Department of Geriatric Medicine, ZGT, Netherland.
Department of Clinical Epidemiology, Biostatistics and Bioinformatics, Academic Medical Center, University of Amsterdam.
Injury. 2020 Aug;51(8):1846-1850. doi: 10.1016/j.injury.2020.05.017. Epub 2020 May 21.
Hip fracture in the elderly is associated with increased morbidity and mortality. Pneumonia during hospitalization is not uncommon and is associated with poorer outcomes, such as an increased risk of readmission and higher mortality rates. We aim to identify independent predictive factors for developing pneumonia during hospitalization in this group and also assessed the effect pneumonia has on hospital-stay, in-hospital and 30-day mortality.
Retrospective cohort study with prospectively collected data from hospitalized elderly hip fracture patients between January 2015 and January 2017. Examined predictors were age, gender, pre-fracture living situation, pre-fracture mobility score, pre-fracture ADL-status, history of dementia, diabetes, congestive heart failure, chronic obstructive pulmonary disease and prior stroke, ASA-score, anemia at admission, surgery within 48 hours, surgical procedure and anesthesia used. Multivariable regression analysis including resampling methods (bootstrapping) was used to examine the effects of predictors.
Of 407 patients, 62 (15.2%) were treated for pneumonia during hospitalization. Only gender, surgery within 48 hours and history of COPD differed significantly at baseline between the groups with and without pneumonia. Adjusted for age and gender, we observed a 1.6 times longer hospital-stay (95% CI 1.4-1.9, p<0.001), higher in-hospital mortality (OR 8.0, 95% CI 2.97-22.29) and 30-day mortality (OR 3.22, 95% CI 1.44-6.94) in the pneumonia-group. Pneumonia explained 9.1% of the variance in the length of hospital-stay aside from age and gender. Eight candidate predictors from the univariate analyses with a p<0.20 were selected for a multivariable logistic regression in 1000 bootstrap samples. Gender and history of COPD were most often found to have a p<0.10 (61.3% and 58.2%, respectively) in the bootstrap analyses and more than 80% stability in their B-coefficient signs. The discriminative quality of these two variables alone resulted in an AUC of 61.7% (95% CI 54%-69%).
Pneumonia resulted in longer hospital-stay and higher mortality rates. Of the 15 selected potential risk-factors for developing pneumonia during admission, male gender and history of COPD appeared to have the best potential as predictors. The other risk-factors had poorer performance, probably due to the few events and limited occurrence of some candidate variables in our study population.
老年人髋部骨折与发病率和死亡率的增加有关。住院期间肺炎并不少见，且与较差的预后相关，如再入院风险增加和更高的死亡率。我们旨在确定该人群住院期间发生肺炎的独立预测因素，并评估肺炎对住院时间、住院期间和 30 天死亡率的影响。
这是一项回顾性队列研究，对 2015 年 1 月至 2017 年 1 月期间住院的老年髋部骨折患者前瞻性收集数据。检查的预测因素包括年龄、性别、骨折前生活状况、骨折前活动能力评分、骨折前日常生活活动能力状态、痴呆史、糖尿病、充血性心力衰竭、慢性阻塞性肺疾病和既往卒中、ASA 评分、入院时贫血、48 小时内手术、手术类型和使用的麻醉。使用多变量回归分析（包括重采样方法[自举法]）来检查预测因素的影响。
在 407 名患者中，有 62 名（15.2%）在住院期间因肺炎接受治疗。仅性别、48 小时内手术和慢性阻塞性肺疾病病史在有肺炎和无肺炎组之间的基线水平上有显著差异。在调整年龄和性别后，我们观察到肺炎组的住院时间延长了 1.6 倍（95%CI 1.4-1.9，p<0.001），住院期间死亡率更高（OR 8.0，95%CI 2.97-22.29）和 30 天死亡率（OR 3.22，95%CI 1.44-6.94）。肺炎除年龄和性别外，还解释了住院时间长短差异的 9.1%。单变量分析中 p<0.20 的 8 个候选预测因素在 1000 次自举样本的多变量逻辑回归中被选择。在自举分析中，性别和慢性阻塞性肺疾病史最常发现 p<0.10（分别为 61.3%和 58.2%），并且其 B 系数符号的稳定性超过 80%。这两个变量单独的判别质量导致 AUC 为 61.7%（95%CI 54%-69%）。
肺炎导致住院时间延长和死亡率升高。在 15 个选定的入院期间发生肺炎的潜在风险因素中，男性性别和慢性阻塞性肺疾病史似乎是最好的预测因素。其他风险因素的表现较差，可能是由于我们研究人群中的事件较少和一些候选变量的发生有限。
