Accurately Assessing Substance Use Disorders in Children and Adolescents

Psychological Assessment | 2023, Vol. 35, No. 2, 178- 187

Diagnostic accuracy of the Child and Adolescent Symptom Inventory (CASI-4R) substance use subscale in detecting substance use disorders in youth

Angelina Pei-Tzu Tsai, Eric Youngstrom, Kenneth Gadow, Sarah Horwitz, Mary Fristad, Stacey Daughters, Andrea Young, L. Eugene Arnold, Boris Birmaher, Stephanie Salcedo, The LAMS Group, & Robert Findling

Identifying substance use disorders (SUDs) early and accurately improves case formulation and treatment. Previous studies have investigated validity and reliability of the Child and Adolescent Symptom Inventory (CASI) for anxiety, mood, and behavior problems. The current study’s aim was to test if the embedded CASI Substance Use (SU) subscale can discriminate adolescents and young adults (AYA) with and without a SUD diagnosis accurately enough to justify clinical application within an evidence-based assessment framework. N=479 outpatient AYA (age 14-21) and their caregivers completed K-SADS-PLW semi-structured diagnostic interviews; caregivers completed the CASI and adolescents completed a parallel version, the Youth (Self-Report) Inventory (YI). K-SADS-PLW indicated that 33 youth met DSMIV criteria for SUDs. Receiver Operating Characteristic (ROC) analyses found that both CASI and YI Substance Use subscale scores significantly identified KSADS-diagnosed SUDs in AYA: Caregiver area under curve (AUC)= .91, p <.0005; YI(AUC)= .90, p <.0005. There was no significant difference in diagnostic accuracy between informants. Both subscales showed diagnostic and clinical utility in identifying AYA SUDs in outpatient mental health settings. Findings suggest that the CASI-4R subscale could be a helpful screening instrument for AYA SUDs. A case vignette illustrates the clinical application of study findings. Future research should examine rapport as a moderator of reporting accuracy, and replicate use of these measures under varying clinical scenarios.

Substance use disorder, adolescent, young adult, sensitivity, specificity, screening

“Substance use disorders (SUDs) are prevalent and among the largest contributors to disability and impairment in adolescents and young adults…Approximately 60% to 90% of adolescents with SUDs present diagnosable symptoms satisfying criteria for one or more co-occurring psychiatric disorders. Because they are prevalent, highly comorbid, and complicate treatment of other mental health issues, early identification would be valuable ....In outpatient and community settings, rating scales and checklists are often used to evaluate clients’ symptoms and guide the next course of actions, such as administering a disorder-specific measure…

However, most substance use (SU) instruments such as the Brief Screener for Tobacco, Alcohol, and Other Drugs, Problem Oriented Screening Instrument for Teenagers, CRAFFT, and Screening to Brief Intervention are stand-alone measures that focus on SUDs only; they do not assess common comorbid disorders. They also are underused in mental health settings, despite many of them being free. Finding and implementing these creates barriers to implementation…The Child and Adolescent Symptoms Inventory is a promising option, because the CASI includes a SU subscale while also assessing a broad band of psychopathologies” (p. 2- 3).

“The current study extends previous work by investigating the CASI SU subscale, a DSM-referenced screening instrument designed to assess symptoms of a set of common psychiatric disorders, including SU on AYA, in an older age cohort (e.g., late adolescence and early adulthood versus school-age through high school). CASI items correspond to DSM symptom criteria for target conditions, and screen for more diagnoses with more complete item coverage than the corresponding ASEBA “DSM-oriented” scales. …Secondary analyses used the 96-month final wave data (N=484; Mage=17.3±1.9) from the Longitudinal Assessment of Manic Symptoms (LAMS) study. LAMS was designed to examine the trajectory of emotional and behavioral dysregulation over time in youth seeking initial outpatient mental health services” (p. 4).

“The CASI-4R is a 163-item rating scale evaluating symptoms of emotional and behavioral disorders using DSM-IV diagnostic criteria for youth aged 5 to 18. The SU subscale in the updated DSM-5-referenced version of the CASI and the CASI-4R are identical… Unlike the ASEBA, the CASI and YI also include an impairment item, ‘How often do the behaviors in Category O interfere with youth’s ability to do schoolwork or get along with others?’... The current study used scores at the 96-month wave as participants would have aged into the period of higher risk for substance misuse” (p. 5).

“Both caregiver- and self-report significantly predicted SUD diagnoses, as hypothesized, even after adjusting for demographics and number of comorbid diagnoses. Both CASI and YI SU subscales also provided clinically useful individual-level information about risk. The two scores did not perform significantly differently compared head-to-head, but youth report showed significant incremental validity, whereas caregiver report did not. Our study recruited youth from outpatient clinics and included a comparison group comprising a wide spectrum of disorders, resembling a typical clinical outpatient setting… Both of these embedded scales showed clinically meaningful results as a way of potentially identifying cases in mental health settings deserving of more thorough consideration of “dual diagnosis” or comorbid substance use issues” (p. 8).

“... Clinicians can directly follow-up elevated scores with careful interviewing, and then integrate findings to gauge severity and develop an appropriate case formulation and treatment plan. With multiple-thresholds, clinicians can use a more objective predictive system, reducing potential heuristics and sources of bias in clinical judgement” (p. 8- 9).

“Overall, results suggest that both the youth (YI) and caregiver (CASI) reports provide valid information about possible SUD in teens and young adults. Accuracy is high enough to be helpful in clinical outpatient settings, where the SU subscale is embedded as part of an assessment instrument with multiple other syndrome scales. The CASI and YI cover symptoms of frequently comorbid conditions as well as alerting to the presence of SU patterns that merit more clinical evaluation” (p. 10).

“... Replication in additional samples would be helpful. Data used were the last wave of a longitudinal study in which participants had been working with clinical research staff annually. In contrast, intake is conducted at the beginning of treatment, when rapport is just established. Additionally, data were collected at universities-affiliated clinics rather than specialized substance use treatment centers, which explains a lower base rate in the current sample compared to the similar age group in the population-based studies. Further research should examine rapport as a moderator in evaluating the underreporting rate of SUDs, and replicate use of the CASI and YI under varying clinical scenarios” (p. 9).

As always, please feel free to check out the resources below and comment to add more to the discussion!

• Substance Use Disorder
  
  
• Mental Health and Substance Use Associated with Criminal Activity in Adolescent Inpatients