Department of Speech, Language, and Hearing Sciences, Purdue University, West Lafayette, Indiana.
Department of Psychological Sciences, Purdue University, West Lafayette, Indiana.
JAMA Netw Open. 2024 May 1;7(5):e2411190. doi: 10.1001/jamanetworkopen.2024.11190.
Finding effective and scalable solutions to address diagnostic delays and disparities in autism is a public health imperative. Approaches that integrate eye-tracking biomarkers into tiered community-based models of autism evaluation hold promise for addressing this problem.
To determine whether a battery of eye-tracking biomarkers can reliably differentiate young children with and without autism in a community-referred sample collected during clinical evaluation in the primary care setting and to evaluate whether combining eye-tracking biomarkers with primary care practitioner (PCP) diagnosis and diagnostic certainty is associated with diagnostic outcome.
DESIGN, SETTING, AND PARTICIPANTS: Early Autism Evaluation (EAE) Hub system PCPs referred a consecutive sample of children to this prospective diagnostic study for blinded eye-tracking index test and follow-up expert evaluation from June 7, 2019, to September 23, 2022. Participants included 146 children (aged 14-48 months) consecutively referred by 7 EAE Hubs. Of 154 children enrolled, 146 provided usable data for at least 1 eye-tracking measure.
The primary outcomes were sensitivity and specificity of a composite eye-tracking (ie, index) test, which was a consolidated measure based on significant eye-tracking indices, compared with reference standard expert clinical autism diagnosis. Secondary outcome measures were sensitivity and specificity of an integrated approach using an index test and PCP diagnosis and certainty.
Among 146 children (mean [SD] age, 2.6 [0.6] years; 104 [71%] male; 21 [14%] Hispanic or Latine and 96 [66%] non-Latine White; 102 [70%] with a reference standard autism diagnosis), 113 (77%) had concordant autism outcomes between the index (composite biomarker) and reference outcomes, with 77.5% sensitivity (95% CI, 68.4%-84.5%) and 77.3% specificity (95% CI, 63.0%-87.2%). When index diagnosis was based on the combination of a composite biomarker, PCP diagnosis, and diagnostic certainty, outcomes were concordant with reference standard for 114 of 127 cases (90%) with a sensitivity of 90.7% (95% CI, 83.3%-95.0%) and a specificity of 86.7% (95% CI, 70.3%-94.7%).
In this prospective diagnostic study, a composite eye-tracking biomarker was associated with a best-estimate clinical diagnosis of autism, and an integrated diagnostic model including PCP diagnosis and diagnostic certainty demonstrated improved sensitivity and specificity. These findings suggest that equipping PCPs with a multimethod diagnostic approach has the potential to substantially improve access to timely, accurate diagnosis in local communities.
找到有效的、可扩展的解决方案来解决自闭症诊断延迟和差异问题是公共卫生的当务之急。将眼动追踪生物标志物整合到基于社区的自闭症评估分层模型中的方法有望解决这个问题。
确定一系列眼动追踪生物标志物是否可以在初级保健环境中的临床评估中，从社区推荐的样本中可靠地区分有自闭症和无自闭症的幼儿，并评估将眼动追踪生物标志物与初级保健医生（PCP）诊断和诊断确定性相结合是否与诊断结果相关。
设计、地点和参与者：自闭症早期评估（EAE）中心的初级保健医生连续将一批儿童推荐给这个前瞻性诊断研究，进行盲法眼动追踪指标测试和后续专家评估，时间从 2019 年 6 月 7 日至 2022 年 9 月 23 日。参与者包括 7 个 EAE 中心连续推荐的 146 名儿童（年龄 14-48 个月）。在 154 名入组的儿童中，有 146 名儿童至少提供了 1 项可用的眼动追踪测量数据。
主要结果是基于显著眼动追踪指标的综合（即指数）测试的敏感性和特异性，与参考标准专家临床自闭症诊断进行比较。次要结果指标是使用指数测试和 PCP 诊断和确定性的综合方法的敏感性和特异性。
在 146 名儿童中（平均[标准差]年龄 2.6[0.6]岁；104[71%]男性；21[14%]西班牙裔或拉丁裔和 96[66%]非拉丁裔白人；102[70%]有参考标准自闭症诊断），113 名（77%）儿童的指数（综合生物标志物）和参考结果之间的自闭症结果是一致的，敏感性为 77.5%（95%CI，68.4%-84.5%），特异性为 77.3%（95%CI，63.0%-87.2%）。当指数诊断基于复合生物标志物、PCP 诊断和诊断确定性的组合时，127 例中有 114 例（90%）与参考标准一致，敏感性为 90.7%（95%CI，83.3%-95.0%），特异性为 86.7%（95%CI，70.3%-94.7%）。
在这项前瞻性诊断研究中，综合眼动追踪生物标志物与自闭症的最佳临床诊断相关，包括 PCP 诊断和诊断确定性的综合诊断模型显示出更高的敏感性和特异性。这些发现表明，为初级保健医生配备多方法诊断方法有可能大大改善当地社区及时、准确诊断的机会。
Zotero 插件