Impact of Time to Report Positive Hemoculture on Mortality in Surgical Critically Ill Patients with Septicemia
Petch Wacharasint MD*1, Chatdanai Angsusakun MD*2, Preecha Jongstapongpun MD*2
Affiliation : *1 Division of Pulmonary and Critical Care Medicine, Department of Medicine, Phramongkutklao Hospital, Bangkok, Thailand *2 Department of Anesthesiology, Phramongkutklao Hospital, Bangkok, Thailand
Objective : In patients receiving inappropriate empirical antimicrobial agents, delaying of time to report positive hemoculture
(TRH) may cause delay switching from inappropriate antimicrobial agents to appropriate antimicrobial agents, and thus may
increase mortality. We hypothesized that the patients with septicemia from different types of pathogens may have differ
duration of TRH, and duration of TRH may had an impact on hospital mortality.
Material and Method: We performed observational study on the patients who were reported to have bacteremia or fungemia
and admitted to the surgical intensive care unit, Phramongkutklao Hospital during a 2-year period. Type of pathogens grew
in blood cultures and the sensitivities to antimicrobial agents were collected from blood culture reports. Patients were
categorized into three groups based on their blood culture reports which were gram-positive (GP) bacteria, gram-negative
(GN) bacteria, and fungus. Primary outcome was duration of TRH and secondary outcome was hospital mortality among the
three groups of patients.
Results : There were 9, 32, and 7 patients for whom growth of GP bacteria (18.8%), GN bacteria (66.7%), and fungus
(14.6%) were reported in their blood cultures respectively. Patients with fungemia had the longest TRH (130 hours, interquartile
range (IQR) 100-137 hours), followed by patients with GN (64 hours, IQR 48-78 hours), and GP bacteremia (55 hours, IQR
42-71 hours) (p = 0.001). There was no difference in hospital mortality (GP 89%, GN 66%, fungus 71%, p = 0.4). TRH was
found significantly longer in survivors (n = 14) (81 hours, IQR 56-105 hours) than non-survivors (n = 34) (64 hours, IQR
48-71 hours) (p = 0.035). In multivariate analysis, we found that every 1-hr increasing of TRH was associated with lower risk
of hospital mortality with adjusted odd ratio of 0.95 (0.92-0.99, p = 0.01).
Conclusion : Patients with fungemia had significantly longer TRH than patients with GP and GN bacteremia. TRH was found
significantly longer in patients with septicemia who survived than non-survivors.
Keywords : Hemoculture, Mortality, Septicemia, Bacteremia, Fungemia
