, we examined the provision of a broad range of alcohol screening and counseling services to adolescents, as well as the quality of those services. Variations in service delivery and quality as a function of selected physician and practice characteristics were also examined.
Study participants were recruited by mail using a stratified random sample that was drawn from the American Medical Association Masterfile Registry. The Registry is the most complete list of physicians in the United States who have completed requirements for practicing medicine. The sampling frame was stratified by specialty, and equal numbers of primary care physicians in family practice and pediatrics were drawn.
Physicians were eligible for inclusion in the sampling frame if they were actively practicing in the United States, reported spending at least 50% of their time in direct patient care, and had a primary specialty in pediatrics or family practice. Physicians who reported having a secondary specialty were included only if the secondary specialty was one in which routine visits would be expected. These included the following specialties: Adolescent Medicine, Family Practice, General Practice, General Preventive Medicine, Gynecology, Internal Medicine/Pediatrics, Pediatrics, Public Health and General Preventive Medicine, and Sports Medicine (in Family Practice or Pediatrics). Board- and nonboard-certified physicians were eligible. Physicians who did not see adolescent patients for routine visits were eliminated, as were physicians who reported seeing, on average, <1 adolescent patient per week.
Physicians were mailed packets that contained a cover letter describing the study, the 7-page study questionnaire (in Scantron form), a stamped, addressed return envelope, and a $10 bill as a token of appreciation for their participation. A letter of endorsement from the American Academy of Pediatrics went out to pediatricians asking for their support; we were unable to include a corresponding letter from the American Academy of Family Physicians.
Reminder postcards were sent to all potential subjects 9 days after the initial mailing. Nonresponders were sent a second questionnaire 3 weeks after the initial mailing. Eight weeks after the initial mailing, return reply postcards were sent in an attempt to ascertain why physicians had failed to respond (ie, wrong address, felt they were ineligible) and to allow physicians to request another study packet.
Based on previous experience with the Masterfile, we anticipated that many of the people who failed to respond would be found to be ineligible for the study, had they responded. We therefore conducted subsequent follow-up efforts in an attempt to identify these individuals via strategic telephone follow-up. This was done for 3 categories of physicians who were likely to have high rates of ineligibility: 55 years old or older; practice in settings other than solo or group; and secondary specialties unrelated to adolescents.
Of the 4000 questionnaires mailed, 254 (6.4%) were undeliverable because the physician had moved and left no forwarding address, or was deceased. Of the remaining 3746 physicians who presumably received study materials, 1024 (27.3%) were found to be ineligible to participate based on study entry criteria, 1862 (49.7%) were eligible and returned surveys, 260 (6.9%) actively refused to participate, and 600 (16%) failed to respond to repeated requests for information. Among eligible physicians (n = 2722), the participation rate was 68.4%. We eliminated 20 of these because of excessive missing data, resulting in a useable sample size of 1842.
Given the relatively large sample size and our desire to detect clinically relevant results, levels for statistical significance were set at P < .001 for all analyses. Using this criterion, rates of study participation did not vary as a function of physicians’ age, graduation year, or geographic region. (Participants and eligible nonparticipants did not differ on age, graduation year, board certification, or public health region.) Females had higher participation rates than males (χ2 = 10.67 (1); P = .001), and pediatricians were more likely to participate than family practitioners (χ2 = 33.82 (1); P < .001).
The 7-page survey instrument included questions about the physicians’ sociodemographic background, personal alcohol use, training and experience, characteristics of his or her practice, beliefs and attitudes concerning prevention and alcohol use, and delivery of alcohol preventive services.
Physician and Practice Characteristics
Demographic characteristics included participants’ age, gender, and ethnicity. Participants’ personal alcohol use included the age at which the physician first tried alcohol without parental supervision and their current drinking status (nondrinker, drink rarely, drink moderately, drink frequently). Beliefs concerning adolescent alcohol management skills were assessed using a 4-item scale (Crohnbach’s α = 0.83), which consisted of items concerning: familiarity with alcohol abuse criteria, confidence with ability to diagnose alcohol abuse, sufficiency with training regarding alcohol problem management, and familiarity with methods to influence adolescent behaviors. Items were written as semantic differentials anchored by positive and negative modifiers, with 7-point response scales.
The American Medical Association Masterfile was used to assess selected training and experience factors: physicians’ specialty, whether s/he was board certified, and the year of graduation from medical school. Survey questions concerning practice characteristics included the following: the percentage of time spent in direct patient care, the total number of patients seen weekly, the number of adolescent patients seen weekly (<1, 1–5, 6–10, or >10), type of practice setting (private solo practice, private group practice, public/community clinic, free-standing health maintenance organization, or other institutional setting), the proportion of adolescent patient care revenues obtained under managed care contracts (none, <50%, ∼50%, >50%, all), and perceptions of access to alcohol-related assessment and counseling referral resources (either within the practice or the community). Masterfile data were used to ascertain whether the practice was located in a metropolitan or nonmetropolitan area.
Physicians’ Delivery of Preventive Services
Physicians were queried about the percentage of their adolescent patients who were asked about 1) having tried alcohol; 2) whether their friends drink alcohol, 3) whether any family members had alcohol problems; and 4) a history of riding in a motor vehicle with an intoxicated driver. A single item asked about the percentage of patients who were educated about the risks of alcohol. Physicians responded on a scale of percentages that were presented in deciles, from 0% to 100%. Separate estimates were elicited for new patients who were younger (ages 10–14 years) and older (ages 15–18 years) adolescents. Estimates were also obtained for younger and older adolescents who were established patients. Based on subsequent analyses showing little or no differences in practice patterns for new and established patients, these estimates were averaged and the combined data are reported.
Types of Information and Education Provided
Physicians were asked how often, in their educational and counseling efforts with adolescents, they used each of 10 approaches that differ in terms of their general effectiveness.23–25 At the lowest level of effectiveness were items that reflected risk-focused information provision and counseling: information on the general risks of alcohol use, risks of alcohol abuse, and the relationship of alcohol use and injury, as well as warnings about the risks of alcohol use, and the risks of drinking and driving. Advice giving, which is generally more effective than risk-focused intervention, was assessed by 2 items: advising drinkers to stop drinking and advising nondrinkers to maintain abstinence. At the highest level of effectiveness were items concerning the provision of information about alcohol-related situations the adolescent may encounter, normative information about the number of adolescents who do not drink, and practice using refusal skills. For each item, participants noted whether they never, sometimes, or frequently used the approach when providing information to and counseling adolescents.
Analyses are presented separately by specialty group because of significant differences in the demographic and practice composition for pediatricians and family physicians (Table 1). We also separate analyses of males and females, given the different gender distribution in the 2 samples and previous evidence of large gender variation in service delivery.26 Variations in preventive services delivery as a function of physician and practice factors were examined using multiple regression analyses. In these analyses, the effect of each factor was estimated after controlling for the effects of all other factors. Because of the number of analyses conducted, we report as significant only those differences with P < .01. We describe differences with P greater than .01 but less than .05 as “marginal” or “tending to predict” the outcome of screening or educating. We report the exact P value in these cases. Analyses were conducted using SPSS-PC (SPSS Inc, Chicago, IL).
Study participants were 1842 primary care physicians who were board-certified in pediatrics (n = 1008) or family practice (n = 834). The demographic composition and practice characteristic of the 2 specialty groups were significantly different (Table 1). About half of the pediatricians were males who were board-certified; most were white (71%) and in private practice (74%) in a metropolitan area (90%). Compared with pediatricians, family physicians were more likely to be board-certified (61%), female (75%), white (84%), in private practice (80%), and more recent graduates, but less likely to be practicing in metropolitan areas (72%). The practices of family physicians included larger proportions of Medicaid patients, whereas pediatricians obtained more of their patient revenues from managed care.
One fifth of the physicians surveyed (40%) saw 1 to 5 adolescent patients (ages 11–17 years) per week; 40% saw 6 to 15 adolescents per week, and 40% saw >15 teens per week.
An examination of the percentage of physicians who provide alcohol prevention services to all of their younger and older adolescent patients gives an indication of how many providers are following published guidelines (Table 2). Overall, rates of adherence are low. Adherence is highest for asking adolescents whether they use alcohol (23%–43% of pediatricians and 14%–27% of family physicians), and is lowest for asking the adolescent about riding under the influence of alcohol (8%–16% of pediatricians and 4%–7% of family physicians). Across all gender-specialty groups, physicians were more likely to ask adolescents the most basic screening question (ie, whether the adolescent drinks alcohol) than more extensive, in-depth screening questions (mean difference: 17%; [95% confidence interval: 16%–18%]; P < .001).
The proportion of younger and older adolescent patients receiving alcohol prevention services is disaggregated by physician gender and specialty in Table 3. Paired t tests showed that physicians in all gender-specialty subgroups (as well as in the full sample) screened a significantly larger proportion of older adolescent patients (P < .001) than younger adolescent patients (P < .001). Over the entire sample, mean differences in the proportion of older and younger adolescents being screened were: 23.8% (adolescent’s use); 15.9% (friends use); 10.9% (family use); and 14.7% (riding under the influence).
Significant gender differences also emerged. Male physicians reported screening adolescents less often than did female physicians (P < .001). This was true for pediatricians and family physicians, and for screening of younger and older adolescents. The only exceptions were as follows: there was a nonsignificant trend for gender differences in pediatricians’ screening for friends use in younger adolescents (P = .01) and in family physicians’ screening for family use in older adolescents (P < .004); there were no gender differences in pediatricians’ screening for riding under the influence in either younger or older adolescents.
The quality of physicians’ screening practices was based on the proportion of patients receiving basic and expanded screening and could range from low (level 1) to high (level 7). Physicians who conducted only the most basic screening (ie, asking whether the adolescent drinks) with a small (10%–30%) proportion of their adolescent patients were categorized as providing the lowest quality screening (level 1; 9.3% of sample). Physicians who provided basic screening as well as some expanded screening (eg, asking about friends use, family use, and/or riding under the influence) were assigned levels 2 to 7, depending on the proportion of patients receiving the different kinds of screening. Providing basic and expanded screening to a small proportion of patients accorded a level 2 score (10.8% of sample); when the proportion of patients receiving basic screening was moderate (between 40%–60%), the quality level assigned was 3 (14.6% of sample). Providing basic and expanded screening to a moderate proportion of patients accorded a level 4 score (12.1% of sample). Physicians who provided basic screening to a large (between 70%–100%) proportion of patients and expanded screening to a small proportion were assigned level 5 (20.6% of sample); when the proportion of patients receiving expanded screening was moderate the quality level assigned was 6 (13.7%). The highest quality level (7; 15.2% of sample) reflected physicians who conducted both basic and expanded screening with a large (70%–100%) proportion of their adolescent patients. Sixty-eight participants (3.7% of sample) had missing data that precluded the calculation of screening quality scores.
There was great variation in the quality of screening reported. Approximately one fifth of the sample could be described as providing relatively poor quality screening (levels 1 and 2), whereas almost 30% were screening at the highest levels (levels 6 and 7).
Most physicians in the sample reported educating at least some of their adolescent patients about alcohol risks (Tables 2–3). On average, physicians reported providing at least minimal alcohol education for significant numbers of their younger (37%–50%) patients and a majority (55%–68%) of their older adolescent patients. However, rates of educating were lower than screening rates (mean difference: 11.6%, [95% confidence interval: 12.7%–10.5%]; P < .001). Physicians in all gender-specialty subgroups educated a significantly larger proportion of older adolescent patients (P < .001) than younger adolescent patients (P < .001). Over the entire sample, mean differences in the proportion of older and younger adolescents being educated were 17.4%.
Physicians reported using a variety of approaches in their educational and counseling efforts with adolescents (Table 4). Providing information on the general risks of alcohol use, warnings about the risks of drinking and driving and advising drinkers to stop drinking were the most commonly used methods across the 4 gender-specialty subgroups. The least frequently used methods across all 4 groups were some of the more effective ones: the provision of normative information about peer drinking and practice using refusal skills.
We categorized the quality of physicians’ educational/counseling practices as ranging from low (level 1) to high (level 7) based on the level of effectiveness of the interventions used and their frequency of use (ranging from sometimes to frequently). This was done as follows. First, we determined the highest level of intervention effectiveness used by a physician. This resulted in 3 categories of physicians: those who provided risk-focused interventions only (quality level = 1; 2.1% of sample), those whose most effective interventions were advice-focused (quality level = 2; 14.5% of sample), and those who provided highly effective interventions, at least some of the time (quality levels 3–7). In this latter group, additional distinctions were made on the basis of the frequency with which highly effective interventions were used (sometimes, at least 1 on a frequent basis, >1 on a frequent basis), along with the frequency of use of advice-oriented interventions, (sometimes or frequently). Physicians who sometimes used highly effective interventions included some who sometimes offered advice-focused intervention (quality level = 3; 25.6% of sample) and some who frequently provided advice-focused intervention (quality level = 4; 19.9% of sample). Physicians at quality levels 5 (5% of sample) and 6 (10.5% of sample) used at least 1 highly effective approach on a frequent basis; those who also used advice on a frequent basis were designated as quality level 6. The highest quality ratings were assigned to physicians who frequently provided both advice-focused interventions, as well as 2 or more of the most effective interventions (17.9% of sample).
A minority (17%) of physicians reported never using the most effective methods (levels 1 and 2). Almost 30% were educating at the highest levels (levels 6 and 7). Approximately 42% of the sample failed to use any methods on a frequent basis (levels 1–3).
Correlates of Screening and Counseling
Multiple regression analyses were used to identify significant correlates of physicians’ screening and education practices. Separate analyses were conducted in each of the 4 gender-specialty subgroups and for each of 4 dependent variables: prevalence of screening; prevalence of counseling; quality of screening; and quality of counseling. Thus, 16 analyses were conducted. Ten predictor variables were entered into the regression equation: year of medical school graduation, board certification status, percentage of time spent delivering direct patient care, ratio of younger to older adolescent patients, personal beliefs about alcohol management skills, the presence of referral resources, whether the practice is located in a metropolitan area, proportion of revenue derived from managed care (<50%, 50%, >50%), proportion of Medicaid patients, and type of practice setting (coded as solo, large group, health maintenance organization, clinic, other).
In terms of screening and counseling rates, there was a significant (P < .001) relationship between the set of predictor variables and physicians’ service delivery for all 8 prevalence models tested; these are shown in Table 5. Multiple correlations were moderate in size, ranging from 0.356 to 0.544. Screening and counseling rates were significantly higher among physicians who had more positive beliefs about their alcohol management skills in all of the gender-specialty subgroups. Among all groups except female pediatricians, the availability of referral resources was associated with higher rates of screening and counseling; screening rates were also higher for physicians who saw a greater proportion of older adolescent patients.
Factors that emerged as significant, but inconsistent, predictors of higher screening and/or counseling rates included the following: more recent graduation from medical school, practicing in a clinic (vs private practice) setting, being board-certified, practicing in a metropolitan location, having a greater dependence on revenue from managed care, having a higher proportion of Medicaid patients, and spending a smaller proportion of time providing direct patient care. Among these inconsistent predictors, only the first 2 were predictive in more than a single group.
There was also a significant (P < .001) relationship between the set of predictor variables and the quality of physicians’ screening and counseling services delivery for all 8 quality models tested (Table 6). Multiple correlations ranged from 0.289 to 0.596. The quality of screening and counseling was significantly higher among physicians who had more positive beliefs about their alcohol management skills. The perceived availability of referral resources was a significant predictor of screening quality for everyone except female pediatricians. Other inconsistent correlates were identified for family physicians; no additional correlates emerged for pediatricians.
Summarizing across Tables 5 and 6, we can see that the prevalence models yielded a greater number of significant correlates than did the quality models. Factors that were correlated with higher rates of screening and/or counseling (in at least 1 group) but were not associated with the quality of services included: being board-certified, having a greater dependence on revenue from managed care, having a higher proportion of Medicaid patients, and spending less time providing direct patient care.
How well are physicians providing alcohol prevention services to their adolescent patients? On the positive side, it is encouraging to see that the typical physician in our sample reports providing some of these services to many, and in some cases a majority, of their adolescent patients. Most physicians engaged in at least some screening activity and those who never screened their adolescent patients represented a small minority. Furthermore, although participants in the study were more likely to screen adolescents than to educate them about alcohol risks, they did nonetheless report providing some form of preventive education or counseling for significant numbers of their younger patients and a majority of their older adolescent patients.
The results also point to areas where improvement is needed. Professional guidelines recommend that all adolescents be screened for whether they use alcohol, yet less than one quarter of physicians in our sample reach this standard. Recommendations for more extensive, in-depth screening are followed even less often. Screening for alcohol use in family and peers is important for early identification and intervention of at-risk adolescents,27 yet the typical physician performed such screening for a minority of adolescents. They were even less likely to query patients about their use of alcohol while in motor vehicles, despite the significant negative consequences of such behavior and its high frequency among older adolescents.3
The fact that most participants in our sample screened for alcohol use at least some of the time raises the question of why they do not screen more often, and in more depth. Time constraints are a frequently mentioned impediment in physicians’ practices28 and could certainly lead to inconsistent screening. The observed screening patterns may also reflect physicians’ concerns that more in-depth screening will increase the likelihood of detecting serious problems that they are unable to deal with and for which there are inadequate referral resources.29 Few primary care physicians are likely to have the time or skill for treating serious alcohol abuse problems, so resources for referring such patients are essential. We found that physicians’ perceptions that there were resources available to deal with alcohol managements problems were consistently associated with higher rates and quality of screening and counseling. Given the cross-sectional nature of our data, we cannot confirm the direction of the association. It is possible that physicians who deliver better services are also better informed about the resources available to them. However, a recent prospective study found that the presence of a health educator in a pediatric setting significantly increased physicians’ screening for behavioral issues,30 suggesting that the availability of resources does influence screening practices. If so, the general lack of availability of treatment resources for substance abuse and for mental health, which is a related factor in substance abuse, is a serious concern.31
Although most physicians provided some educational/counseling services, less than one-third adhere to recommendations that all adolescents be educated about the risks of alcohol use. This is regrettable, because adolescents express interest in discussing alcohol with their health care providers,32 and physicians are seen as credible and preferred information sources about health-related issues.33
Also discouraging were the low rates at which the most effective intervention methods were used. The success of any intervention depends on how well it is implemented and to insure quality, physicians need to know the best methods for intervening as well as having the skills to apply that knowledge.34 Our data support this conclusion; perceived skills were the single best predictor of service quality for all groups. However, physicians often report feeling they do not have adequate knowledge and skills for addressing and counseling patients about behavioral issues such as alcohol use, smoking, or sexual behavior.17,19,29,35,36
To address this issue, the Department of Health and Human Services has recommended that the basic curriculum in schools of medicine, nursing, and other health professional training schools include core competencies in health promotion and disease prevention.37 Significant efforts have taken place to define these competencies, and attention to prevention in medical school curricula has increased in recent years.38 Nevertheless, important gaps remain and will require attention if we are to be assured that the recommended competencies are achieved. For physicians in practice, training in continuing education venues could also be helpful to develop and sustain these competencies, to the extent that they include skills training and not just didactic material.39
Consistent with other studies on the delivery of other preventive services to youth, we found that younger adolescents were less likely to receive these services than older adolescents. Given the early age of onset of alcohol use, this is unfortunate. Use of alcohol is evident in 18% of 12- to 13-year-olds, and 44% of 14- to 15-year-olds2; by the eighth grade (ages 13–14 years), 52% will have used alcohol.1
Some physicians may be unaware of these statistics, in which case education regarding the early age of alcohol use onset is warranted. In an earlier study, we found that fears of offending parents influence the provision of STD preventive services to adolescents, and similar fears might operate in the area of alcohol use as well.40
A number of study limitations, beyond those referred to earlier, warrant mention. First, like most studies of physician behavior we relied on physician self-report to assess behavior and such data are subject to bias. Although both over- and underreporting have been found, the more prevalent and serious problem is overreporting,20 and we would guess that this is true here as well. If so, it would paint a more negative picture of the current state of service delivery than we have portrayed.
Comparisons between specialty groups must be tempered by the fact that pediatricians were sent a letter of endorsement from the American Academy of Pediatrics, whereas family physicians were not sent a similar letter from their professional organization. This could have introduced sampling bias, and it is almost certainly the reason for the differential response rates. In a previous study in which endorsement letters were included for both specialty groups, response rates were similar,41 and rates for family physicians in the current study seem lower than those previously reported.
Our measures of quality have some limitations. Counseling quality was restricted to an assessment of the content of the intervention, without more explicit descriptions of how those interventions were delivered. Screening quality assessments also reflected judgment calls on our part. For example, our algorithm gives more weight to the value of asking of in-depth screening questions than it does to the provision of basic services to a larger proportion of patients. Nevertheless, our efforts represent an early attempt to quantify the quality of services delivered, which is ultimately the essential question.
Our study was not designed to identify factors that influence the delivery of preventive services. As such, we examined only a small number of potentially important predictors. Much remains to be learned about the causal factors, and prospective studies will be required toward this end. Still, the few prospective studies available support our conclusions regarding the importance of skills40 and resources.30
Finally, it is important to note that although this study focuses on the prevention of alcohol use, providing comprehensive clinical preventive services to adolescents requires attention to a broader range of behaviors and concerns.10,11 However, alcohol use may be of particular importance given its association with other risk behaviors, and the fact that it is typically one of the earliest to emerge, serving as a “gateway” to more serious involvement with drugs.42 As such, an emphasis on alcohol prevention provides an opportunity to intervene with a behavior that may be less firmly established as part of a “lifestyle.” Furthermore, because of the considerable burden of suffering associated with alcohol use and its role in adolescent mortality, preventing use even infrequently is likely to have benefits to the population as a whole.9
Many of the necessary ingredients for facilitating the delivery of alcohol prevention services are at hand. A majority of adolescents see a physician at least yearly,43 and most of these visits are to pediatricians and general/family practice physicians.44 Alcohol prevention is ranked high in importance among both pediatricians and general/family practice physicians; they also view alcohol use as amenable to change.45 The fact that most physicians are able to incorporate at least some degree of alcohol prevention into their practice is also encouraging. Now the challenge is to find ways to improve the quality of these services by assuring the use of the most effective methods, delivering these more consistently, and to a greater proportion of younger and older patients.
This study was funded by the National Institute on Alcohol Abuse and Alcoholism (RO1AA10875, to Dr Millstein). Additional support for the authors was provided by the Maternal and Child Health Bureau of the Department of Health and Human Services to the University of California, San Francisco (2T71 MC00003).
- Received March 15, 2002.
- Accepted September 27, 2002.
- Reprint requests to (S.G.M.) University of California, San Francisco, School of Medicine, Department of Pediatrics, Division of Adolescent Medicine, 3333 California St, Suite 245, San Francisco, CA 94143-1236. E-mail:
This paper was presented in part at the Annual Meetings of the Society for Adolescent Medicine, March 22–26, 2000; Society for Pediatric Research, May 12–16, 2000; and the Association for Health Services Research, June 25–27, 2000.
- ↵Johnston LD, O’Malley PM, Bachman JG. Monitoring the Future: National Survey Results on Drug Use, 1975–1999. Bethesda, MD: National Institute on Drug Use; 2000
- ↵Substance Abuse and Mental Health Services Administration. National Household Survey on Drug Abuse: Main Findings 1996. Rockville, MD: Substance Abuse and Mental Health Administration; 1998. DHHS Publ. No. 98-3200
- ↵American Academy of Pediatrics, Committee on Substance Abuse. Role of the pediatrician in prevention and management of substance abuse. Pediatrics.1993;91 :1010– 1013
- ↵American Academy of Pediatrics, Committee on Substance Abuse. Alcohol use and abuse: a pediatric concern. Pediatrics.1995;95 :439– 442
- ↵US Preventive Services Task Force. Guide to Clinical Preventive Services. 2nd ed. Alexandria, VA: International Medical Publishing; 1996
- ↵Green ME, Palfrey JS. Bright Futures: Guidelines for Health Supervision of Infants, Children, and Adolescents. 2nd ed. Arlington, VA: National Center for Education in Maternal and Child Health; 2000
- ↵Elster AB, Kuznets NJ. AMA Guidelines for Adolescent Preventive Services (GAPS): Recommendations and Rationale. Chicago, IL: American Medical Association; 1994
- ↵National Institute on Alcohol Abuse and Alcoholism. Screening for Alcoholism in Primary Care Settings. Rockville, MD: Alcohol, Drug Abuse, and Mental Health Administration; 1987: 37–39
- Klein JD, Allan MJ, Elster AB, et al. Improving adolescent preventive care in community health centers. Pediatrics.2001;107 :318– 327
- ↵Kokotailo PK, Adger H Jr, Duggan AK, Repke J, Joffe A. Cigarette, alcohol, and other drug use by school-age pregnant adolescents: prevalence, detection, and associated risk factors. Pediatrics.1992;90 :328– 334
- ↵Children and Mental Health. Mental Health: A Report of the Surgeon General. Washington, DC: US Department of Health and Human Services; 1999:123–220
- ↵Levenson PM, Morrow JR, Morgan WC. Health information sources and preferences as perceived by adolescents, pediatricians, teachers and school nurses. J Early Adolesc.1986;6 :183– 195
- ↵US Department of Health and Human Services. Healthy People 2010: Understanding and Improving Health. 2nd ed. Washington, DC: US Government Printing Office; November 2000
- ↵Ziv A, Boulet JR, Slap GB. Utilization of physician offices by adolescents in the United States. Pediatrics.1999;104 :35– 42
- ↵Schneider D. Setting priorities for children’s health: viewpoints of family physicians and pediatricians. J Am Board Fam Pract.1994;7 :387– 394
- Copyright © 2003 by the American Academy of Pediatrics