Public Health Institute, Federal University of Rio de Janeiro, Rio de Janeiro, Brazil.
Botucatu Medical School, São Paulo State University, Botucatu, Brazil.
PLoS One. 2020 Oct 9;15(10):e0240229. doi: 10.1371/journal.pone.0240229. eCollection 2020.
We aimed to examine whether the number of previous hospitalizations and the main diagnoses of those hospitalizations are associated with increased in-hospital hip fracture mortality for older people. That assessment is relevant because if those variables are shown to be associated with increased mortality, that finding could support their use as proxies for comorbidity burden for case-mix adjustment in statistical models seeking to compare the performance of hospitals regarding hip fracture mortality in settings with limited hospital information systems.
In this retrospective cohort study of all public hospital admissions for older adults with hip fractures in the city of Rio de Janeiro between 2010 and 2011, we used data from the Hospital Admission Information System database to examine the association between in-hospital mortality and the number of hospitalizations in the previous two years and their main diagnoses through logistic regression.
Among 1938 patients included in the study there were 103 (5.3%) in-hospital deaths. Although the presence of hospitalization episodes within the two years preceding the index hip fracture was associated with increased mortality (OR: 1.78, 95%CI: 1.07 to 2.97) we did not find evidence of a gradient of increased mortality with a growing number of previous hospitalizations. Additionally, several diseases recorded as main diagnoses of previous hospitalizations were not associated with increased mortality rates, as was expected based on existing knowledge on risk factors for decreased survival in older adults with hip fractures.
Our results suggest that, in settings where local hospital information systems have limited access to secondary diagnoses, the use of the number of previous hospitalizations or the main diagnoses associated with those hospitalizations as proxies for the profile of comorbidities of older adults with hip fractures may not be an effective way to adjust for case-mix when comparing in-hospital mortality rates among hospitals.
我们旨在研究老年人既往住院次数及其主要诊断与院内髋部骨折死亡率增加的关系。这种评估是相关的，因为如果这些变量与死亡率增加相关，那么这一发现可以支持将其作为共病负担的替代指标，用于调整统计模型中的病例组合，以比较在医院信息系统有限的情况下，医院在髋部骨折死亡率方面的表现。
在这项对 2010 年至 2011 年里约热内卢市所有老年髋部骨折患者的回顾性队列研究中，我们使用来自医院入院信息系统数据库的数据，通过逻辑回归检验了住院次数和前两年主要诊断与院内死亡率之间的关系。
在纳入研究的 1938 名患者中，有 103 人（5.3%）在院内死亡。尽管在索引性髋部骨折前两年内存在住院事件与死亡率增加相关（比值比：1.78，95%置信区间：1.07 至 2.97），但我们没有发现随着既往住院次数的增加死亡率呈梯度增加的证据。此外，一些被记录为既往住院主要诊断的疾病与增加的死亡率无关，这与已知的髋部骨折老年患者生存风险因素相一致。
我们的结果表明，在当地医院信息系统难以获取次要诊断的情况下，使用既往住院次数或与这些住院相关的主要诊断作为髋部骨折老年患者共病谱的替代指标来调整病例组合，可能不是比较医院间院内死亡率的有效方法。
