[EngO4-5] Mortality prediction among sepsis patients using a combination of qSOFA, National Early Warning Score, age, gender and serum lactate levels
INTRODUCTION:
The definition of sepsis was revised in 2016 for early recognition of sepsis. The definition captures a wide variation, from patients who require only intravenous antibiotics to critically-ill patients who require intensive care such as intubation and continuous renal replacement therapy which are associated with high mortality.
OBJECTIVE:
To predict critically-ill sepsis patients using the quick Sequential Organ Failure Score (qSOFA), the National Early Warning Score (NEWS) and some variables measured easily on admission.
METHODS:
This was a single-center retrospective cohort study. Of all patients who admitted to our emergency room and intensive care unit, we reviewed 77 patients who were diagnosed as sepsis between March 2014 and December 2017. Age, gender, vital signs (heart rate, body temperature, and respiratory rate), consciousness level, and serum lactate level were measured on admission. Primary outcome measure was in-hospital mortality. We calculated the qSOFA and the NEWS with the variables measured on admission. Then, we created a prediction model using qSOFA or NEWS, age, gender and serum lactate level and performed receiver operating characteristic (ROC) analysis with estimation of the corresponding areas under the curve (AUC).
RESULTS:
Among the eligible 77 patients, the median age was 74 years old (interquartile range [IQR]: 63-82) and 53 (68.8%) was male. Median lactate level was 4.7mmol/l (IQR: 2.2-7.9). About the NEWS on arrival at hospital, low clinical risk patients, middle clinical risk patients and high clinical risk patients were 7 (9.1%), 9(10.4%) and 62 (80.5%), respectively. About the qSOFA on arrival, the score 0, 1, 2 and 3 were 3(3.9%), 21(27.3%), 34(44.2%) and 19(24.7%), respectively. The all-cause mortality rate was 45.5%. Continuous hemodiafiltration was performed for 19 patients (24.7%). ROC curves showed that NEWS had an AUC 0.44 (95% confidential interval [CI], 0.31-0.57); however, the combination of NEWS, age over 65 years, gender and serum lactate level yielded an AUC 0.76 (95% CI, 0.65-0.87). Similarly, qSOFA had an AUC 0.47 (95% confidential interval [CI], 0.34-0.60) and the combination of NEWS, age over 65 years, gender and serum lactate level yielded an AUC 0.78 (95% CI, 0.68-0.88).
CONCLUSION: Among sepsis patients, combined use of NEWS or qSOFA, age, gender and serum lactate levels may be better predictor to detect critically-ill patients who were associated with high mortality.
The definition of sepsis was revised in 2016 for early recognition of sepsis. The definition captures a wide variation, from patients who require only intravenous antibiotics to critically-ill patients who require intensive care such as intubation and continuous renal replacement therapy which are associated with high mortality.
OBJECTIVE:
To predict critically-ill sepsis patients using the quick Sequential Organ Failure Score (qSOFA), the National Early Warning Score (NEWS) and some variables measured easily on admission.
METHODS:
This was a single-center retrospective cohort study. Of all patients who admitted to our emergency room and intensive care unit, we reviewed 77 patients who were diagnosed as sepsis between March 2014 and December 2017. Age, gender, vital signs (heart rate, body temperature, and respiratory rate), consciousness level, and serum lactate level were measured on admission. Primary outcome measure was in-hospital mortality. We calculated the qSOFA and the NEWS with the variables measured on admission. Then, we created a prediction model using qSOFA or NEWS, age, gender and serum lactate level and performed receiver operating characteristic (ROC) analysis with estimation of the corresponding areas under the curve (AUC).
RESULTS:
Among the eligible 77 patients, the median age was 74 years old (interquartile range [IQR]: 63-82) and 53 (68.8%) was male. Median lactate level was 4.7mmol/l (IQR: 2.2-7.9). About the NEWS on arrival at hospital, low clinical risk patients, middle clinical risk patients and high clinical risk patients were 7 (9.1%), 9(10.4%) and 62 (80.5%), respectively. About the qSOFA on arrival, the score 0, 1, 2 and 3 were 3(3.9%), 21(27.3%), 34(44.2%) and 19(24.7%), respectively. The all-cause mortality rate was 45.5%. Continuous hemodiafiltration was performed for 19 patients (24.7%). ROC curves showed that NEWS had an AUC 0.44 (95% confidential interval [CI], 0.31-0.57); however, the combination of NEWS, age over 65 years, gender and serum lactate level yielded an AUC 0.76 (95% CI, 0.65-0.87). Similarly, qSOFA had an AUC 0.47 (95% confidential interval [CI], 0.34-0.60) and the combination of NEWS, age over 65 years, gender and serum lactate level yielded an AUC 0.78 (95% CI, 0.68-0.88).
CONCLUSION: Among sepsis patients, combined use of NEWS or qSOFA, age, gender and serum lactate levels may be better predictor to detect critically-ill patients who were associated with high mortality.