- Poster presentation
- Open Access
Institutional evaluation of a new methodology for early sepsis risk identification in hospitalized patients
Critical Carevolume 11, Article number: P22 (2007)
The effectiveness of sepsis, severe sepsis and septic shock management on prognosis depends strongly on early clinical suspicion and rigorous diagnosis methods. Early clinical suggestive infection sign recognition is therefore also a cornerstone of successful treatment.
To evaluate a new institutional methodology for early sepsis risk identification in hospitalized patients.
A before – after study design with prospective consecutive data collection in a 124-bed private medical center. Twelve months after the institutional Surviving Sepsis Campaign implementation and current use of the respective treatment bundles, this medical center adopted a standardized hospital maneuver to anticipate the identification of two or more suggestive infection signs. Demographic data, the time interval for recognition of two or more infection risk signs, and the mortality rate are evaluated during the next 5 months (phase II) and compared with the same data obtained during the initial 12 months (phase I).
A total of 85 patients with two or more suggestive infection signs were enrolled. Thirty-two patients were recognized with severe sepsis during phase I and 22 patients in phase II. Demographic variables and severity of illness measured by the APACHE II score (P = 0.12) were similar for both groups. The phase I severe sepsis patients were identified after 29 ± 32 hours from the initial presentation of two or more infections signs. On the other hand, during phase II this time was lower: 14.5 ± 16 hours (P < 0.07). The hospital mortality was greater in the phase I group (50%) when compared with the phase II group (27.3%) (P < 0.08).
These preliminary data suggest that the implementation of the proposed methodology for early sepsis risk identification in hospitalized patients was associated with early severe sepsis recognition and reduced mortality.