Bacteremic sepsis leads to higher mortality when adjusting for confounders with propensity score matching

One can falsely assume that it is well known that bacteremia is associated with higher mortality in sepsis. Only a handful of studies specifically focus on the comparison of culture-negative and culture-positive sepsis with different conclusions depending on study design. The aim of this study was to describe outcome for critically ill patients with either culture-positive or -negative sepsis in a clinical review. We also aimed to identify subphenotypes of sepsis with culture status included as candidate clinical variables. Out of 784 patients treated in intensive care with a sepsis diagnosis, blood cultures were missing in 140 excluded patients and 95 excluded patients did not fulfill a sepsis diagnosis. Of 549 included patients, 295 (54%) had bacteremia, 90 (16%) were non-bacteremic but with relevant pathogens detected and in 164 (30%) no relevant pathogen was detected. After adjusting for confounders, 90-day mortality was higher in bacteremic patients, 47%, than in non-bacteremic patients, 36%, p = 0.04. We identified 8 subphenotypes, with different mortality rates, where pathogen detection in microbial samples were important for subphenotype distinction and outcome. In conclusion, bacteremic patients had higher mortality than their non-bacteremic counter-parts and bacteremia is more common in sepsis when studied in a clinical review. For reducing population heterogeneity and improve the outcome of trials and treatment for sepsis, distinction of subphenotypes might be useful and pathogen detection an important factor.