现在的位置: 首页研究点评, 进展交流>正文
[JAMA作者回复]:急诊科使用qSOFA评分
2017年05月22日 研究点评, 进展交流 暂无评论

Comment & Response

May 9, 2017

Use of the qSOFA Score in the Emergency Department—Reply

Yonathan Freund, Ben Bloom, Bruno Riou

JAMA. 2017;317(18):1910. doi:10.1001/jama.2017.3511

In Reply

Dr Scheer and colleagues have 3 concerns about our study: questionable definitions, inconsistent methods, and a data set that includes missing values. The definition of “severe sepsis” that we used may not capture all patients in this category, particularly those with organ dysfunction and normal lactate levels. However, the definition of “organ dysfunction” in severe sepsis is equivocal.1 If we defined severe sepsis as any SOFA component of at least two, 127 more patients would have been classified as having severe sepsis, giving an area under the receiver operating curve (AUROC) for prediction of in-hospital mortality of 0.73. qSOFA still had better prognostic accuracy (incremental AUC, 0.07; 95% CI, 0.02-0.12). Scheer and colleagues suggest these 2 classifications performed equally because there was little difference in their positive and negative predictive values. Because our study was not powered to detect such differences, we do not believe this is a valid conclusion. Furthermore, the requirement for 2 elements of SIRS in severe sepsis resulted in poor sensitivity (47%), ie, in a substantial proportion of seriously ill patients misclassified as not having severe sepsis. This risk of misclassification has previously been reported.2 For these reasons, among others, the Sepsis-3 task force removed the SIRS criteria and focused only on organ dysfunction.3 Our results support these changes. What Scheer and colleagues suggest for severe sepsis is actually what was done with the new definition of sepsis that focuses on organ dysfunction, without the SIRS criteria.

Scheer医生及其同事对我们的研究提出了3点疑虑:有问题的定义,相互矛盾的方法学,以及包括大量缺失数据的数据集。我们所采用的“严重全身性感染”定义并未包括所有患者,尤其是存在器官功能障碍但乳酸水平正常的患者。然而,严重全身性感染患者“器官功能障碍”的定义尚不确定。如果我们将严重全身性感染定义为任何器官的SOFA评分至少为2分,那么将有另外127名患者被纳入严重全身性感染组,预测住院病死率的AUROC为0.73。qSOFA的预后准确性仍然更好(AUC增加0.07;95%CI 0.02-0.12)。Scheer及其同事认为,这两种评分的预测准确性相似,因为其阳性及阴性预期值仅有很小的差异。因为我们的研究不足以发现这一差异,我们不认为这是正确的结论。另外,严重全身性感染要求满足SIRS的2项指标,从而导致敏感性降低(47%),即,相当一部分重症感染患者被错误的分至非严重全身性感染组。错误分组的风险已经见诸报道。基于这些原因,Sepsis-3工作组将SIRS标准排除,而仅仅强调器官功能障碍。我们的结果支持上述改变。Scheer及其同事的建议实际上指出,全身性感染新定义强调器官功能障碍,而忽略SIRS标准。

The primary objective of our study was to assess the prognosis accuracy of qSOFA in patients in the emergency department with suspected infection. Because qSOFA only has 3 components, we decided that we could not analyze patients with any missing components. Also, it seemed artificial to perform multiple imputation. One component is binary (altered mental state) and respiratory rate is highly skewed, so multiple imputation would be subject to bias.4 We accepted missing SOFA data because this is a pragmatic approach and reflects how patients would be managed in an emergency department. We do not agree that this represents an inconsistency in our methods.

我们研究的主要目的在于对可疑感染的急诊科患者评价qSOFA评分的预后准确性。由于qSOFA评分仅有3项指标,我们认为无法分析那些指标缺失的患者。而且,进行多重替代法处理缺失数据也存在人为因素。qSOFA中一项指标为二分变量(意识状态改变),呼吸频率呈严重的偏态分布,因此,多重替代法存在偏倚。我们接受SOFA数据存在缺失的事实,因为这是实效性做法,反映了急诊科处理患者的现状。我们不同意这会导致方法学相互矛盾的说法。

Missing SOFA data could have led to the normal value assumption being inaccurate. However, we built a multiple imputation model that resulted in a larger proportion of patients with a SOFA score of 2 or higher (324 vs 297) and a nonsignificant improvement of SOFA AUROC (0.80 vs 0.77). We did not conclude that qSOFA was superior to SOFA, but that both scores performed well for the prediction of in-hospital mortality. The advantage of qSOFA over SOFA in the emergency department lies in its rapidity, simplicity, and absence of requirement of laboratory variables.

SOFA数据缺失可能导致患者指标正常的假设并不准确。然而,我们建立了多重替代模型,发现更多患者SOFA评分≥2分,但SOFA AUROC并未显著改善(0.80 vs 0.77)。我们并未得出qSOFA由于SOFA的结论,而是认为两种评分对于住院病死率的预测同样准确。在急诊科,qSOFA相较于SOFA的优点在于快捷,简单,无需实验室数据。

References

1. Bone  RC, Balk  RA, Cerra  FB,  et al.  Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis.  Chest. 1992;101(6):1644-1655.PubMedArticle

2. Kaukonen  K-M, Bailey  M, Pilcher  D, Cooper  DJ, Bellomo  R.  Systemic inflammatory response syndrome criteria in defining severe sepsis.  N Engl J Med. 2015;372(17):1629-1638.PubMedArticle

3. Singer  M, Deutschman  CS, Seymour  CW,  et al.  The third international consensus definitions for sepsis and septic shock (Sepsis-3).  JAMA. 2016;315(8):801-810.PubMedArticle

4. Sterne  JAC, White  IR, Carlin  JB,  et al.  Multiple imputation for missing data in epidemiological and clinical research: potential and pitfalls.  BMJ. 2009;338:b2393.PubMedArticle

给我留言

您必须 [ 登录 ] 才能发表留言!

×
腾讯微博