Early clinical and laboratory risk factors of intensive care unit requirement during 2004-2008 dengue epidemics in Singapore: a matched case-control study
Early clinical and laboratory risk factors of intensive care unit requirement during 2004-2008 dengue epidemics in Singapore: a matched case-control study
复制标题
DOI：
10.1186/s12879-014-0649-2
复制
发表时间：
2014-12-05
影响因子：
3.7
通讯作者：
Lye, David C.
中科院分区：
文献类型：
--
作者：
Pang, Junxiong;Thein, Tun-Linn;Lye, David C.
背景：登革热感染可导致严重的临床表现，需要重症监护。有效的分流对于早期临床治疗以减少发病率和死亡率至关重要。然而，对重症监护病房(ICU)需求的早期危险因素了解有限。本研究旨在确定首次入院时和ICU要求前24小时的早期临床和实验室危险因素。方法：采用1：4配对病例对照研究方法，对2004-2008年间27例需要ICU的登革热患者和108例不需要ICU的登革热患者进行病例对照研究，按登革热出现的年份匹配。进行单因素和多因素条件Logistic回归分析。结果：与非ICU登革热患者相比，ICU登革热患者的年龄显著增加(P=0.003)，并伴有糖尿病(P=0.031)。ICU患者中有7人在发烧后7天的中位数死亡。最初介绍时，世卫组织2009年登革热严重程度分类与重症监护病房显著相关(P<；0.001)，但与世卫组织1997年分类无关。与ICU要求相关的早期临床危险因素有：红细胞压积改变<20%合并血小板<50 K[95%可信区间=2.46~30.53]、低蛋白血症(95%CI=1.09~19.74)、低血压(95%CI=1.83~31.79)和严重器官损害(95%CI=3.30~331)。早期实验室危险因素包括中性粒细胞比例(95%CI=1.0 4~1.17)、血清尿素(95%CI=1.0 2~1.5 6)和丙氨酸氨基转移酶水平(95%CI=1.001~1.0 6)。该预测模型的灵敏度和特异度高达88%。ICU前2 4h的早期实验室危险因素为淋巴细胞(95%CI=1.0 3~1.3 8)、单核细胞比例(95%CI=1.0 2~1.78)、脉率(95%CI=1.002~1.14)和血压(95%CI=0.92~0.996)。该预测模型的灵敏度和特异度分别高达88.9%和78%。结论：据我们所知，这是第一个匹配的病例对照研究，确定了住院期间ICU需求的早期临床和实验室危险因素。这些因素提示登革热患者在ICU要求之前就有不同的病理生理背景，这可能反映了登革热严重程度的发病机制。在较大规模的独立研究中进行验证后，这些风险模型可能有助于临床医生对患者进行分类。
Background: Dengue infection can result in severe clinical manifestations requiring intensive care. Effective triage is critical for early clinical management to reduce morbidity and mortality. However, there is limited knowledge on early risk factors of intensive care unit (ICU) requirement. This study aims to identify early clinical and laboratory risk factors of ICU requirement at first presentation in hospital and 24 hours prior to ICU requirement.Method: A retrospective 1:4 matched case-control study was performed with 27 dengue patients who required ICU, and 108 dengue patients who did not require ICU from year 2004-2008, matched by year of dengue presentation. Univariate and multivariate conditional logistic regression were performed. Optimal predictive models were generated with statistically significant risk factors identified using stepwise forward and backward elimination method.Results: ICU dengue patients were significantly older (P = 0.003) and had diabetes (P = 0.031), compared with non-ICU dengue patients. There were seven deaths among ICU patients at median seven days post fever. At first presentation, the WHO 2009 classification of dengue severity was significantly associated (P < 0.001) with ICU, but not the WHO 1997 classification. Early clinical risk factors at presentation associated with ICU requirement were hematocrit change >= 20% concurrent with platelet < 50 K [95% confidence-interval (CI) = 2.46-30.53], hypoproteinemia (95% CI = 1.09-19.74), hypotension (95% CI = 1.83-31.79) and severe organ involvement (95% CI = 3.30-331). Early laboratory risk factors at presentation were neutrophil proportion (95% CI = 1.04-1.17), serum urea (95% CI = 1.02-1.56) and alanine aminotransferase level (95% CI = 1.001-1.06). This predictive model has sensitivity and specificity up to 88%. Early laboratory risk factors at 24 hours prior to ICU were lymphocyte (95% CI = 1.03-1.38) and monocyte proportions (95% CI = 1.02-1.78), pulse rate (95% CI = 1.002-1.14) and blood pressure (95% CI = 0.92-0.996). This predictive model has sensitivity and specificity up to 88.9% and 78%, respectively.Conclusions: This is the first matched case-control study, to our best knowledge, that identified early clinical and laboratory risk factors of ICU requirement during hospitalization. These factors suggested differential pathophysiological background of dengue patients as early as first presentation prior to ICU requirement, which may reflect the pathogenesis of dengue severity. These risk models may facilitate clinicians in triage of patients, after validating in larger independent studies.