Abstract
Although the prevalence rate of panic disorder (PD) among adolescents is relatively low, epidemiological research suggests that panic attacks and subclinical panic disorder symptoms occur in a substantial portion of the adolescent population. Retrospective and prospective studies also suggest that adolescence is a critical developmental period for the onset of PD symptoms. Given the negative academic, social, and emotional outcomes associated with undetected and untreated PD, effective prevention and intervention are warranted. Identifying adolescents with current PD or who may be at-risk for future PD is an important step in such treatment efforts. Among professionals working with youth, physicians and medical staff may be at a particularly advantageous position to screen for adolescent panic symptoms, given the high utilization of medical services among those experiencing such PD symptoms. Although limited time and resources within primary care settings frequently hinder effective mental health screening procedures, the use of time-and cost-effective screening instruments may aid professionals in detection efforts. With this in mind, the current study sought to validate a brief screening tool previously studied with adults for use with adolescents seen at pediatrics primary care practices. The screening instrument was evaluated both in terms of its ability to effectively detect adolescents with PD and in terms of the association between positive screen status and cognitive, symptom, and broader impairment variables associated with PD. Participants included 165 adolescents (57% male) ages 12 to 17 (M = 14.40; SD = 1.77) recruited from two general pediatrics clinics in Miami-Dade County, Florida. The sample was 42.3% White, Non-Hispanic, 41.1% Hispanic, 7.9% Black (African-American and Caribbean American), 1.2% Asian American, 7.4% mixed ethnicity or other, and 1.2% unknown. At Time 1, while in the waiting room of a pediatrics clinic, participants completed the Autonomic Nervous System Questionnaire (ANS; Stein et al., 1999), a five-item screening measure of panic symptoms. Of this larger sample, 45 participants (25 screening positive for potential panic disorder and 20 with negative screens, matched by age and gender to the positive screen group) completed telephone-administered follow-up measures at Time 2. Follow-up measures included a more comprehensive diagnostic assessment of PD and agoraphobia, as well as adolescent-report measures of anxiety sensitivity, interpretive biases, overall anxiety and depression, and functional impairment. At Time 1, 65 participants (39.4%) screened positive on the ANS, as indicated by endorsing the first and/or second item on the measure. Of those screening positive, roughly one-third of participants (33.84% of positive screens) endorsed moderate to severe anticipatory anxiety about future panic attacks. The ANS displayed excellent sensitivity (Se = 1.00), with two participants from the positive screen group meeting criteria for PD, and no control participants meeting criteria. However, as expected, specificity of the ANS was lower (Sp = .43), indicating a high degree of false positives (e.g., those screening positive but not meeting criteria for PD). In addition, as hypothesized, the ANS demonstrated good test-retest reliability (r = .74). Independent samples t-tests revealed that positive screen participants had significantly higher self-reported anxiety sensitivity, interpretive biases, anxious and depressive symptoms (including panic), and functional impairment than negative screen participants. This difference remained significant for overall symptom T-scores on the Revised Child Anxiety and Depression Scales (RCADS; Chorpita et al., 2000), even after controlling for group differences in anxiety sensitivity and interpretive biases. Finally, further analyses revealed that participants endorsing both starter items on the ANS (n = 7) had higher elevations on self-reported anxiety sensitivity and panic symptoms, compared to those not endorsing either item or those endorsing the first item (e.g., "In the past six months, did you ever have a spell or an attack when all of a sudden you felt frightened, anxious, or very uneasy?"), but not higher than those endorsing only the second item ("In the past six months, did you ever have a spell or attack when for no reason your heart suddenly began to race, you felt faint, or you couldn't catch your breath?"). These findings offer preliminary validation for the ANS as a screening measure for PD in adolescence, given its high sensitivity and ability to adequately "catch" patients with PD (e.g., low false negative rate). Perhaps even more importantly, those screening positive on the ANS demonstrated higher scores on cognitive correlates of PD and elevated internalizing symptoms and functional impairment, compared to participants screening negative. Based on these analyses, current recommendations for physicians and medical staff are to monitor and follow-up with adolescents screening positive on the ANS for the development of anxiety and panic disorder symptoms, particularly among those who endorse both starter items. However, given the relatively small sample size, replication of these findings in a larger sample is needed to further validate these recommendations. Finally, implications for prevention and intervention within pediatrics settings are discussed.