Sensitivity and Specificity of a Signaling Question for Surveillance of Cognitive Functioning in Pediatric Sickle Cell Disease

Background

Sickle cell disease (SCD) confers neurological risks that contribute to cognitive and academic difficulties. Clinical guidelines state that cognition should be monitored using signaling questions. However, evidence is lacking regarding the extent to which signaling questions accurately identify children with cognitive issues.

Procedure

Caregivers of youth with SCD between ages 7 and 16 years (n = 89) responded to a signaling question about their child's academic performance: “How would you rate your child's academic performance relative to their peers?” Response choices included below grade level, on grade level, or above grade level. Youth with SCD completed the Wechsler Intelligence Scale of Intelligence, Fifth Edition to assess cognitive functioning.

Results

Most children were described by caregivers as being on grade level (67%), followed by below grade level (21%), and above grade level (11%). Cognitive testing revealed that 36% of participants had a low IQ (≤85; i.e., ≥1 SD below the mean), 60% had an IQ within normal limits (86–114; i.e., <1 SD below or above the mean), and 4% had a high IQ (≥115; i.e., ≥1 SD above the mean). Academic performance grouping was significantly associated with differences across cognitive domains (all p < 0.001). Cognitive functioning varied by academic grouping for younger (ages 7–11; all p ≤ 0.002) but not older children (ages 12–16; all p ≥ 0.05). The signaling question's sensitivity for identifying those with cognitive difficulties was low (42%–54%); specificity, negative predictive value, and positive predictive value estimates were acceptable (87%–93%, 74%–84%, and 53%–79%, respectively).

Conclusions

Asking caregivers of children with SCD to report on academic performance is a simple method of monitoring cognition. Children rated as below grade level should be referred for cognitive screening or evaluation. However, this question will likely result in a high rate of false negatives, and additional questions may be necessary to determine when to refer patients without obvious academic concerns.