Screening Children and Adolescents (5–18 Years)
Screening for child and adolescent behavioral disorders using the Pediatric Symptom Checklist (PSC) is widely used in many medical practices and Medicaid programs. The current literature documents the ability of this brief, one-page instrument to identify children in need of further behavioral evaluation. Unfortunately, there are no randomized, controlled studies that document outcomes on screened individuals or groups, compared with populations not screened. PSC screening is classified “targeted” rather than “general” because the studies needed to provide a firmer evidence base have not been done.
A. Screening for Evidence of Behavioral Disorder
The PSC is a brief, one-page, 35-question instrument designed for use by parents in the doctor’s waiting room. The questionnaire is designed to detect behavioral and psychosocial problems in children from 2 to 16 years of age, and it has been used effectively in persons up to 18 years of age (Bernal et al., 2000). Each of the questions can be answered with a “never,” sometimes,” or “often,” with scores of 0, 1, or 2, respectively, attributed to each answer. Scores of 24, 28, or higher, depending on the age of the child, are considered indicative of a possible behavioral or psychosocial problem and will warrant further exploration by the clinician (Jellinek & Murphy, 1999).
The PSC has been suggested as a tool for universal use with children 2 to 16 years of age to screen for behavioral and psychosocial problems (Jellinek & Murphy, 1988; Walker, LaGrone, & Atkinson, 1989; Murphy, Arnett, Bishop, Jellinek, & Reede, 1992; Jellinek & Murphy, 1999, Gardner, 2002; Jellinek et al., 1999). In use since the 1970s, the PSC has been tested and used in tens of thousands of children; scored well in a test of its usefulness to the Medicaidsponsored Early and Periodic Screening, Diagnosis, and Treatment (EPSDT) program (Murphy et al., 1996); and is used in several States in the context of their EPSDT programming (Jellinek & Murphy, 1999; Bernal et al., 2000; Gardner, Kelleher, & Pajer, 2002).
The PSC has been found to be acceptable to parents, regardless of socioeconomic status or ethnicity, and to clinicians and clinic office staff (Murphy et al., 1992; Murphy, Reede, Jellinek, & Bishop 1992; Jellinek et al., 1999; Navon, Nelson, Pagano, & Murphy, 2001). It has been validated against more elaborate classification instruments—the Child Behavior Checklist (CBCL) and the Clinician’s Global Assessment Scale (CGAS (Walker et al., 1989; Jellinek & Murphy, 1999). It also routinely generates prevalence rates for pediatric psychosocial and behavioral disorders of approximately 12 percent, which is consistent with other estimates of pediatric behavioral and psychosocial disorders (Jellinek & Murphy, 1999; Jellinek, 1999). The expected increase in psychosocial dysfunction with lower socioeconomic class (Jellinek, Little, Murphy, & Pagano, 1995) and the expected correlation with maternal psychological distress and marital adjustment (Sanger, MacLean, & Van Slyke, 1992) have been clearly documented.
The primary outcome measure noted in the PSC literature has been the percentage of children referred for behavioral or psychosocial evaluation and treatment. This rate of referral has dramatically increased with the introduction of the PSC in every study where this measure has been reported (Navon et al., 2001). In one study, the referral rate increased from 1.5 percent before implementation of the PSC to 12 percent, then dropped back to 2 percent after the PSC screening was discontinued (Murphy et al., 1992). This review found no studies that address the behavioral and psychosocial benefits to the children screened or costs associated with referral of false-positive cases.
One study published in 2000 (Bernal et al., 2000) reported average log costs for health and psychiatric care for all children studied at $393 per year, and costs of those with anxious, depressed symptoms at $805 per year. Chronically ill children showed the highest health care costs, with average log costs of $1,138 per year. Psychosocial dysfunction was associated with higher costs. Unfortunately, this study did not explore whether detection and treatment of the psychosocial dysfunction could lower these costs. With a documented minimum sensitivity (accurately detecting true “positives” or those with the illness) of 80 percent, and a specificity (detecting those without disease) of 68 percent or better (Jellinek et al., 1988; Jellinek & Murphy, 1999), this screening instrument may miss up to 20 percent of children who have serious problems, and refer up to 32 percent of well children to diagnostic interviews that prove negative for any treatable behavioral or psychosocial behavior. Although these efficacy statistics are within acceptable ranges for screening instruments, they do speak to costs of program implementation that need to be considered. Like virtually all other screening programs, little or no benefit will accrue without follow-up treatment for those found to be in need of such treatment. Owing to the research findings, the PSC may be considered a “targeted” service for use in health care delivery settings with providers and health care systems wishing to use it.
The available literature leaves unanswered the possible use of the PSC when the primary care practitioner suspects a significant behavioral problem but does not have enough information to confirm or deny this impression. For such cases, health care systems may wish to make this instrument available to providers for selective use, at their discretion.
The PSC, along with articles describing its proper use on the Pediatric Development and Behavior Web site, is available at www.dbpeds.org/handouts/ (Jellinek & Murphy, 1999) under “screening.” It should be used without modification, other than for translation when working with non- English-speaking families.
The PSC consists of 35 very brief statements to which the parent responds “never,” “sometimes,” or “often.” Presented on a single page with check-off boxes, sample statements include: “Complains of aches/pains; tires easily, little energy; has trouble with a teacher; acts as if driven by a motor … .” The responses are graded on a zero-to-two scale. Depending on age, a score of 24, 28, or greater is considered indicative of significant psychosocial impairment (Jellinek & Murphy, 1999).
Summary: Children and Adolescents 5–18 Years
Screening for potential child and adolescent behavioral disorders using the PSC is widely used in medical practices and Medicaid programs. Because of its low burden (brief), ease of use, wide applicability, and validity, the literature supports its use by health plans with all children in a health care system. In this report, such screening is classified as a “targeted” service rather than “general” because no randomized controlled trials that could document outcomes have been attempted.
