Adolescent Substance Use Disorder in Primary Care: Challenges in Treatment Referral Beyond Access Availability

Abstract

Objective:

Fatal overdoses are the third leading cause of death in the pediatric population. Substance use disorders (SUD) screening is not routinely done in primary care practices. Early screening and intervention for adolescent SUD could mitigate future harm.

Methods:

We conducted a 3-month pilot adapting universal screening using the CRAFFT tool in patients aged 12 to 17 presenting to an urban and a rural primary care practice during well-child and acute/sick-child visits. We collaborated with our pediatric addiction service to ensure access availability for further assessment and treatment for all positively screened patients; this was broadly communicated to primary care providers.

Results:

There was a higher CRAFFT completion rate in the urban site (90%, vs 52.6% in our rural site). The majority of CRAFFT questionnaires were completed during acute/sick-child visits in both study sites. Moreover, we found a higher positive screen rate in our rural practice (14.6%, vs 2.4% in our urban practice). Only 27% of positively screened patients had substance use addressed by their providers. No pediatric addiction referrals were made.

Conclusions:

Findings suggest provider-level barriers exist despite having adequate specialty referral sources and institutional encouragement. Future work is needed to explore these barriers.

Introduction

The rate of fatal drug overdoses in the United States has exceeded 100 000 since 2021, the majority involving opioids. ¹ Of particular concern is the rise in overdose deaths among adolescents since 2019, which now constitutes the third leading cause of death in the pediatric population.^2,3 Prior research demonstrated that the majority of adolescents with severe substance use disorder (SUD) at age 18 continue to have symptoms of SUD in adulthood. ⁴ Adolescent drug use is associated with high-risk sexual behaviors, unintended pregnancies, exposure to violence, and increased risk of mental health problems.^5,6 Early identification and treatment for adolescent SUD could mitigate harm later on in life.

Although screening, brief intervention, and providing low-barrier, evidence-based treatment have all been recommended,^7,8 there remains a paucity of standardized guidelines. The United States Preventive Services Task Force’s most recent recommendation summary states that “current evidence is insufficient to assess the balance of benefits and harms of screening” in adolescents. ⁹

Guidelines from our institution recommend annual screening for drug use beginning at age 15 during primary care well-child visits using the CRAFFT (Car, Relax, Alone, Forget, Friends, Trouble) tool, which is a self-administered questionnaire taking approximately 5 min to complete through the patient portal or at the time of rooming. ¹⁰ Our internal audit revealed a screening rate between 39.7% and 63.4% across our Family Medicine and Pediatric practices between September 2022 and March 2023. Our practice encompasses 50 clinics in southern Minnesota and western Wisconsin covering one major urban center (population greater than 100 000) and multiple rural sites. Positive screening rates ranged between 6.7% and 10% in this period.

Our goal was to improve SUD screening rates by incorporating the CRAFFT in all visit types (acute/sick child visits in addition to well-child visits) through a 3-month pilot study in selected primary care practices. We hypothesized that by adding CRAFFT as a screening metric in all visit types, we would increase the total number of CRAFFT screens completed by eligible patients since certain patient groups, such as Hispanic and black children, have historically not attended well-child visits as routinely compared to other groups. ¹¹

To ensure positively screened patients can get prompt referrals for further assessment and treatment, we partnered with our institution’s pediatric addiction service to create an expedited referral workflow and confirmed adequate capacity for all referred patients to be assessed virtually or in-person within 1 week.

Methods

Implementation Strategy

Based on previous literature we anticipated multiple potential barriers to screening substance use with adolescent patients, including inadequate time, lack of training on addressing positive screens, and other competing priorities. ¹² Communication apprehension related to self-efficacy or previous negative experience with adolescent patients and/or their parents in discussing substance use has also been suggested. ¹³ Studies on adult patients with opioid use disorder have shown that inadequate provider mentorship, lack of referral sources and perceived institutional support as significant barriers.^14,15

Based on a modified Consolidated Framework for Implementation Research (CFIR) framework, ¹⁶ we selected implementation strategies, as detailed in the sections below, to facilitate a multifaceted approach to overcome these barriers that focused on stakeholder interrelationship, stakeholder education and clinician support.

Study Sites

Our urban site and one of our rural sites in southeast Minnesota were selected for this 3-month pilot. In each site, 1 family medicine and 1 pediatric practice were selected. Two provider champions were selected from the urban and rural sites to promote awareness of our project in their practices.

The urban site consists of 5 family physicians, 5 family medicine advanced practice providers (APP), 1 pediatric physician, and 2 pediatric APPs. The rural site consists of 6 family physicians, 5 family medicine APPs, 2 pediatric physicians, and 1 pediatric APP.

The urban practice pilot took place between October 2 and December 29, 2023. The rural practice pilot took place between November 6, 2023 and February 2, 2024. Approval was obtained from medical and nurse leadership from each participating site. A full description of our study goals, duration, and work involved was distributed to providers, nurses, and patient rooming staff in each participating site before the pilot study through in-person meetings and electronic communication.

Before the study, we consulted with our institution’s legal service to clarify the legal status in our state (Minnesota) regarding minor patients’ ability to provide informed consent without parental or legal guardian’s permission, as well as parental access to records when minors consent to treatment. We made this information widely available to providers in our pilot sites, as we anticipated questions and concerns about patient’s privacy and willingness to disclose their substance use and treatment if their parents or legal guardians were to have full knowledge of their drug use.

Per our legal department’s advice, any minors in our state may give effective consent for medical, mental health, or other health services for alcohol and other drug use. In addition, when a minor has consented to SUD treatment, the health records cannot be released to parents/guardians unless, in the judgment of the healthcare professional, failure to inform parent or guardian would seriously jeopardize the health of the minor patient.

CRAFFT Questionnaire

This is a validated 9-item self-administered questionnaire used to screen for substance misuse in adolescents. ¹⁰ The first 3 items ask respondents to provide the number of days in the past 12 months they consumed alcohol, cannabis, and any other substances to achieve euphoric effects. The final 6 items are yes/no questions. A “yes” to any of these items yields a score of 1. A score of 2 or higher indicates a positive screen, and is associated with an increased risk of having a substance use disorder diagnosis, placing the adolescent in an at-risk category. ¹⁷

Patient Screening and Workflow

All patients between the ages 12 to 17 presenting to the study sites during the pilot period were asked to complete the CRAFFT on an electronic tablet during rooming. Providers were notified of a positive screen (score of 2 or higher) by rooming staff.

Information was made available to providers of additional community resources for addictions treatment which could be directly accessed through the electronic health record (EHR), as well as an internal referral pathway to pediatric addiction service. Figure 1 outlines our workflow.

Figure 1.

Community Resources through FindHelp (can access directly through Electronic Health Record).

Data Analysis

Descriptive data was obtained on the total number of patients aged 12 to 17 seen in each pilot site during the study period, the number and percentage of patients in each site completing the CRAFFT, and the number and percentage of patients with a positive CRAFFT screen in each site.

A chart review of all positively screened patients was completed to identify if a positive screen was acknowledged and if subsequent action was taken and documented by the provider. Additionally, the number of referrals to addiction services, the number of appointments to addiction services scheduled after referrals were made, and the number of patients who showed up for their appointments with addiction services were tracked.

Chi-square test for homogeneity of proportions was used to determine if a significant difference exists in the proportion of positively screened patients (CRAFFT score ≥ 2) between the urban versus rural site.

Our project was exempt from Institutional Review Board review, and in accordance with the Code of Federal Regulations, 45 CFR46.102.

Results

In our urban site, 231 patients aged 12 to 17 were seen during the pilot period and all patients were given the opportunity to complete the CRAFFT; 208 patients completed the CRAFFT (90.0% completion), 88 of those completed questionnaires were administered during well-child visits (42.3%) and the remainder during acute/sick-child visits. Of those completed, 5 patients scored 2 or higher (2.4% positive screen rate), Figure 2. Only 1 of these positive screens came from well-child visits; the remaining 4 positive screens came from acute/sick-child visits. The median age of those positively screened was 15 (3 females; 2 males).

Figure 2.

CRAFFT Completion Rates and Positive Screening Rates in Urban versus Rural Sites during Well-Child and Acute/Sick-Child Visits.

In the rural site, 95 patients aged 12 to 17 were seen during the pilot period and all were given the opportunity to complete the CRAFFT; 50 patients completed (52.6% completion), 17 of those completed questionnaires were administered during well-child visits (34%) and the remainder during acute/sick-child visits. Of those completed, 6 patients scored 2 or higher (14.6% positive screen rate). Median age of those positively screened was 16 (4 females; 2 males). Two of the positive screens came from well-child visits; the remaining 4 positive screens came from acute/sick-child visits.

Chi-square test for homogeneity of proportions confirms that the proportion of positively screened patients in the rural site is higher compared to the urban site, χ² = 13.172, df = 1, P = .0003.

Chart review of the 11 positively screened patients revealed substance use to include alcohol, cannabis, and cocaine. The primary care pediatric or family medicine provider acknowledged the positive CRAFFT screen and addressed substance use during the appointment in 3 of these patients (27%); mental health concerns were addressed without addressing substance use or the positive CRAFFT screen in 5 patients (45%); and the provider did not address substance use/positive CRAFFT screen or mental health concerns in 3 patients (27%), Figure 2.

No referrals to our pediatric addiction services were made in either the urban or rural location during the pilot study.

Discussion

Findings from our study demonstrated a higher positive substance use screening rate in our rural practice (14.6%) compared to our urban practice (2.4%). In 27% of the positively screened patient, the primary care provider did not acknowledge the positive screen, nor was substance use or mental health issues discussed with the patient. Only 27% of the positively screened patients were followed by documented action or discussions about substance use by the provider, whereas 45% of positively screened patients had mental health concerns addressed without provider documentation of substance use or the positive CRAFFT. These findings raise the concern that despite being alerted of a positive CRAFFT screen during the rooming process and having received widespread communication of referral availability to our pediatric addiction service, substance use was only addressed with a minority of those screened positive and there were no referrals for further assessment. These findings suggest that increased awareness of substance use screening and referral services may not be adequate.

Our group attempted to reduce referral barriers and improve perceived institutional support by ensuring our pilot study practices have adequate access to our pediatric addictions service, links to community resources, and clarification from legal counsel on adolescent patients’ ability to consent to treatment and confidentiality from parents and legal guardians. Despite our efforts, only a small percentage of positively screened patients had their substance use addressed, suggesting that other barriers need to be further examined and addressed. Future studies to evaluate our efforts in changing provider’s perceptions could also be explored.

Our results also demonstrated a significantly higher proportion of positively screened adolescent patients in the rural practice. It is unclear if this is unique to this practice alone, or if this could be generalized to other rural regions. However, our findings are consistent with previous reports suggesting higher rates of substance misuse among rural adolescents compared to their urban counterparts, including higher levels of binge drinking, prescription opioid misuse, and nicotine-containing products.^2,18,19

Interestingly, more CRAFFT questionnaires were administered and the majority of positive screens were identified during acute/sick-child visits. Our approach led to increased screening and captured more patients with potential substance misuse issues.

There are several limitations to our study. We conducted our study by looking at 2 regions of Minnesota (one rural and one urban) during a 3-month period. Additionally, since we only conducted our study within our primary care practice, adolescents living in our region who do not routinely receive medical care or who receive care outside of our healthcare system are not captured. All this could limit the generalizability of our findings. Work is underway to expand our efforts to more practices within our institution.

Moreover, the CRAFFT response rate in our rural site was only 52.6% (vs 90% in our urban site). We did not follow up with non-respondents about reasons for not completing the CRAFFT, thus the positive screening rate could have been different if a higher percentage of patients completed the screening at our rural site. Response bias could also have contributed to our findings. It is possible the lower response rate could be related to site-specific processes, as our rural site has a lower support staff-to-provider ratio. Therefore, utilizing the same screening workflow for both urban and rural practices may not be realistic. Work is ongoing to gain provider and support staff feedback on our workflow.

Finally, we acknowledge the role of social desirability bias in our study. Patients may not have responded truthfully when accompanied by parents during their appointments or if they were concerned that their parents could view their responses through the patient portal. Therefore, it is possible that our results may be an underestimate of the true extent of substance misuse in the adolescent population. Future studies utilizing indirect questioning techniques, such as randomized response approach, ²⁰ may be helpful to reduce the effect of social desirability bias.

Conclusion

Our pilot study suggests that embedding a universal substance use screening and improving referral access to pediatric SUD support alone may not be adequate to drive provider-initiated discussions and action to address substance misuse in the adolescent population. Consistent with previous reports, our findings suggest a higher rate of substance misuse among rural adolescents. Future efforts to identify barriers to addressing substance use-related issues with primary care providers caring for adolescent patients could improve screening rates and facilitate early intervention.