现在的位置: 首页研究点评, 进展交流>正文
[JAMA述评]:中低收入国家全身性感染患者的qSOFA评分
2018年06月17日 研究点评, 进展交流 暂无评论

Editorial

May 20, 2018

qSOFA Score for Patients With Sepsis in Low- and Middle-Income Countries

Neill K. J. Adhikari, Gordon D. Rubenfeld

JAMA. Published online May 20, 2018. doi:10.1001/jama.2018.6413

In the past 2 years, sepsis has received considerable attention from the global community. In 2017, the World Health Assembly passed a resolution urging all 194 UN Member States to implement actions to reduce the burden of sepsis, and for the World Health Organization to report on the public health implications of sepsis and its global consequences.1 It has become clear that early resuscitation targeted to hemodynamic goals is more expensive and no more effective than routine care in high-income countries.2 In Zambia, a protocolized resuscitation strategy actually increased mortality.3 These data and other studies4continue to suggest that some evidence on sepsis therapy does not generalize on a global scale.

In addition, in 2016, the Sepsis-3 expert panel5 developed a new consensus definition of sepsis as “life-threatening organ dysfunction caused by a dysregulated host response to infection,” with organ dysfunction operationalized by a 2-point increase in the Sequential (Sepsis-Related) Organ Failure Assessment score (SOFA).6 The SOFA score (range, 0-24) has 6 variables that require 2 clinical and 4 laboratory assessments.

In the revamped definition, the decades-old concept of systemic inflammatory response syndrome (SIRS)7was abandoned for several reasons. First, SIRS criteria of deranged temperature, white blood cell count, heart rate, and respiratory rate do not uniformly reflect organ dysfunction and are not concordant with the current conceptual model of sepsis. Second, SIRS criteria lack specificity, with approximately 50% of patients in hospital but not in the intensive care unit having at least 2 criteria at some time, of whom far fewer have infection and organ dysfunction.8 Third, determination of SIRS mandates measurement of the white blood cell count, which impedes feasibility.

Simply based on the distribution of the global population and epidemiologic data from high-income countries, it is estimated that almost 90% of the annual 19 million sepsis cases occur in low- and middle-income countries (LMICs), where the barriers to improving outcomes are numerous.9 These challenges start well before any acute medical care is provided, with major disparities in income, living conditions, sanitation of the built environment, and access to basic public health interventions and ambulatory care. For patients with sepsis who survive to access an often distant hospital, treatment might include counterfeit antimicrobials, limited surgical and obstetrical interventions, and intensive care that might or might not have life support interventions. Amid these daunting circumstances, the prompt identification of patients with sepsis among those that access acute care, to administer timely antibiotics and resuscitation, is at least as great a challenge in resource-limited settings in LMICs as it is for wealthier regions.

The 2016 Sepsis-3 Consensus Conference recognized that using SOFA as part of the sepsis definition outside the intensive care unit would be impractical. Therefore, the group used data from US and German hospitals to derive the quick SOFA (qSOFA) score, which includes 2 vital signs and a brief neurological assessment in a single measure.10 The score consists of 1 point for each of hypotension (systolic blood pressure ≤100 mm Hg), tachypnea (respiratory rate ≥22/min), and altered mental status; a positive score is defined as 2 or 3 points. Considering the domains of reliability, validity, and feasibility when evaluating diagnostic criteria,11 qSOFA is superior to the SIRS criteria regarding content validity (agreement with the concept of organ dysfunction as part of sepsis) and feasibility (no laboratory tests required).

In the initial description, qSOFA had moderate ability to discriminate between survivors and decedents in nonintensive care unit patients with suspected infection (area under the receiver operating characteristic curve [AUC], 0.81) compared with SIRS (AUC, 0.76). A recent systematic review12 of studies conducted primarily in high-income countries found lower discrimination for both qSOFA and SIRS compared with the original description10 (AUC, 0.72 for qSOFA and 0.64 for SIRS for patients outside the intensive care unit, calculated from the reported diagnostic odds ratios13), with qSOFA having lower sensitivity and higher specificity for mortality compared with SIRS.

In this issue of JAMA, Rudd and colleagues14 analyzed 8 cohort studies and 1 randomized clinical trial, including 6569 adult patients with suspected infection from 10 LMICs in Africa, Asia, and the Caribbean. There was substantial clinical heterogeneity in types of infection (any infection, suspected Lassa fever, malaria, dengue), location within the hospital (emergency department, ward, intensive care unit), and hospital mortality (range, 1.3%-22%). In the 9 studies, 83% of the patients (7 studies) were from hospitals with intensive care units that included mechanical ventilation and vasopressors, which is a higher level of resource than available in many low-income countries. The extent of missing variables was problematic: 61% for white blood cell counts, 20% for altered mental status, and 5.3% for mortality, for example, with wide range of missing variables among studies. The authors developed a predictive model for hospital mortality using baseline risk factors of age, sex, HIV status, and transfer status from another center, and then evaluated the discrimination using this model and with addition of qSOFA or SIRS. As a predictor of mortality, qSOFA was overall better than SIRS (AUC, 0.69 and 0.59, respectively); when added to the baseline model (AUC, 0.56; close to a coin toss), the model with qSOFA (AUC, 0.70) was better than with SIRS (AUC, 0.59).

For researchers studying sepsis in low-resource settings in LMICs, the feasibility of qSOFA is an important and useful feature. Although the Sepsis-3 publications make it clear that sepsis is not defined by qSOFA, its performance in the study by Rudd and colleagues14 suggests it might be used that way in resource-limited settings. This approach would be similar to ultrasonography and oxygen saturation as substitutions for chest radiography and arterial blood gas analysis in the Kigali modification of the acute respiratory distress syndrome definition.15 It is particularly reassuring to see the performance of qSOFA validated across the diverse pathogens of HIV, tuberculosis, malaria, and dengue routinely encountered in many LMICs, all of which would have been uncommon or absent in the derivation data sets and whose protean manifestations might have affected qSOFA performance. For local epidemiologic and quality improvement studies, identifying sepsis cases without laboratory testing—if shown to be sufficiently comparable to SOFA-defined sepsis—would facilitate comparative studies of hospital-based incidence and temporal trends, and initiatives to improve outcomes. For clinical trials, enrollment of patients with sepsis defined by a positive qSOFA would be a form of prognostic enrichment, selecting patients at higher risk of a sepsis-related poor outcome.

Even if the available data show that qSOFA risk stratifies patients with suspected infection somewhat better than SIRS, there remains confusion over what, if any, clinical direction it provides. Does a positive qSOFA mandate further investigation, as suggested by Sepsis-3,5 or increased monitoring? Can an infected patient with a negative qSOFA be safely discharged home? For clinicians in LMICs, where clinician time, investigations, and intensive care are all markedly constrained, these questions are more pressing and the answers even less clear.

qSOFA might be used to define sepsis in LMICs where SOFA scoring is unfeasible, but that is definitely not how Sepsis-3 intends it to be used.5 qSOFA might be used, incorrectly, to identify infected patients; however, it is unlikely to outperform SIRS, which includes fever and elevated white blood cell count, or clinical evaluations targeted at specific sources of infection.16 qSOFA might be used as a triage aid like the Pneumonia Severity Index,17 which also risk stratifies infected patients and has been validated as a decision tool; however, Sepsis-3 specifically indicates that qSOFA is not to be used in that fashion.5 qSOFA might simply be a valuable warning score indicating which patients deserve additional testing and monitoring. However, its superiority to other early warning scores has not been demonstrated,18 and there are reasons to favor a more generalizable early warning strategy to improve outcomes of all patients, not just those with sepsis. Particularly in LMICs, such a strategy is likely to include more than a warning and may involve protocols to improve vital signs measurement and documentation, with algorithms for clinicians to initiate emergency treatment.19

Ultimately, the most relevant questions about qSOFA are whether it can be used to improve patient outcomes in resource-limited settings by providing a fast and feasible definition of sepsis, or in wealthier hospitals by ensuring that the next step of clinical and laboratory assessment of organ failure is rapidly done. On these questions, the jury is still out; answers will require further data from rigorously conducted studies.

给我留言

您必须 [ 登录 ] 才能发表留言!

×
腾讯微博