The Legend score synthesizes Wells, PERC, Geneva, D-dimer and predicts acute pulmonary embolism prior to imaging tests

Yunfeng Zhao; Yi Cheng; Hongwei Wang; He Du; Jinyuan Sun; Mei Xu; Yong Luo; Song Liu; Xuejun Guo; Wei Xiong

doi:10.1016/j.pulmoe.2023.10.002

The Legend score synthesizes Wells, PERC, Geneva, D-dimer and predicts acute pulmonary embolism prior to imaging tests

Pulmonology. 2023 Nov 10:S2531-0437(23)00195-2. doi: 10.1016/j.pulmoe.2023.10.002. Online ahead of print.

Authors

Yunfeng Zhao¹, Yi Cheng², Hongwei Wang², He Du³, Jinyuan Sun², Mei Xu⁴, Yong Luo², Song Liu², Xuejun Guo², Wei Xiong⁵

Affiliations

¹ Department of Pulmonary and Critical Care Medicine, Punan Hospital, Pudong New District, Shanghai, China.
² Department of Pulmonary and Critical Care Medicine, Xinhua Hospital, Shanghai Jiaotong University School of Medicine, Shanghai, China.
³ Department of Medical Oncology, Shanghai Pulmonary Hospital, Tongji University School of Medicine, Shanghai, China.
⁴ Department of General Practice, North Bund Community Health Service Center, Hongkou District, Shanghai, China.
⁵ Department of Pulmonary and Critical Care Medicine, Xinhua Hospital, Shanghai Jiaotong University School of Medicine, Shanghai, China; Department of Cardiovascular Medicine, Graduate School of Medicine, Kyoto University, Kyoto, Japan. Electronic address: xiongwei@xinhuamed.com.cn.

PMID: 37953212
DOI: 10.1016/j.pulmoe.2023.10.002

Abstract

Introduction: The prediction rules of acute pulmonary embolism(PE) before imaging tests recommended by the commonly used guidelines have low diagnostic efficiency if not combined with D-dimer, therefore it is necessary to seek for a prediction rule with higher diagnostic efficiency.

Methods: We designed a new score named Legend by synthesizing the scores of Wells, PERC, and Geneva, as well as D-dimer with patients in the development group(n = 2112), and then validated it in patients of validation group(n = 388). Diagnostic efficiency was also compared between Legend score and Wells+D-dimer (DD), PERC+DD, Geneva+DD, and YEARS+DD(YEAR algorithm).

Results: The Legend score comprised active cancer, D-dimer≥1000 ng/mL, DVT symptoms and/or signs, previous venous thromboembolism (VTE) history, and surgery, trauma, or immobilization in the past month. The sensitivity, specificity, Youden index, and area under the curve(AUC) were 0.985, 0.744, 0.729, and (0.861[0.796-0.925], P<0.001), respectively, for original Legend score, whereas were 0.982, 0.778, 0.760, and (0.871[0.823-0.920], P<0.001), respectively, for simplified Legend score. The Kappa coefficient and P value of McNemar test were 0.988 and 1.000, respectively, between the original and simplified Legend scores. In the validation group, the sensitivity, specificity, Youden index, and C-index were 0.971, 0.749, 0.720, and (0.838[0.781-0.896], P<0.001), respectively, for the original Legend score, whereas were 0.986, 0.715, 0.701, and (0.816[0.750-0.880], P = 0.001) respectively, for the simplified Legend score. The Kappa coefficient and P value of McNemar test between original Legend score and Wells+DD, PERC+DD, Geneva+DD, and YEARS+DD were (0.563, 0.001), (0.139, <0.001), (0.631, 0.006), and (0.732, 0.029), respectively. The Kappa coefficient and P value of McNemar test between simplified Legend score and aforementioned scores were (0.675, 0.009), (0.172, <0.001), (0.747, 0.001), and (0.883, 0.012), respectively.

Discussion: In view of the fact the Legend score reserves the efficient predictors and eliminates the inefficient ones in Wells, PERC, and revised Geneva scores, and incorporates D-dimer into it, a more efficient, modified, and user-friendly one has replaced the original ones.

Conclusions: The Legend score yields excellent diagnostic efficiency with good safety in the pretest prediction of acute PE prior to imaging tests. It also avoids more unnecessary imaging tests than Wells+DD, PERC+DD, Geneva+DD, or YEARS+DD.

Keywords: Diagnosis; Prediction rule; Pretest probability; Pulmonary embolism; Score.