Serial SIRS Scores in Distinct Phenotypes of Presumed Sepsis in Culture Positive & Negative Patients.

Determine the utility of using scoring systems in a serial fashion to identify distinct phenotypes of sepsis. Sepsis is a heterogenous clinical syndrome with distinct phenotypes and no universally accepted gold standard for diagnosis. Current scoring systems are usually used once during a presentation. Single-center, retrospective study that screened all adults who activated the hospital's SIRS (systemic inflammatory response syndrome) based sepsis alert into culture positive (Cx+) & culture negative (Cx-) groups. Pre-defined location-based subgroups (emergency department–ED, Step down–SDU & floor) were also analyzed. SIRS scores were calculated at t=0 (time of activating > 2 SIRS criteria) as well as 3,6,12 & 24 hours before and after t=0. Primary outcome was a difference in SIRS scores amongst those that were Cx+ or Cx- at any time point. Secondary outcomes were comparing SIRS scores in subgroups over consecutive time points. 7,955 patients met inclusion criteria, including 4,701 Cx+ and 3254 Cx- patients. Differences in both length of stay (LOS) (Cx+ 16.9 days vs Cx- 9.7 days) and in-hospital mortality (IHM) (Cx+ 14.1% vs Cx- 10.5%) were statistically significant (p<.05). LOS and IHM for each subgroup are shown below [table1]. Statistically significant differences were seen in overall SIRS score between Cx + and Cx- groups at hours 6 (Cx+ 1.40+1.04 vs Cx- 1.35+1.01) & 12 (Cx+ 0.95+0.95 vs Cx- 0.90+0.90) after t=0. For patients who met SIRS criteria in the ED the difference between culture positive and negative sepsis was only significant at 24 hours after t=0 (Cx+ 1.10+0.98 vs Cx- 1.00+0.88). For patients on the floor the only significant difference was at t=0 (Cx+ 1.99+1.02 vs Cx- 1.88+0.95). Lastly, the only significant difference for patients in the SDU was at 3 hours after t=0 (Cx+ 1.10+0.98 vs Cx- 1.00+0.88). SIRS scores at all time points for all subgroups and all patients can be found below [table2]. Using readily available clinical scoring systems such as SIRS in a serial fashion over time shows a clear increase in SIRS score (window of bioreactivity) before t=0 and a subsequent decrease in scores after t=0. While not all locational subgroups showed these differences, these findings suggest potential utilities in early detection, intervention, and a real time provision for monitoring response to therapeutics. Further studies need to address whether the lack of differences in subgroups is an issue with power versus too much clinical heterogeneity. [ABSTRACT FROM AUTHOR]