IntroductionNewborn screening (NBS) is a major public health programme enabling early detection and treatment of severe congenital conditions. The development of next-generation sequencing (NGS), declining genomic costs, and increasing therapeutic options for rare diseases have renewed interest in expanding NBS by genomic approaches (gNBS) [1,2,3,4,5].NBS policy has long been guided by the criteria set out by Wilson and Jungner for the World Health Organization in 1968, which require an important health problem with a well-understood natural history, an acceptable test, an available treatment, and benefits outweighing costs and harms. Andermann et al. revisited them in 2008 in the light of genomic technologies, adding requirements on equity of access, programme governance and quality assurance, informed choice and privacy.These developments have raised international debates on gNBS. In the late 2010s, initiatives such as the BabySeq Project in the United States helped structure these discussions [6, 7]. At that time, French NBS remained limited to severe childhood-onset conditions, mainly identified through measurement of analytes in a dried blood spot, whether metabolites, hormones or enzyme activity, with genetics used as a second-tier tool. Genetic testing of an asymptomatic person was then permitted only for cascade testing of relatives, once an affected index case had been identified. The recent revision of the French bioethics law in late 2021 authorised genetic testing within population-based screening, and therefore as a first-line test in NBS. Yet no structured national assessment of healthcare professionals’ (HCPs) views on gNBS was available.In this context, the Fédération Hospitalo-Universitaire TRANSLAD launched the SeDeN project (Séquençage et Dépistage Néonatal) to examine expectations and barriers related to NBS expansion, including genomic approaches. The project comprises several components: policy analysis (SeDeN-p1), HCPs (SeDeN-p2), parents (SeDeN-p3), and policymakers and stakeholders (SeDeN-p4) (Supplementary Fig. SF1).This article reports findings from SeDeN-p2. HCPs are central to any expansion. They inform families, obtain parental consent, take the sample and return the results. Whether a gNBS can be implemented therefore depends on their willingness and their capacity to deliver it, and insufficient preparedness is a recurrent barrier in the international literature [8,9,10,11,12,13,14,15,16,17]. The study assessed how HCPs involved in perinatal care and genetics in France perceive potential NBS expansion, with and without first-line genetic testing. It aimed to measure levels of acceptability across types of results, to identify areas of consensus and divergence between specialties, and to characterise the ethical, organisational and informational conditions that HCPs attach to NBS expansion.MethodsSeDeN-p2 combined an exploratory qualitative phase, aimed at designing a questionnaire aligned with field-related concerns, with a quantitative phase based on a nationwide self-administered survey of HCPs in France.The preliminary interviews were conducted in accordance with the COREQ guidelines. A purposive sample of fifteen participants, including obstetricians, paediatricians, midwives, biologists, clinical geneticists and public policymakers, was selected to ensure variation in professional background and practice setting. Semi-structured interviews were conducted online by CL between November 2020 and April 2021, audio-recorded, transcribed, anonymised, and analysed thematically using NVivo (version 12 Pro). This exploratory phase was conducted for questionnaire development purposes only. The sample was designed to cover the range of professional concerns, not to reach thematic saturation on a research question of its own, and the interview material is not analysed as such here.This phase identified four key domains: (i) Criteria for selecting conditions; (ii) Genetic findings to be returned (iii) Integration of NGS into NBS; and (iv) Parental information and consent.Accordingly, the questionnaire was constructed as a four-module questionnaire, each module corresponding to one of these domains. The criteria module presented two sets of ten items: the criteria for organised screening adapted from Wilson and Jungner and published in French by Guessous et al. [18], hereafter referred to as the Wilson and Jungner criteria, and the ten criteria of the Andermann synthesis [19], translated into French. The wording of each item and the identifier used in the figures and text are given in Supplementary Methods SM1. The four modules were administered as a single fixed sequence, and respondents could stop at any point. Because completion rates differed across modules, the number of responses included in the analyses varied by module (N = 1077, N = 874 and N = 751 for the criteria, genetic findings, and NGS and family information modules respectively). A module-based analytical strategy was therefore defined a priori: analyses were conducted separately within each thematic block, and for each module, respondents were retained if they had answered at least half of its mandatory items, prioritising interpretative coherence over maximising a single overall sample size. Detailed descriptions of the questionnaire content, response scales, and analytical procedures, including inclusion thresholds and comparability checks, are provided in the Supplementary Methods SM2 and Supplementary Fig. SF2. The questionnaire also included a module presenting five clinical vignettes, which is reported separately.The questionnaire was administered online between June and December 2021. It was disseminated nationwide through HCPs networks and learned societies representing paediatricians, medical geneticists, genetic counsellors, gynaecologists, and midwives. Dissemination relied on professional mailing lists and, for several learned societies, publication of the survey link on their official websites. Participation was voluntary and not accompanied by any financial reward. All participants were invited to provide informed consent for the processing of data in accordance with the research protocol.Analyses were primarily descriptive. Categorical variables were summarised as counts and percentages within each module-specific analytical subsample. Missing data were reported explicitly and were not imputed, and open text responses were used illustratively. Items on categories of genetic findings and on NGS were displayed in two-dimensional landscape plots (R 4.4.2), showing for each item overall agreement, variability across specialties, and the min-max range of specialty-specific responses. “No opinion” and “Not my field of expertise” were retained as distinct categories over the full module sample. Because the propensity to select “not my field of expertise” varied across specialties, agreement and between-specialty comparisons were also computed among positioned respondents only, that is those expressing agreement or disagreement. Response distributions were then compared across professional and sociodemographic characteristics. All tests were global tests of association. A chi-squared test with Monte Carlo simulated p-values was used, and Fisher’s exact test when expected cell counts fell below five.ResultsSociodemographic and professional characteristicsRespondents covered a wide range of specialties and practice settings, and the composition of the three analytical samples was closely similar (Table 1). Paediatricians were the largest group (462, 43% in the criteria module), followed by midwives (262, 24%), gynaecologists (124, 12%), medical geneticists (119, 11%) and genetic counsellors (66, 6%). Two thirds were under 50. Half practised in a university hospital and just over half reported prior experience in genetics practice. A full description of each sample is given in Supplementary Table ST1.Table 1 Characteristics of respondents in each analytical sample.Full size tableCriteria for selecting conditionsAll ten criteria adapted from Wilson and Jungner were rated as important, but the level of importance attributed to them differed (Fig. 1a, N = 1077). Four received the highest support, each rated very important by 83 to 87% of respondents: the superiority of early treatment outcomes (WJ4), the existence of a detectable early stage (WJ3), test performance (WJ5) and test acceptability to the population (WJ6). Two were more often rated of secondary importance: the requirement that costs be offset by expected benefits (WJ10, 35% very important) and the possibility of repeating the test (WJ8, 23%). The superiority of early treatment outcomes was designated the single most important criterion by 59% of respondents, far ahead of any other; the cost criterion was designated the least important by 36% and the possibility of repeating the test by 30% (Supplementary Fig. SF3).Fig. 1: Views on the criteria adapted from Wilson and Jungner (a, WJ1-WJ10) and on the Andermann criteria (b, And1-And10) (N = 1,077).Full size imageStacked bars show the distribution of responses for each criterion (“not at all important”, “moderately important”, “very important”, “no opinion”). Percentages are calculated within each criterion, among respondents who answered it, so denominators vary slightly across criteria. Values within segments are the number of respondents and the corresponding percentage; segments accounting for less than 5% of responses are not labelled. The full wording of each item and its identifier are given in Supplementary Methods SM1.Among the Andermann criteria (Fig. 1b), equity of access to screening for the entire target population (And8) received the strongest support (94% very important), ahead of evidence of programme effectiveness (And4, 85%). Support was markedly lower for the criteria concerning programme design and governance: evaluation planned from the outset (And9, 63%), response to a recognised need (And1, 59%) and a defined target population (And3, 49%). The requirement that benefits outweigh harms, present in both frameworks, was strongly supported in each (WJ9, 84%; And10, 89%). Sixty-seven HCPs (6% of sample) suggested additional criteria. None was new: HCPs reiterated treatability and early intervention, with direct benefit to the child prevailing over prevalence or cost (Supplementary Analysis SA1).Response distributions differed by specialty, prior experience in genetics, academic activity and practice setting, most frequently for the Andermann criteria concerning governance, evaluation and safeguards and for the forced ranking (Supplementary Analysis SA2).Genetic findings to be returnedIn this module (N = 874), 44% of HCPs considered the current scope of NBS in France insufficient, 29% judged it adequate, and 27% expressed no opinion.Support for returning a category of findings followed its actionability rather than the age at which the condition manifests (Table 2). Agreement was highest for childhood-onset conditions amenable to a treatment that would change the prognosis (98% strongly or somewhat agree) or lead to the disappearance of symptoms (96%) and remained high when the same benefit applied only in adulthood (82% and 72% respectively). It fell sharply for conditions for which no medical action is possible, whether in childhood (23%) or in adulthood (14%), and for variations with uncertain clinical implications (VUS, 12%). Findings whose benefit is indirect occupied an intermediate position: pharmacogenetic variations modifying the child’s medical management (66%) and conditions whose screening could reduce long-term costs, irrespective of age of onset or availability of treatment (58%). Findings with no medical benefit for the child received limited support: information on the health of relatives (31%) and heterozygous carrier status (30%). Full distribution is given in Supplementary Fig. SF4.Table 2 Support for returning each category of genetic finding, overall and by professional specialty (N = 874).Full size tableSupport for childhood-onset actionable conditions was uniformly high across specialties, ranging from 94% to 100%. On every other category, medical geneticists were the least supportive and midwives the most (all p