Factors associated with decisions to refer possible abusive head trauma to a hospital-based child protection team in Aotearoa New Zealand

Background: Aotearoa New Zealand has a high incidence of abusive head trauma (AHT) and a national children's hospital with an established multi-disciplinary child protection team (CPT). Staff concerned about possible AHT are expected to refer to the CPT, but there has been no research into the factors which might influence those referral decisions.<br /> Objective: To identify factors associated with decisions to refer head-injured children to the CPT, and to assess whether those factors are consistent with the literature. Participants and setting: All children <5 years old admitted from 2010 to 2019 with skull fracture or intracranial injury.<br /> Methods: Retrospective review comparing 25 variables in cases referred and not referred. Multivariable analysis estimated the independent role of each variable and modelled their combined contribution to decisions to refer. The area under the receiver operator characteristic curve (AUROC) and 95 % CI were used to describe performance of the model.<br /> Results: Of 631 head-injured children, 265 (42 %) were referred and 121 (19 %) diagnosed as AHT/undetermined. Variables associated with referral decisions were age < 1 year, p = .0001; injury inconsistent with the history, p < .0001; certain categories of history (motor vehicle accident, no history, history of abuse, fall <1 m, blunt force and penetrating trauma), p < .0001; delayed presentation, p < .0001; past history of injury, p = .0001; social/behavioral concerns, p < .0001 and subdural hemorrhage, p = .01. The AUROC was 0.95 (95 % CI 0.93, 0.97).<br /> Conclusions: Factors associated with referral are generally consistent with the literature. The percentage referred seems justified given the number diagnosed as AHT/undetermined.