The pediatric glucocorticoid toxicity index
The pediatric glucocorticoid toxicity index
复制标题
DOI：
10.1016/j.semarthrit.2022.152068
复制
发表时间：
2022-07-30
影响因子：
5
通讯作者：
Stone, John H.
中科院分区：
文献类型：
--
作者：
Brogan, Paul;Naden, Ray;Stone, John H.
目的：开发儿科糖皮质激素毒性指数（pGTI），这是一种标准化的加权临床结局评估，可测量糖皮质激素（GC）毒性随时间的变化。方法：7个专科14名专家参加。医生专家代表了GC在炎症性疾病治疗中发挥主要作用的多个亚专业：肾脏学，流变学，肿瘤学，内分泌学，遗传学，精神病学和母胎医学。9名研究者来自加拿大、欧洲或新西兰，5名来自美国。采用群体一致性方法和多准则决策分析。pGTI是对常见、重要和动态GC毒性的综合评估。这些毒性被组织成健康领域，分为轻微、中度或严重，并根据严重程度进行加权。使用1000 MindsTM软件平台，通过群体共识和多标准决策分析得出相对权重。两个定量评分包括源自pGTI数据的总体毒性特征：（1）累积恶化评分;和（2）总体改善评分。pGTI还包括GC副作用的定性、未加权记录，称为损害清单，其记录了不太常见的毒性，尽管可能严重，但不太可能随着GC剂量的变化而改变。结果：pGTI中包括107项毒性，损伤检查表中包括32项。为了评估pGTI对应于专家临床判断的程度，研究者根据临床判断将15例病例从最高到最低的GC毒性进行排名。然后将专家评级与pGTI的病例评级进行比较，得出了极好的一致性（加权kappa 0.86）。pGTI在开发和初步验证后迁移到数字环境中。该数字平台旨在确保临床上的易用性、应用的严谨性和评分的准确性。诊所工作人员输入生命体征、实验室结果和与pGTI评分相关的药物变化。临床医生记录GC肌病、皮肤毒性、情绪障碍和感染的发现。然后，pGTI算法将权重应用于这些原始数据并计算评分。嵌入式逻辑解释了年龄和性别相关的参考范围对几个健康领域的影响：血压，脂质代谢和骨密度。其他算法解释了生长领域中使用的身高Z评分的预期变化，从而解决了儿童GC毒性特有的问题。损伤检查表可确保GC毒性的全面测量，但不影响pGTI评分，因为评分的领域强调GC毒性的表现，这些表现可能在试验过程中发生变化。结论：我们描述了一个加权的综合毒性指数的发展和初步评价，用于评估儿童和青少年使用GC相关的发病率。开发pGTI数字平台对于执行必要的细微计算至关重要，以确保临床和研究环境中的严谨性，准确性和易用性。
Objectives: To develop a Pediatric glucocorticoid toxicity index (pGTI), a standardized, weighted clinical outcome assessment that measures change in glucocorticoid (GC) toxicity over time. Methods: Fourteen physician experts from 7 subspecialties participated. The physician experts represented multiple subspecialties in which GCs play a major role in the treatment of inflammatory disease: nephrology, rheumatology, oncology, endocrinology, genetics, psychiatry, and maternal-fetal medicine. Nine investigators were from Canada, Europe, or New Zealand, and 5 were from the United States. Group consensus methods and multi-criteria decision analysis were used. The pGTI is an aggregate assessment of GC toxicities that are common, important, and dynamic. These toxicities are organized into health domains graded as minor, moderate, or major and are weighted according to severity. The relative weights were derived by group consensus and multi-criteria decision analysis using the 1000MindsTM software platform. Two quantitative scores comprise the overall toxicity profile derived from pGTI data: (1) the Cumulative Worsening Score; and (2) the Aggregate Improvement Score. The pGTI also includes a qualitative, unweighted record of GC side-effects known as the Damage Checklist, which documents less common toxicities that, although potentially severe, are unlikely to change with varying GC dosing. Results: One hundred and seven (107) toxicity items were included in the pGTI and thirty-two (32) in the Damage Checklist. To assess the degree to which the pGTI corresponds to expert clinical judgement, the investigators ranked 15 cases by clinical judgement from highest to lowest GC toxicity. Expert rankings were then compared to case ranking by the pGTI, yielding excellent agreement (weighted kappa 0.86). The pGTI was migrated to a digital environment following its development and initial validation. The digital platform is designed to ensure ease-of-use in the clinic, rigor in application, and accuracy of scoring. Clinic staff enter vital signs, laboratory results, and medication changes relevant to pGTI scoring. Clinicians record findings for GC myopathy, skin toxicity, mood dysfunction, and infection. The pGTI algorithms then apply the weights to these raw data and calculate scores. Embedded logic accounts for the impact of age-and sex-related reference ranges on several health domains: blood pressure, lipid metabolism, and bone mineral density. Other algorithms account for anticipated changes in the height Z-scores used in the growth domain, thereby addressing a concern unique to GC toxicity in children. The Damage Checklist ensures comprehensive measurement of GC toxicity but does not contribute to pGTI scoring, because the scored domains emphasize manifestations of GC toxicity that are likely to change over the course of a trial. Conclusions: We describe the development and initial evaluation of a weighted, composite toxicity index for the assessment of morbidity related to GC use in children and adolescents. Developing the pGTI digital platform was essential for performing the nuanced calculations necessary to ensure rigor, accuracy, and ease-of-use in both clinic and research settings.