In a recently released article in Pediatrics, Dr. Elizabeth D’Amico et al. (10.1542/peds.2016-1717) examine the usefulness of old and new adolescent screening tools for alcohol and marijuana use for use in primary care (Ref here). Primary care physicians have multiple mandates and many topics to cover in well teen visits, and this study gives welcome information comparing screens that can identify alcohol and marijuana use and impairment.
Likely none of us need to be convinced that this is a critical part of the well teen visit. Rather, our challenges are (1) how to incorporate screening in a way that is time efficient and socially comfortable for both patient and physician, and (2) how to select a screening tool with good psychometric properties. Additionally, our families are frankly already at risk of receiving too many screening questionnaires in our efforts to meet many needs and metrics: for example a teen-parent dyad at our Practice could easily already be completing an Asthma Control Test (if any history of asthma), an overall Adolescent Questionnaire, a PHQ-9 (Patient Health Questionnaire – a depression screen), and a Health Leads Screen (for any non-medical unmet social needs), and potentially be offered a research questionnaire if there is an ongoing study offering participation. So what can we glean from Dr. D’Amico and colleagues’ work that helps us?
The authors chose to compare a new screening tool, the 2 question NIAAA SG (National Institute on Alcohol Abuse and Alcoholism Screening Guide) with 3 older well studied screening tools, the AUDIT (Alcohol Use Disorder Identification Test – 10 items), the CRAFFT (Car-Relax-Alone-Forget-Family and Friends-Trouble – 6 items) and the PESQ-PS (Personal Experience Screening Questionnaire Problem Severity Scale – 18 items). Each takes about 5-10 minutes to complete, although the brevity of the 2 question screen (NIAAA SG) is tough to beat for time efficiency. The 1573 teens ages 12-18 years that participated in the study were a geographically, racially and ethnically diverse population recruited from underserved populations cared for at community health clinics in Los Angeles and Pittsburgh.
The geographic diversity of the two cohorts makes it easy to generalize results to similar populations. Teens with alcohol or marijuana use disorder were identified using the DSM-5 criteria following completion of the computerized version of the Diagnostic Interview Schedule for Children Version IV (DISC-IV). The authors readily acknowledge limitations of their study. The most relevant limitation for practitioners, I believe, is that teens did not report study responses to their physicians, so it is possible that confidentiality and disclosure concerns would alter results in actual clinical use. The authors present results in several clear tables, and at-risk youth were able to be identified using each of the 4 screening tools.
Rather than deluge you with the sensitivity, specificity, PPV (positive predictive value) and NPV (negative predictive value) of each test, which are easily seen in Tables 3 and 4 for identifying teens with “Alcohol Use Disorder” and “Cannabis Use Disorder”, respectively, I’d like to vote for my screening tool of choice. I am going to use the new NIAAA SG. It is only 2 questions in length, it has decent psychometric properties, and it incorporates elements that I am already using, specifically by asking “Do any of your friends drink alcohol?” which for me has been a reliable and nonthreatening way to open the door to this conversation. For those of you interested, the other question is more specific, and asks, “In the past year, on how many days have you had more than a few sips of beer, wine, or any drink containing alcohol?” and the questions are asked in different order depending on age, with responses leading to one of 4 risk categories.
Read, vote and share your thoughts with us as to what tool you prefer by responding to this blog, sharing a comment on our journal website, or posting your thoughts on our Facebook or Twitter sites.
