Although screening for behavioral health problems in ambulatory medical settings had been strongly advocated, little progress had been made toward realizing this goal. Thus, we conducted a series of studies to assess whether providers and patients wanted screening and then developed a screening tool to be used in Primary Care. First, focus groups were conducted at four hospitals across Pennsylvania to discuss behavioral health issues including screening. Most providers agreed that computerized screening for behavioral health would be a great asset. Second, a feasibility study was conducted with 24 adolescents, who completed the Behavioral Health Screening-Primary Care (BHS-PC; Formerly AHSPC) before their medical appointment. After the appointment, adolescents reported that they a) liked the computer program, b) completed the tool in around 10 minutes, c) understood the questions, d) said they reported honestly, and e) most important, found it helpful during the appointment.
We developed a computer-based screening tool, Behavioral Health Screening-Primary Care, in an effort to standardize best-practice screening efforts for ambulatory medical settings. In this study we aimed to establish the instrument’s psychometric properties in three phases: 1) test-retest and internal consistency reliability, 2) concurrent validity 3) as well as predictive validity (predicting diagnosis). Adolescents aged 12-17 presenting for their well visits or non-acute return visits at primary care offices (e.g., follow-up to an acute visit) were recruited into the study. We recruited 415 adolescents across the sites who completed the BHS-PC, as well as a battery of established instruments measuring the same constructs as six of the BHS-PC subscales.
Four hundred adolescents completed the BHS-PC and a validation battery. Each BHS-PC scale was matched with a standardized, well-validated measure with norms and cutoffs. The sample was recruited out of primary care offices over 14 months. Analyses demonstrated that thebehavioral health screening (BHS) scales are unidimensional, internally consistent, and capable of accurately and efficiently discriminating among adolescents with a range of diagnostic symptoms (e.g. depression, anxiety, suicide and trauma) (Diamond et al., 2010). The tool is brief, yet psychometrically comparable to longer validated scales. The BHS resolves many practical and clinical barriers to behavioral health screening in primary care.
Analyses were used to establish cutoff scores on the BHS scales that maximized both true positive and true negative classifications of clinically significant behavioral health problems, as indicated by scores on the validation measures (Bevans, Diamond, & Levy, 2012). The operating curve characteristics of the BHS subscales are generally satisfactory. For each subscale, the probability of correctly distinguishing adolescents who have clinically significant behavioral health symptoms from those who do not exceeded 75%. The greatest risk of misclassification is falsely identifying adolescents with subclinical anxiety symptoms as anxious (specificity = 67%).
In another study, the BHS was modified to be used in the Children’s Hospital of Philadelphia Emergency Department (ED). The study began with qualitative interviews with 60 adolescents, their parents, and 45 medical providers. All three groups supported the idea of computerized screening in the ED, with some concerns about privacy, healthcare provider sensitivity, time constraints and lack of referral options (Pailler & Fein, 2009). Over a one year period, 857/3979 (22%) of eligible subjects completed the BHS-ED (Fein et al., 2010). There was a significant increase in the identification of mental illness or behavioral problems after initiation of the Behavioral Health Screening-Emergency Department (BHS-ED) (10.5% vs. 2.5%, OR = 4.58, 95% CI 3.53, 5.94) and more frequent ED-based behavioral health assessments by social workers or psychiatrists (8.3% vs. 1.7%, OR 5.12, 95% CI 3.80, 6.88).