Predictive Validity of the qSOFA Score for Sepsis in Adults with Community-Onset Staphylococcal Infection in Thailand.

Thumbnail Image
Change log
Tandhavanant, Sarunporn 
Suntornsut, Pornpan 
Chetchotisakd, Ploenchan 

The quick sequential organ failure assessment (qSOFA) score has had limited validation in lower resource settings and was developed using data from high-income countries. We sought to evaluate the predictive validity of the qSOFA score for sepsis within a low- and middle-income country (LMIC) population with culture-proven staphylococcal infection. This was a secondary analysis of a prospective multicenter cohort in Thailand with culture-positive infection due to Staphylococcus aureus or S. argenteus within 24 h of admission and positive (≥2/4) systemic inflammatory response syndrome (SIRS) criteria. Primary exposure was maximum qSOFA score within 48 h of culture collection and primary outcome was mortality at 28 days. Baseline risk of mortality was determined using a multivariable logistic regression model with age, gender, and co-morbidities significantly associated with the outcome. Predictive validity was assessed by discrimination of mortality using area under the receiver operating characteristic (AUROC) curve compared to a model using baseline risk factors alone. Of 253 patients (mean age 54 years (SD 16)) included in the analysis, 23 (9.1%) died by 28 days after enrollment. Of those who died, 0 (0%) had a qSOFA score of 0, 8 (35%) had a score of 1, and 15 (65%) had a score ≥2. The AUROC of qSOFA plus baseline risk was significantly greater than for the baseline risk model alone (AUROCqSOFA = 0.80 (95% CI, 0.70-0.89), AUROCbaseline = 0.62 (95% CI, 0.49-0.75); p < 0.001). Among adults admitted to four Thai hospitals with community-onset coagulase-positive staphylococcal infection and SIRS, the qSOFA score had good predictive validity for sepsis.

Staphylococcus, Thailand, sepsis, sequential organ failure assessment scores, systemic inflammatory response syndrome
Journal Title
J Clin Med
Conference Name
Journal ISSN
Volume Title