Abstract：
OBJECTIVE: To analyze the prognostic accuracy of the scores NEWS, qSOFA, GYM used in hospital emergency department (ED) in the assessment of elderly patients who consult for an infectious disease.
METHODS: Data from the EDEN (Emergency Department and Elderly Need) cohort were used. This retrospective cohort included all patients aged ≥65 years seen in 52 Spanish EDs during two weeks (from 1-4-2019 to 7-4-2019 and 30/3/2020 to 5/4/2020) with an infectious disease diagnosis in the emergency department. Demographic variables, demographic variables, comorbidities, Charlson and Barthel index and needed scores parameters were recorded. The predictive capacity for 30-day mortality of each scale was estimated by calculating the area under the receiver operating characteristic (ROC) curve, and sensitivity and specificity were calculated for different cut-off points. The primary outcome variable was 30-day mortality.
RESULTS: 6054 patients were analyzed. Median age was 80 years (IQR 73-87) and 45.3% women. 993 (16,4%) patients died. NEWS score had better AUC than qSOFA (0.765, 95CI: 0.725-0.806, versus 0.700, 95%CI: 0.653-0.746; P < .001) and GYM (0.716, 95%CI: 0.675-0.758; P = .024), and there was no difference between qSOFA and GYM (P = .345). The highest sensitivity scores for 30-day mortality were GYM ≥ 1 point (85.4%) while the qSOFA score ≥2 points showed high specificity. In the case of the NEWS scale, the cut-off point ≥4 showed high sensitivity, while the cut-off point NEWS ≥ 8 showed high specificity.
CONCLUSIONS: NEWS score showed the highest predictive capacity for 30-day mortality. GYM score ≥1 showed a great sensitivity, while qSOFA ≥2 scores provide the highest specificity but lower sensitivity.
METHODS: Data from the EDEN (Emergency Department and Elderly Need) cohort were used. This retrospective cohort included all patients aged ≥65 years seen in 52 Spanish EDs during two weeks (from 1-4-2019 to 7-4-2019 and 30/3/2020 to 5/4/2020) with an infectious disease diagnosis in the emergency department. Demographic variables, demographic variables, comorbidities, Charlson and Barthel index and needed scores parameters were recorded. The predictive capacity for 30-day mortality of each scale was estimated by calculating the area under the receiver operating characteristic (ROC) curve, and sensitivity and specificity were calculated for different cut-off points. The primary outcome variable was 30-day mortality.
RESULTS: 6054 patients were analyzed. Median age was 80 years (IQR 73-87) and 45.3% women. 993 (16,4%) patients died. NEWS score had better AUC than qSOFA (0.765, 95CI: 0.725-0.806, versus 0.700, 95%CI: 0.653-0.746; P < .001) and GYM (0.716, 95%CI: 0.675-0.758; P = .024), and there was no difference between qSOFA and GYM (P = .345). The highest sensitivity scores for 30-day mortality were GYM ≥ 1 point (85.4%) while the qSOFA score ≥2 points showed high specificity. In the case of the NEWS scale, the cut-off point ≥4 showed high sensitivity, while the cut-off point NEWS ≥ 8 showed high specificity.
CONCLUSIONS: NEWS score showed the highest predictive capacity for 30-day mortality. GYM score ≥1 showed a great sensitivity, while qSOFA ≥2 scores provide the highest specificity but lower sensitivity.
摘要：
目的：分析NEWS评分的预后准确性，qSOFA,GYM在医院急诊科（ED）中用于评估咨询传染病的老年患者。
方法：使用来自EDEN（急诊科和老年人需求）队列的数据。该回顾性队列包括在2周内（从2019年1月4日至2019年7月4日和2020年3月30日至5月4日）在52个西班牙ED中观察到的所有年龄≥65岁的患者，并在急诊科诊断为传染病。人口统计变量，人口统计学变量，合并症,记录Charlson和Barthel指数以及所需的评分参数。通过计算受试者工作特征（ROC）曲线下的面积来估计每个量表对30天死亡率的预测能力，并计算不同截止点的敏感性和特异性。主要结果变量是30天死亡率。
结果：对6054例患者进行分析。中位年龄为80岁（IQR73-87），女性占45.3％。993例（16.4%）患者死亡。新闻评分的AUC优于qSOFA（0.765，95CI：0.725-0.806，与0.700，95CI：0.653-0.746；P<.001）和GYM（0.716，95CI：0.675-0.758；P=.024），qSOFA和GYM之间没有差异（P=.345）。30天死亡率的最高敏感性评分为GYM≥1分（85.4％），而qSOFA评分≥2分显示出高特异性。在新闻规模的情况下，截止点≥4显示高灵敏度，而截止点NEWS≥8显示出高特异性。
结论：NEWS评分显示30天死亡率的预测能力最高。GYM评分≥1分表现出很高的敏感性，而qSOFA≥2分的特异性最高，但敏感性较低。
方法：使用来自EDEN（急诊科和老年人需求）队列的数据。该回顾性队列包括在2周内（从2019年1月4日至2019年7月4日和2020年3月30日至5月4日）在52个西班牙ED中观察到的所有年龄≥65岁的患者，并在急诊科诊断为传染病。人口统计变量，人口统计学变量，合并症,记录Charlson和Barthel指数以及所需的评分参数。通过计算受试者工作特征（ROC）曲线下的面积来估计每个量表对30天死亡率的预测能力，并计算不同截止点的敏感性和特异性。主要结果变量是30天死亡率。
结果：对6054例患者进行分析。中位年龄为80岁（IQR73-87），女性占45.3％。993例（16.4%）患者死亡。新闻评分的AUC优于qSOFA（0.765，95CI：0.725-0.806，与0.700，95CI：0.653-0.746；P<.001）和GYM（0.716，95CI：0.675-0.758；P=.024），qSOFA和GYM之间没有差异（P=.345）。30天死亡率的最高敏感性评分为GYM≥1分（85.4％），而qSOFA评分≥2分显示出高特异性。在新闻规模的情况下，截止点≥4显示高灵敏度，而截止点NEWS≥8显示出高特异性。
结论：NEWS评分显示30天死亡率的预测能力最高。GYM评分≥1分表现出很高的敏感性，而qSOFA≥2分的特异性最高，但敏感性较低。
