探讨改良Morse跌倒风险评估量表对老年住院患者跌倒风险的预测价值。

选取2020年6月至2021年12月中国人民解放军联勤保障部队第九〇三医院老年病诊治中心接诊且经老年综合评估（CGA）软件筛查的老年患者238例，其中包括48例跌倒患者。对比改良前后的Morse跌倒风险评估量表预测高危跌倒人群的情况，并利用受试者工作特征（ROC）曲线比较量表预测跌倒风险的曲线下面积（AUC）、灵敏度、特异度和约登指数等。

Morse跌倒风险评估量表结果显示，跌倒高风险86例，中风险109例，低风险43例；改良Morse跌倒风险评估量表结果显示，跌倒高风险86例，中风险111例，低风险41例。48例跌倒患者改良前及改良后的评分均高于未跌倒患者（P<0.01），但改良后量表评估为中风险患者的占比较改良前上升（改良前后分别为6.42%和8.10%），而低风险患者的占比下降（改良前后分别为11.63%和7.31%），高风险患者占比无变化。ROC曲线分析结果显示，Morse跌倒风险评估量表预测跌倒的AUC为0.828（95%CI：0.773～0.873，P<0.01），改良Morse跌倒风险评估量表的AUC提升至0.848（95%CI：0.795～0.891，P<0.01），但两者的差异无统计学意义（Z=1.116，P>0.05）。Morse跌倒风险评估量表的灵敏度为64.58%，特异度为95.79%，约登指数为0.604；改良Morse跌倒风险评估量表的灵敏度为60.42%，特异度为97.89%，约登指数为0.583。

基于CGA改良后的Morse跌倒风险评估量表可在一定程度上提高跌倒风险预警的准确性。

To improve the Morse Fall Risk Assessment Scale using the Geriatric Comprehensive Assessment (CGA) technique and explore its role in fall risk management for elderly hospitalized patients.

From June 2020 to December 2021, our center admitted 238 patients who completed the Morse Fall Assessment Scale and Comprehensive Geriatric Assessment (CGA) software screening, including 48 patients with falls. Compare the Morse Fall Risk Assessment Scale before and after improvement to predict high-risk fall populations, and use receiver operating characteristic (ROC) curves to compare the area under the curve (AUC), sensitivity, specificity, and Jordan index of the scale in predicting fall risk.

According to the Morse Fall Risk Assessment Scale, among the 238 patients, 86 were of high risk of falling, 109 were of medium risk, and 43 were of low risk; According to the modified Morse Fall Risk Assessment Scale, among the 238 patients, 86 were of high risk of falling, 111 were of medium risk, and 41 were of low risk.The scores of 48 fall patients before and after improvement were higher than those of non fall patients (P < 0.01), but the proportion of medium risk patients evaluated by the improved scale increased compared to before improvement (6.42% and 8.10% respectively), while the proportion of low-risk patients decreased (11.63% and 7.31% respectively), and the proportion of high-risk patients remained unchanged. The ROC curve analysis results showed that the AUC of the Morse Fall Risk Scoring Scale for predicting falls was 0.828 (95%CI: 0.773-0.873, P < 0.01), and the AUC of the modified Morse Fall Risk Scoring Scale was improved to 0.848 (95%CI: 0.795-0.891, P < 0.01), but the difference between the two was not statistically significant (Z=1.116, P > 0.05). The sensitivity of the Morse Fall Risk Scoring Scale was 64.58%, the specificity was 95.79%, and the Youden index was 0.604; The sensitivity of the modified Morse Fall Risk Scoring Scale was 60.42%, the specificity was 97.89%, and the Youden index was 0.583.

The Morse Fall Risk Assessment Scale, improved based on CGA, can improve the accuracy of fall risk prediction.