Test-Retest Reliability of a Computerized Neurocognitive Battery in School-Age Children with HIV in Botswana

Area of Science:

• Neuroscience
• Pediatric Infectious Diseases
• Global Health

Background:

• Human immunodeficiency virus (HIV) infection significantly impacts children and adolescents in Botswana.
• Standardized neurocognitive testing is limited in this high-burden, low-resource setting.
• The culturally adapted Penn Computerized Neurocognitive Battery (PennCNB) aims to simplify neurocognitive assessment for youth.

Purpose of the Study:

• To evaluate the test-retest reliability of the culturally adapted PennCNB.
• To determine the reliability of summary scores, domain scores, and individual tests within the PennCNB.
• To identify potential practice effects influencing test-retest reliability.

Main Methods:

• The study involved 65 school-age children (ages 7-17) living with HIV in Botswana.
• Test-retest reliability was assessed using the culturally adapted PennCNB.
• Intraclass correlation coefficients (ICCs) were calculated for summary, domain, and individual test scores.

Main Results:

• PennCNB summary scores demonstrated high reliability (ICCs > 0.80).
• Domain scores showed good to excellent reliability (ICCs = 0.66-0.88).
• Individual tests exhibited more variable reliability (ICCs = 0.50-0.82), with memory tests being least reliable. Practice effects were noted in memory and complex cognition.

Conclusions:

• The adapted PennCNB offers adequate test-retest reliability at the domain level for pediatric HIV populations in resource-limited settings.
• Reliability varies significantly for individual PennCNB tests, requiring careful consideration during interpretation.
• Findings highlight the need to account for test-specific reliability and practice effects when implementing the PennCNB in similar populations.