209 - Trends in substance use screening results before and after the onset of the COVID-19 pandemic among youth in pediatric primary care.

Poster Number: 209
Publication Number: 209.400

Maddie O'Connell, Boston Children's Hospital, Allston, MA, United States; Julia A. Plumb, Boston Children's Hospital, Jamaica Plain, MA, United States; Barbara J. Howard, The Johns Hopkins University School of Medicine, Baltimore, MD, United States; Raymond Sturner, Johns Hopkins University School of Medicine, Baltimore, MD, United States; Lydia A. Shrier, Boston Children's Hospital, Boston, MA, United States; Sion K. Kim Harris, Boston Children's Hospital/Harvard Medical School, Boston, MA, United States

Background: Youth may increase substance use (SU) when stressed. Trends in pediatric practice screening results may illuminate the immediate and ongoing impact of the COVID-19 pandemic on youth SU behaviors.

Objective: To assess trends in youth SU behaviors before and during the pandemic via screening results in a large national database of pediatric primary care practices.

Design/Methods: We analyzed the CRAFFT screen for SU and SU risks from visits by 12- to 22-year-olds at 226 U.S. pediatric practices utilizing the CHADIS online clinical process support system. CRAFFT assesses past-12-month use of alcohol, cannabis, and other drugs, driving/riding with a driver under the influence (Car), using to Relax, using Alone, Forgetting while drunk or high, Family/Friend concern about use, and getting in Trouble for use; ≥2 yes responses indicates problematic SU. We compared CRAFFT data in the same 6-month period (P) for 2018-2021 to control for seasonal effects (P1=Mar-Aug 2018, P2=Mar-Aug 2019, P3=Mar-Aug 2020, P4=Mar-Aug 2021). The start of P3 (Mar 2020) marked pandemic onset. We stratified analyses by age group (12-17 vs. 18-22) due to known SU differences in younger vs. older youth. We used logistic regression modeling; Generalized Estimating Equations accounted for data clustering within practices and patients.

Results: Patients (n=68,904) were 51% females, 53% from the South, and 84% age 12-17 (M+SD 15.5+2.3). In P4 (Mar-Aug 2021) vs. P1 (Mar-Aug 2018), both 12- to 17-year-olds and 18- to 22-year-olds reported lower rates of any SU and of alcohol use; these rates did not differ across P1-P3. Rates of problematic SU did not change over P1-P4 in either group. SU-related car risk was lower in P2 (Mar-Aug 2019; pre-pandemic) and P3 (Mar-Aug 2020; immediately post-pandemic onset) for the 12-to-17 group, vs. P1, but no different in P4. Youth 12-17 years old were more likely to report using alone immediately post-pandemic onset (P3) vs. P1. For 18- to 22-year-olds, using alone had begun to increase pre-pandemic (P2), a trend that continued during the pandemic (P3 and P4).Conclusion(s): Youth in pediatric primary care reported less SU during vs. before the COVID-19 pandemic, likely related to less peer contact and more family time. However, problematic SU did not decrease. Decreased SU-related car risk for younger youth before and immediately post-pandemic onset was not sustained. Increased SU when alone may reflect isolation from peers. Routine SU screening in pediatric primary care remains critical for addressing the pandemic’s effects on SU risk in younger and older youth.
Table 1. Trends in youth substance use screening results in pediatric primary care, by age group and time periods before and during the COVID-19 pandemic.CHADIS_CRAFFT_Data_Table FINAL.jpega Unique patients within time period.
b AOR=Adjusted Odds Ratio. CI=Confidence Interval. Adjusted for patient age in years, respondent (self-administered versus clinician interview), and region (Northeast, Midwest, South, West). Generalized Estimating Equations account for data clustering within practices and patients.
c Substance Use (SU) includes alcohol, cannabis, and other substance use (“anything else to get high”).
*Significant at p < 0.05
**Significant at p < 0.01
***Significant at p < 0.001