Aims This study systematically reviews the literature on assessment tools for atypical feeding behaviour (AFB) in children and adolescents with autistic spectrum disorder (ASD).

Method Articles were identified by electronic database and manual searches and critically reviewed using established criteria. Outcomes were accuracy & validity of tests and whether or not the tests were found to be clinically useful or indicative of prognosis.

Results 19 relevant articles were identified describing 34 different tools. But only one questionnaire was specifically developed for AFB in children with ASD, the Brief Autism Mealtime Behaviour Inventory (BAMBI). A second tool, Screening Tool of Feeding Problems was found to be useful in children as well as adults. The BAMBI correlated well with the gold standard scale for feeding difficulties. For all factors, children in the autism group had higher scores than children in the typically developing children (χ² p<0.0001). Score on the PSI-SF (parental stress inventory-short form) if significantly lower at the end -compared with the beginning of treatment- can be used as an outcome measure for successful treatment of feeding problems (F (2, 72)=3.68/3.39, p<0.05). Scales using observable parent behaviour are promising but at an early stage in terms of outcomes. The Parent Mealtime Action Scale showed that if children with ASD have poor diet variety and low body mass index it is more likely their parents prepare special meals for them (multiple regression: β=0.318, t=4.29, p=0.14).

Conclusion The BAMBI and PSI emerged as screening instruments most likely to be of value in assessing AFB in children with ASD. No UK based original research was found in this study and it may be difficult to generalise from the evidence found to the UK. More research is needed in this area.