Background Prematurity increases the rate of ASD. Therefore preterm infants should be screened for it. Screening tests are designed for 24 months onwards. However, some traits of ASD are present at an earlier age and early intervention seems to improve prognosis.

Aims (1) Study the prevalence of ASD at 12 and 24 months postmenstrual age. (2) Compare results at both ages. (3) Explore correlation with perinatal data.

Methods Infants born <33 weeks. At 12 months, parents completed a questionnaire (CSBS-DP), which measures three areas: Communication, Expressive Speech and Symbolic abilities. At 24 months they completed the CSBS-DP and another questionnaire (M-CHAT). Results were compared at both ages and were correlated with neonatal morbidity data.

Results Correctly questionnaires were obtained in 121 infants at 12 m and 43 infants at 24 m. Mean gestational age (GA) was 29.31 weeks (SD 2.39). At 12 m, 19% infants had abnormal total scores (Communication 15.7%, Speech 25%, Symbolic 20.7%). At 24 m, 16.7% had abnormal scores (Communication 23.8%, Speech 21.4%, Symbolic 14.3%), and 24.3% abnormal M-CHAT. Rate change from 12 to 24 m was significant for Communication (p = 0.006). PPV ranged 0.81–0.91 and NPV 0.31–0.60. Abnormal results correlated with umbilical artery pH <7.15 (p = 0.034), Apgar scores <7 at 5 min (p = 0.035), twin pregnancy (p = 0.049), abnormal brain scans (p = 0.005), surgery (p = 0.027) and a trend with male gender and lower GA.

Conclusions The CSBS-CP questionnaire correlates with neonatal morbidity. It shows a high PPV although a variable NPV when compared at 24 m. It seems a good early screening, although a close follow-up is always needed.