Comparing diagnostic outcomes of children with fetal alcohol syndrome in South Africa with diagnostic outcomes when using the updated Institute of Medicine diagnostic guidelines.

INTRODUCTION: During fetal alcohol spectrum disorder (FASD) prevalence studies in South Africa, cases of fetal alcohol syndrome (FAS) were identified that presented differently from the 2016 Hoyme et al. modified Institute of Medicine (IOM) criteria. We compared diagnostic outcomes of children diagnosed with FAS using a combination of the 2005 Hoyme et al. criteria and the "gestalt method" in South Africa to the diagnosis they would have received using the latest Hoyme et al. criteria. The frequency with which dysmorphic features presented was compared to the frequency with which they were reported in the revised criteria which drew on a larger sample.

METHODS: Data were gathered from four South African FASD prevalence studies. Dysmorphology data, anthropometric data, and final diagnosis for participants (N = 917) were extracted.

RESULTS: Of the 390 participants with diagnoses of "full FAS," 175 would not have received a "full FAS" diagnosis using the 2016 criteria. Of these, 21 would have received a pFAS diagnosis, and 154 would have received a diagnosis of ARND or a "no-FASD" diagnosis. The frequency of all but five dysmorphic features differ significantly between this sample and the sample examined for the 2016 criteria. There is more variability in the features present in the current sample.

DISCUSSION: Differences regarding diagnostic outcomes and prevalence of dysmorphic features suggest that strict application of the diagnostic criteria may miss children who present with FAS. We recommend including gestalt-based screening in a research setting where the clinical experience is available to inform future guidelines.