Background and aims Very preterm children have increased risk of behavioural problems. Easily administered valid screening tests are required for epidemiological studies. This study aims at exploring the relation between the Child Behaviour Checklist (CBCL) and the shorter Strengths and Difficulties Questionnaire (SDQ) in a subsample of the Italian area-based ACTION cohort followed up to school age.

Methods The ACTION follow-up study uses the SDQ parent version to assess the emotional and behavioural problems of children born <32 weeks gestation and survived to school age. For the purposes of the study, a subsample (n.223) of the cohort recruited in Lazio region was administered both SDQ and CBCL, parent-reported. We computed Spearman’s rho coefficients to carry out correlation analyses. The K-Cohen Test was used to measure agreement SDQ and CBCL instruments

Results 121 children were males (54.3%). 44 (19.7%) were <28 weeks gestation. Seven had cerebral palsy, 4 were blind or almost blind, and 9 used hearing aids.

According to CBCL, 20 children (9%) had significant and 18 (8.1%) borderline difficulties. SDQ identified as pathological 14 children (6.3%), while 13 were classified as borderline (5.8%). Overall agreement was 86.1% (Kappa 0.47, p < 0.0001).

Spearman’s rho was 0.63 for emotional/internalising problems, 0.61 for behavioural/externalising, and 0.74 for total problems. All values were statistically significant (p < 0.001).

Conclusions Correlation and inter-rater agreement between the two instruments are good. Overall, however, SDQ appeared more conservative in the identification of children with problems.

Acknowledgements The ACTION follow-up study was funded by Italian MoH and Chiesi Spa.