The value of genomic testing in severe childhood speech disorders

Childhood apraxia of speech (CAS) is a complex neurodevelopmental disorder affecting 1–2 per 1000 children, characterized by impaired planning and sequencing of speech movements, leading to reduced precision, consistency, and intelligibility. Children with CAS often require intensive therapy, with persistent impacts on psychosocial, literacy, educational, and employment outcomes, and current therapies frequently target symptoms rather than causes. Historically, limited understanding of etiology has hindered targeted interventions and led families through lengthy, costly, and stressful diagnostic odysseys. Recent gene discoveries indicate substantial genetic contributions from single-gene variants, with genomic testing yielding diagnoses in up to one in three children. A genomic diagnosis can shift clinical care from watchful waiting or trial-and-error therapy toward precision medicine. Despite emerging evidence that genomic testing provides personal utility and is valued by patients and society, no prior evidence existed for severe speech disorders. This study aims to elicit preferences and estimate the value and uptake of genomic testing for severe childhood speech disorders from the perspectives of the Australian public and parents of affected children, to inform cost–benefit analyses for translation into the Australian health system.

The paper situates its contribution within a growing body of discrete choice experiment (DCE) studies valuing genomic testing. Prior Australian public estimates include willingness-to-pay (WTP) of AU$5650 for genomic testing in complex pediatric neurological disorders and AU$5470–$15,250 across pediatric genetic scenarios, depending on informational benefits. For critically ill infants/children, the value of genomic testing relative to standard care ranged from AU$9810–$11,500 depending on turnaround time. In Canada, parents of children with rare suspected genetic diseases reported WTP of CAD$6590 to obtain a diagnostic result. The authors also note evidence that clinical utility alone may not capture the total value of genomics; non-health outcomes (e.g., waiting time, knowledge gained, and access to peer/support groups) contribute substantially to perceived benefits and should be incorporated into evaluations.

Study design: A discrete choice experiment (DCE) elicited preferences for attributes of genomic testing in severe speech disorders. Attributes were identified and refined via five deliberative focus groups (May–June 2021) with parents of children with severe speech disorders, guided by literature and expert input. Participants and surveys: Two surveys were administered—(1) a representative Australian public sample (n=951) recruited via a survey vendor with quotas on age, gender, income; and (2) parents of children with severe speech disorders (n=56) recruited from the Genetics in Speech project, Murdoch Children’s Research Institute (MCRI), and the Royal Children’s Hospital (RCH). Surveys included demographics, information about genomic testing and severe speech disorders, attribute descriptions, and an example choice task. The public survey included additional explanatory materials (including a video of a child with CAS) to address familiarity differences. A cheap talk script was used to mitigate hypothetical bias. Attribute set and levels (7 attributes):

• Number of children who receive a genetic diagnosis (20/100, 30/100, 40/100, 50/100)
• Knowledge about future health and development (prognosis): no knowledge, some knowledge, a lot of knowledge
• Chance of improving the process of the child’s medical care now (20%, 30%, 40%, 50%)
• Time to test (1, 3, 6 months)
• Out-of-pocket cost (A$500, A$1,500, A$3,000, A$4,500)
• Allowing access to educational support services (Yes/No)
• Enabling access to relevant genetic-based family support groups (Yes/No) Experimental design: A Bayesian D-efficient design (Ngene) with priors from pilots and literature. Choice tasks were presented in 4 blocks; each participant received 12 tasks with three alternatives (Situation 1, Situation 2, Neither). Blocking minimized correlation and task burden. Surveys were implemented in Qualtrics; pilots included 115 public respondents and 14 parent focus group members; wording refined based on feedback. Statistical analysis: Choice data analyzed using a panel random parameter error component mixed logit model (Nlogit 6) to account for preference heterogeneity and alternative-specific error correlation. Cost and time were continuous with constrained triangular distributions; other attributes dummy-coded with normally distributed random parameters. Random parameters estimated using 500 Halton draws. Attribute importance was normalized to reflect absolute utility changes. Marginal WTPs were derived from individual-level conditional estimates. Expected uptake and total WTP for genomic testing were computed via the compensating variation approach, assuming zero out-of-pocket cost for policy-relevant valuation. A latent class panel model further explored preference heterogeneity. Values reported in Australian dollars.

• Preferences and significance: Statistically significant preferences across all seven attributes in the public sample; in the parent sample, all attributes were significant except access to genetic-based family support groups.
• Direction of effects: Higher diagnostic yield, more knowledge from diagnosis, higher chance of improving current medical care, shorter time to testing, lower cost, and enabling access to educational support and family support groups all increased preference for testing. Time and cost reduced utility as expected.
• Overall value and uptake (assuming zero cost): Public average WTP = AU$7489 (US$5021) with predicted uptake 94.2%; Parents’ average WTP = AU$4452 (US$2985) with predicted uptake 99.6%.
• Attribute importance (public): Cost and waiting time were most influential; knowledge gained, diagnostic yield, and educational support also important.
• Marginal WTP (public, selected): AU$45 per additional child per 100 diagnosed; AU$1478 for some knowledge and AU$3092 for a lot of knowledge vs none; AU$90 per percentage-point increase in chance of improving current care; AU$201 to reduce waiting time by one month; AU$1140 for access to educational support services; AU$520 for access to genetic-based family support groups.
• Marginal WTP (parents, selected): AU$2352 for some knowledge; AU$4207 for a lot of knowledge; AU$93 per percentage-point increase in chance of improving current care; AU$169 to reduce waiting time by one month; AU$3094 for educational support services; AU$658 for family support groups. Parents placed markedly higher value on educational support services than the public.
• Preference heterogeneity: Random parameter variances were significant, indicating heterogeneity. Latent class analysis (public) identified three classes: (1) strong pro-testing preference; (2) cost-sensitive, especially valuing educational support; (3) low valuation of genomic testing regardless of attributes.

The study addresses the gap in quantitative evidence on the value and uptake of genomic testing for severe childhood speech disorders by demonstrating high willingness-to-pay and predicted uptake from both the public and affected families. Findings indicate that beyond clinical utility, non-health outcomes—such as knowledge gained, reduced waiting time, and access to educational and peer support—substantially influence preferences. This underscores the need for decision-makers to incorporate personal utility and broader benefits into health economic evaluations of genomic technologies, ensuring equitable prioritization. The strong value placed on educational support suggests genomic results can facilitate tailored educational planning and inclusion in mainstream settings, potentially improving broader life outcomes. Preference heterogeneity implies that shared decision-making should be tailored to different patient and family profiles, acknowledging groups with strong pro-testing preferences, cost sensitivity, or low valuation of testing.

This study provides empirical evidence of high personal utility and predicted uptake of genomic testing for severe speech disorders (e.g., CAS) from societal and parent perspectives. Both groups value increased diagnostic yield, greater knowledge, improved medical care processes, quicker access, lower cost, and access to educational and peer supports. Results support incorporating personal and non-health benefits into evaluations and will facilitate cost–benefit analyses to inform the translation of genomic testing into publicly funded health care in Australia. Future research should explore preference heterogeneity with larger parent samples and consider the influence of families’ positions within the diagnostic journey.

• Hypothetical nature of DCE choices may not perfectly reflect real-world decisions; a cheap talk script and educational materials (e.g., CAS video) were used to mitigate bias.
• Public sample recruited via a market research company may be subject to self-selection or incentive biases.
• Parent sample consisted of families whose children had undergone genomic testing (with findings in one-third), potentially limiting generalizability to parents without testing experience; small parent sample size reduced power for some attributes.
• The study did not control for families’ stage in the diagnostic journey, which may affect the value placed on testing.