This is the first systematic evaluation of health professionals’ perceptions of the need for and design of screening for FASD in Australia. We found consensus support for targeted screening, and more frequent agreement with screening in childhood than at birth. Although about half the participants supported the need for universal screening for FASD as an ideal and ethically responsible approach to minimise missed cases, consensus agreement was only achieved for targeted screening of individuals with relevant clinical presentations and groups at high risk of FASD. Parent or caregiver concern that their child may have a FASD and clinical presentations associated with the likelihood of prenatal alcohol exposure were most strongly endorsed as indications for targeted screening.
Targeted screening was considered more cost effective and feasible than universal screening, with support for targeted screening linked to constraints on service provision, including the need for provider training and increased demand for diagnostic and intervention services. Participants were also concerned about the limitations of existing screening methods and the need for programs to have acceptable sensitivity, specificity and cost-effectiveness. These findings reflect previously identified difficulties in screening for FASD
[20, 29, 30], and deficits in diagnostic and management capacity to support screening
[9, 14, 15, 29, 44, 45].
We found consensus agreement on the need for a checklist and criteria for referral to support the implementation of screening for FASD. We also found consensus agreement that screening primarily requires health professionals to assess prenatal alcohol exposure and consider it as a possible cause of abnormalities. These findings are consistent with the need for greater awareness of FASD and improved capacity for screening among Australian health professionals
[46, 47], and highlight support for targeted screening as a strategy which can be integrated into usual clinical practice and passive case finding. Over half of participants agreed that most of the information required for FASD screening is assessed at birth or during relevant clinical presentations in childhood, which indicates some capacity for FASD screening in Australia. However, perceptions of participants in this study could differ systematically from health professionals who do not have experience or expertise in FASD screening or diagnosis.
Participants indicated that standard, explicit and efficient screening criteria are needed to determine when referral for a diagnostic evaluation is required. Few formal, validated instruments for FASD screening exist, and no single instrument is suitable for all ages and settings
. The absence of a single common distinguishing feature of FASD that can be used to indicate the need for a diagnostic evaluation has required programs to use aetiological risk factors and diagnostic features to determine whether a diagnostic evaluation is required. Although these screening methods have been considered unreliable, time consuming and rudimentary
, most participants supported the use of screening indicators linked to aetiological and diagnostic factors, and emphasised the need to identify efficient and effective screening criteria.
Inaccurate FASD screening and case ascertainment have been linked to deficits in practitioner training and the use of inconsistent case definitions
[8, 26]. Formal assessment of facial dysmorphology and functional CNS performance at the screening stage was not considered necessary, reliable or appropriate in all settings. The need for valid population reference data for comparison was also identified. Although more than half of participants were not familiar with the formal assessment instruments for alcohol exposure presented, we found consensus agreement on the need for formal assessment methods, and on the usefulness of the AUDIT-C assessment instrument
 following description of this instrument in round 2.
The facial photographic screening tool was considered to provide a feasible alternative to the measurement of facial anomalies by inexperienced assessors, although some participants indicated concerns about the use of facial dysmorphology assessment in FASD screening. While facial anomalies have been used to screen for FAS
[26, 27, 29] facial anomalies are not commonly assessed in FASD screening
[9, 32]. The assessment of facial anomalies has been considered inappropriate in general population screening for FASD due to multiple factors, including its low predictive value for FASD
. However, diagnostic guidelines and criteria for referral recognise the significance of the characteristic FAS facial anomalies as an indicator of the need for diagnostic evaluation, particularly in the absence of confirmed prenatal alcohol exposure
[6, 9, 12].
Criteria have been established to identify when population screening is appropriate and ethically justified, and to ensure that screening does more good than harm
. The variable application of these criteria in practice has been attributed to their inherent subjectivity and a lack of evidence to comprehensively evaluate screening interventions
. Criteria for the evaluation of population-based screening interventions include whether the potential impact of the condition is sufficient to justify screening; whether there is a benefit from early diagnosis and treatment; whether there is a cost-effective and acceptable screening test; and whether appropriate diagnostic services and treatment are available
[49–51]. Our finding of a lack of consensus on the need for universal screening of apparently healthy individuals is consistent with the failure of FASD screening to meet some of these requirements.
Targeted screening for high risk presentations and groups was supported as a strategy to improve the identification of FASD in Australia, consistent with developments in health policy
. Participant comments emphasised the importance of ensuring the effectiveness of screening interventions prior to their use to ensure that the intervention benefits both individuals screened and the community. There is little reliable information on the epidemiology of FASD in Australia, and it is not clear what proportion of individuals with FASD would be identified by targeted screening of high-risk groups. Our findings highlight the complex issues that must be addressed when pursuing the deceptively simple objective of early diagnosis
, and the real need for systematic evaluation of the risks and benefits of proposed screening interventions
The Delphi method provides a strong basis for the construct validity of the study findings, with participants able to validate their initial responses and identify areas of uncertainty
. This study was primarily based on an exploration of established screening methods, and our use of non-representative sampling and a modified Delphi process may have limited our ability to evaluate all potentially relevant information on the design of FASD screening programs in Australia. Nevertheless, participants provided diverse perspectives on the use of screening for FASD, including the identification of risks associated with screening, and the importance of both individual case finding and a population based approach. Further research is needed to evaluate potential strategies to facilitate improved identification of FASD, and to evaluate their performance and acceptability in the Australian context.
We aimed to recruit health professionals who had specific experience or expertise in FASD screening or diagnosis to identify consensus perceptions on the approach to FASD screening in Australia. Paediatricians were most highly represented within the study sample, reflecting the profession’s key role in diagnosis in Australia. The round 1 response in this study exceeded the 70% recommended level
 and was similar to that observed in other Delphi studies involving clinician panels
[53, 54]. Attrition was greatest between the time of agreement to participate in the study and returning the round 1 questionnaire, and similar to that reported by others
. Delay between recruitment of the panel and distribution of the questionnaire, as well as the recruitment of a large panel
 may have reduced response in this study. Although we found response completeness was associated with participant occupation and experience, there was little evidence of differences in perceptions about screening according to response completeness.
Evaluation of pilot screening interventions are required to address the lack of evidence to support the effectiveness of screening for FASD in Australia and identify the impact of screening parameters on program outcomes and effectiveness. This approach to development would enable investigation of practical barriers to success, including the ability to engage with high risk groups. The development of effective screening interventions can improve our understanding of the epidemiology of FASD and its prevalence in high risk groups; contribute to the identification of appropriate strategies for FASD management and prevention; and provide an empirical evidence base for FASD policy in Australia.
It is clear from our findings that additional strategies are also required to improve passive case-finding, referral and diagnosis in Australia. Signs of FASD are frequently non-specific, and barriers to seeking or accessing appropriate diagnostic and intervention services are likely to be important obstacles to passive case finding. Strategies that can be used to improve passive case finding capacity and the identification of FASD in Australia include the development of resources and programs to improve awareness of the disorders and their prevention among health professionals, other relevant professionals and the wider community; the development of standard criteria to guide health professionals on appropriate referral of individuals who require specialist assessment, as implemented in North America
[6, 12]; and improved diagnostic capacity and access to specialist services.