Classification accuracy of the Quick Interactive Language Screener for preschool children with and without developmental language disorder.

INTRODUCTION: This research examined the classification accuracy of the Quick Interactive Language Screener (QUILS) for identifying preschool-aged children (3;0 to 6;9) with developmental language disorder (DLD). We present data from two independent samples that varied in prevalence and diagnostic reference standard. METHODS: Study 1 included a clinical sample of children (54 with DLD; 13 without) who completed the QUILS and a standardized assessment of expressive grammar (Syntax subtest from the Diagnostic Evaluation of Language Variation-Norm Referenced; Structured Photographic Expressive Language Test-Preschool 2nd Edition; or Structured Photographic Expressive Language Test-3 rd Edition). Study 2 included a community sample of children (25 with DLD; 101 without) who completed the QUILS and the Auditory Comprehension subtest of the Preschool Language Scales-5th Edition (PLS-5; Zimmerman et al., 2011). Discriminant analyses were conducted to compare classification accuracy (i.e., sensitivity and specificity) using the normreferenced cut score (< 25th percentile) with empirically derived cut scores. RESULTS: In Study 1, the QUILS led to low fail rates (i.e., high specificity) in children without impairment and statistically significant group differences as a function of children's clinical status; however, only 65% of children with DLD were accurately identified using the norm-referenced cutoff. In Study 2, 76% of children with DLD were accurately identified at the 25th percentile cutoff and accuracy improved to 84% when an empirically derived cutoff (<32nd percentile) was applied. CONCLUSIONS: Findings support the clinical application of the QUILS as a component of the screening process for identifying the presence or absence of DLD in community samples of preschool-aged children.

Within and Across Language Predictors of Word Learning Processes in Dual Language Learners.

This study investigated the relation between Dual Language Learners' (N = 90) vocabulary and grammar comprehension and word learning processes in preschool (aged 3-through-5 years). Of interest was whether: (a) performance in Spanish correlated with performance in English within each domain; and (b) comprehension predicted novel word learning within and across languages. Dual-language experience was evaluated as a potential moderator. Hierarchical linear modeling revealed stronger predictive associations within each language than across languages. Across languages, results varied by experience and domain. Structural sensitivity theory suggests exposure to two languages heightens awareness of parameters along which languages vary and provides a framework for interpreting complex associations within and across languages. Knowledge from one language may influence learning in both.

Impact of highly active antiretroviral therapy on the spectrum of liver disease in HCV-HIV coinfection.

BACKGROUND & AIMS: Hepatitis C virus (HCV) coinfection is common in patients with human immunodeficiency virus (HIV) infection. The mortality associated with HIV has decreased dramatically with the introduction of highly active antiretroviral therapy (HAART). However, the impact of HAART, including protease inhibitors (PIs), nucleoside reverse-transcriptase inhibitors (NRTIs), and non-NRTIs (NNRTIs), on the spectrum of HCV-related liver disease remains unclear. The purpose of this retrospective analysis is to determine the impact of PI and NNRTI use on liver histological characteristics in patients with stable HIV-HCV coinfection (n = 101) compared with HIV-uninfected controls with HCV infection (n = 302). METHODS: The majority of coinfected patients were men (75%), African American (82%), and had genotype 1 HCV (91%). Mean HIV load was 1.52 log copies/mL, 48% had undetectable HIV RNA and a mean CD4 count of 528 cells/microL, and 11% had a CD4 count < 200 cells/microL. Both mean alanine aminotransferase (ALT) level (83 U/L; 54% had a normal ALT level) and Knodell Histological Activity Index score (7.04; 33% had advanced fibrosis) were similar to those of our control population. Ninety-three percent of patients were administered a mean of 3 antiretroviral medications: NRTIs in 98%, NNRTIs in 45%, and PIs in 54%. RESULTS: There were no significant differences in biochemical or histological parameters between patients administered or not administered PIs or NNRTIs. CONCLUSIONS: In this uncontrolled retrospective analysis, we were unable to show a significant impact of either PI or NNRTI use on the spectrum of liver disease.

The clinical spectrum of hepatitis C virus in HIV coinfection.

The biochemical, virologic, and histologic spectrum of hepatitis C virus (HCV) in 66 consecutive patients with HIV-HCV coinfection and 119 HCV controls was compared: 86% of coinfected patients had CD4 counts >200 cells/mm3, 51% had a normal alanine aminotransferase (ALT) value, the mean HCV RNA titer was 5.7 log IU/mL, 92% of coinfected patients were of genotype 1, and the mean histologic activity index was 6.86 with advanced fibrosis in 32% of patients. The biochemical, virologic, and histologic findings of HCV in coinfected patients were similar to those observed in HCV controls. For both groups of patients, no clinical, biochemical, or virologic factors could reliably identify patients with advanced fibrosis or cirrhosis, underscoring the importance of liver biopsy in the evaluation of these patients. The spectrum of liver disease in coinfection includes a significant proportion of patients with normal ALT values, and excluding these patients from previous studies has led to an overestimation of HCV disease severity.