Predictive value of the Ages and Stages Questionnaire® for school performance and school intervention in late preterm‐ and term‐born children

Abstract

Background: The new health supervision guidelines emphasize the importance of surveillance or a formal developmental screening test at 4 years, one of the most used tests is Ages & Stages Questionnaire. Nevertheless, there is still not enough evidence whether these tools will be useful to predict future school performance (SP). Objectives: The objectives of this study were to evaluate the Ages & Stages Questionnaire 3rd edition 48-month interval (ASQ3-48) translated to Spanish for predicting the need for school intervention (SI) and poor SP at 8-9 years of age, in late preterm infants (LPIs) and term-born infants (Terms) and to compare the prevalence of SI and poor SP in the two groups. Methods: Data were collected from a cohort of 75 LPIs and 58 Terms assessed with ASQ3-48 and with a further assessment of academic results at 8-9 years, through a standardized school test of the Education Department of Catalonia. SI data were obtained through a parent report. Area under the curve (AUC), sensitivity, specificity and predictive values were calculated, and logistic regression analysis was used. Results: The prevalence of poor SP was 12.8%, without statistically significant differences between LPIs and Terms. LPIs had higher SI than Terms (29.3% vs. 10.3%, P = 0.001). AUC for poor SP was 0.73 and for receiving SI was 0.56 without differences between the two groups. The sensitivity of the ASQ3-48 for poor SP was 41%, for specificity 92%, and for receiving SI 14% and 89%, respectively. Poor SP was related to having positive screening in the ASQ3-48 (OR 6.5 [95% CI, 1.9-22.2]) while having received SI was related to late prematurity (OR 3.6 [95% CI, 1.3-9.6]). Conclusions: The ASQ3-48 shows acceptable predictive properties for poor SP but not for receiving SI. No differences were found in SP between LPI and Term cohorts, but LPIs are likelier to require SI.

Predictive Validity of Developmental Screening Questionnaires for Identifying Children With Later Cognitive or Educational Difficulties: A Systematic Review

“…Of these, 32 publications met the inclusion criteria, corresponding to 10 cohort studies (Figure 1). The cohorts studied came from South America: Chile (19,27,36,37) and Colombia (26,38,39); Europe: Catalonia, Spain (40)(41)(42)(43), the Netherlands (28,(44)(45)(46)(47); France Poitiers and Nancy (48) and France Loire (49)(50)(51) and Norway (52)(53)(54); Asia: South Korea (55-57) and North India (58-62); Oceania: Australia (63).…”

“…Another factor that could alter developmental trajectories is the interventions carried out in children, data not reported by any of the cohorts. Only in the study from Catalonia was there evidence of a lack of association between the evaluation carried out at the age of 4 years and referral to support programs in development ( 42 ). It is described that in real world, referral rates for early intervention among children with positive screens ranged from 10–86% ( 67 , 68 ).…”

“…Eight of these cohorts describe comparison measures showing adequate capacity to predict later cognitive achievement/school performance. The ORs reported were >3, and the AUC was high (0.66–0.87), showing trade-offs in sensitivity and specificity, which could be explained by the different scoring methods and thresholds used ( 28 , 36 , 42 , 48 , 50 ); the optimal cutoff point, for a screening test, is the one that yields sensitivity and specificity values >70% ( 7 ).…”

The Indonesian version of Ages and Stages Questionnaire III accuracy compared to Bayley Scales of Infant Development III

“…However, if we look at the purpose of the developmental screening program itself, identifying and monitoring at‐risk children (including those who were incorrectly screened positive) gives more benefits that may lessen the future burden rather than losing the opportunity of missing the early intervention (resulting from those who were incorrectly screened negative). A recent study from Martínez‐Nadal et al found that ASQ‐3, specifically the 48‐month questionnaire, can predict school performance at 8–9 years old (Martínez‐Nadal et al, 2021). This opens the future possibility for follow‐up on the ASQ‐3 Indonesian version to determine the predictive validity of cognitive function and school performance at later ages.…”