School of Medicine, The Johns Hopkins University, Baltimore, MD, USA.
Whiting School of Engineering, The Johns Hopkins University, Baltimore, MD, USA.
J Diabetes Sci Technol. 2024 Mar;18(2):302-308. doi: 10.1177/19322968231201654. Epub 2023 Oct 5.
In the pivotal clinical trial that led to Food and Drug Administration De Novo "approval" of the first fully autonomous artificial intelligence (AI) diabetic retinal disease diagnostic system, a reflexive dilation protocol was used. Using real-world deployment data before implementation of reflexive dilation, we identified factors associated with nondiagnostic results. These factors allow a novel workflow, where patients most likely to benefit from pharmacologic dilation are dilated to maximize efficiency and patient satisfaction.
Retrospective review of patients who were assessed with autonomous AI at Johns Hopkins Medicine (8/2020 to 5/2021). We constructed a multivariable logistic regression model for nondiagnostic results to compare characteristics of patients with and without diagnostic results, using adjusted odds ratio (aOR). < .05 was considered statistically significant.
Of 241 patients (59% female; median age = 59), 123 (51%) had nondiagnostic results. In multivariable analysis, type 1 diabetes (T1D, aOR = 5.82, 95% confidence interval [CI]: 1.45-23.40, = .01), smoking (aOR = 2.86, 95% CI: 1.36-5.99, = .005), and age (every 10-year increase, aOR = 2.12, 95% CI: 1.62-2.77, < .001) were associated with nondiagnostic results. Following feature elimination, a predictive model was created using T1D, smoking, age, race, sex, and hypertension as inputs. The model showed an area under the receiver-operator characteristics curve of 0.76 in five-fold cross-validation.
We used factors associated with nondiagnostic results to design a novel, predictive dilation workflow, where patients most likely to benefit from pharmacologic dilation are dilated . This new workflow has the potential to be more efficient than reflexive dilation, thus maximizing the number of at-risk patients receiving their diabetic retinal examinations.
在导致食品和药物管理局首次批准完全自主人工智能（AI）糖尿病视网膜疾病诊断系统的关键临床试验中，使用了反射性散瞳方案。利用在实施反射性散瞳之前的真实世界部署数据，我们确定了与非诊断结果相关的因素。这些因素允许采用一种新的工作流程，即对最有可能从药物散瞳中获益的患者进行散瞳，以最大程度地提高效率和患者满意度。
回顾性分析 2020 年 8 月至 2021 年 5 月在约翰霍普金斯医学中心接受自主 AI 评估的患者。我们构建了一个多变量逻辑回归模型，用于非诊断结果，以比较有和无诊断结果的患者特征，使用调整后的优势比（aOR）。<.05 被认为具有统计学意义。
在 241 名患者中（59%为女性；中位年龄=59 岁），有 123 名（51%）患者的结果为非诊断性的。在多变量分析中，1 型糖尿病（T1D，aOR=5.82，95%置信区间[CI]：1.45-23.40，=.01）、吸烟（aOR=2.86，95%CI：1.36-5.99，=.005）和年龄（每增加 10 岁，aOR=2.12，95%CI：1.62-2.77，<.001）与非诊断结果相关。在特征消除后，使用 T1D、吸烟、年龄、种族、性别和高血压作为输入，创建了一个预测模型。该模型在五重交叉验证中的接收者操作特征曲线下面积为 0.76。
我们使用与非诊断结果相关的因素设计了一种新的预测性散瞳工作流程，对最有可能从药物散瞳中获益的患者进行散瞳。这种新的工作流程有可能比反射性散瞳更有效，从而最大限度地增加接受糖尿病视网膜检查的高危患者数量。
