External validation of the Hospital Frailty Risk Score and comparison with the Hospital-patient One-year Mortality Risk Score to predict outcomes in elderly hospitalised patients: a retrospective cohort study
External validation of the Hospital Frailty Risk Score and comparison with the Hospital-patient One-year Mortality Risk Score to predict outcomes in elderly hospitalised patients: a retrospective cohort study
复制标题
DOI：
10.1136/bmjqs-2018-008661
复制
发表时间：
2019-04-01
影响因子：
5.4
通讯作者：
van Walraven, Carl
中科院分区：
文献类型：
--
作者：
McAlister, Finlay;van Walraven, Carl
目的脆弱是住院患者的一个重要预后因素，但通常需要训练有素的观察员进行面对面的评估才能发现。因此，对于那些有兴趣实施质量改进措施以优化其结果的人来说，从系统的角度来看，虚弱的患者并不是很明显。这项研究旨在外部验证和比较两个最近描述的工具，使用管理数据作为潜在的脆弱标记：医院脆弱风险评分(HFRS)和医院-患者一年死亡风险(HOMR)评分。设计回溯性队列研究。设置加拿大安大略省。参与者所有75岁以上的患者在2004至2010年间至少有一次紧急非精神科住院。主要结果衡量延长住院时间(>；结果在452785例(25.9%为中高危HFRS)患者中，HFRS增加与Charlson评分高、年龄大、基线独立的可能性降低有关。高或中度肾综合征出血热患者延长住院的风险显著增加(70.0%(OR 8.64，95%CI 8.30至8.99)或49.7%(OR 3.66，95%CI 3.60至3.71)vs21.3%)和30天死亡率(15.5%(OR 1.27，95%CI 1.20至1.33)或16.8%(OR 1.39，但30天再入院的风险(10.0%(OR 0.74，95%CI 0.69至0.79)和11.2%(OR 0.84，95%CI 0.82至0.86)vs 13.1%)或急诊(7.3%(OR 0.41，95%CI 0.38至0.45)和11.1%(OR 0.66，95%可信区间为0.38比0.45)，对16.0%)。虽然只有松散的相关性(皮尔逊相关系数0.265，p<；0.0001)，但肾综合征出血热评分和HOMR评分都与每个结局独立相关--HFRS评分与住院时间延长的相关性更强(C统计量为0.71)，HOMR评分与30天死亡率的相关性更强(C统计量为0.71)。两者对再入院30天的预测都很差(C-统计数字，HFRS为0.52，HOMR评分为0.54)。结论HFRS能更好地识别住院老年患者延长住院时间的风险，而HOMR评分能更好地预测30天死亡率。然而，两种评分都不适合预测出院后30天内再入院或急诊的风险。因此，单一的评分不足以预测与脆弱相关的所有结果。
Objective Frailty is an important prognostic factor in hospitalised patients but typically requires face-to-face assessment by trained observers to detect. Thus, frail patients are not readily apparent from a systems perspective for those interested in implementing quality improvement measures to optimise their outcomes. This study was designed to externally validate and compare two recently described tools using administrative data as potential markers for frailty: the Hospital Frailty Risk Score (HFRS) and the Hospital-patient One-year Mortality Risk (HOMR) Score.Design Retrospective cohort study.Setting Ontario, Canada.Participants All patients over 75 with at least one urgent non-psychiatric hospitalisation between 2004 and 2010.Main outcome measures Prolonged hospital length of stay (> 10 days), 30-day mortality after admission and 30-day postdischarge rates of urgent readmission or emergency department (ED) visits.Results In 452 785 patients (25.9% with intermediate or high-risk HFRS), increased HFRS was associated with higher Charlson scores, older age and decreased likelihood of baseline independence. Patients with high or intermediate HFRS had significantly increased risks of prolonged hospitalisation (70.0% (OR 8.64, 95% CI 8.30 to 8.99) or 49.7% (OR 3.66, 95% CI 3.60 to 3.71) vs 21.3% in low-risk HFRS group) and 30-day mortality (15.5% (OR 1.27, 95% CI 1.20 to 1.33) or 16.8% (OR 1.39, 95% CI 1.36 to 1.41) vs 12.7% in low-risk), but decreased risks of 30-day readmission (10.0% (OR 0.74, 95% CI 0.69 to 0.79) and 11.2% (OR 0.84, 95% CI 0.82 to 0.86) vs 13.1%) or ED visit (7.3% (OR 0.41, 95% CI 0.38 to 0.45) and 11.1% (OR 0.66, 95% CI 0.38 to 0.45) vs 16.0%). Although only loosely associated (Pearson correlation coefficient 0.265, p< 0.0001), both the HFRS and HOMR Score were independently associated with each outcome-HFRS was more strongly associated with prolonged length of stay (C-statistic 0.71) and HOMR Score was more strongly associated with 30-day mortality (C-statistic 0.71). Both poorly predicted 30-day readmissions (C-statistics 0.52 for HFRS and 0.54 for HOMR Score).Conclusions The HFRS best identified hospitalised older patients at higher risk of prolonged length of stay and the HOMR score better predicted 30-day mortality. However, neither score was suitable for predicting risk of readmission or ED visit in the 30 days after discharge. Thus, a single score is inadequate to prognosticate for all outcomes associated with frailty.