Pseudomonas aeruginosa remains an important cause of BSI and one of the principal pathogens responsible for severe organ dysfunction. The present study compares the three most useful severity of illness scoring systems applied in studies of bloodstream infection. We decided to compare APACHE II, SAPS II and SOFA on the day of bacteremia and two days prior to positive blood culture through two days afterwards in order to determine which scoring system was most predictive for mortality.
The ability to assess a complex clinical condition such as bloodstream infection, using relatively simple scores may facilitate communication with regards to the severity of the physiologic process. Our intention is to improve the assessment of P. aeruginosa
BSI so as to identify different patterns of organ dysfunction, and thereby enhance our understanding of the infectious process, as well as gaining knowledge of a patient's prognosis. BSIs should not be seen as static phenomena, but rather as a continuum of alterations with changes in the patients' condition seen daily. Many intrinsic variables are involved (e.g., underlying diseases, age, and gender), but interventional variables such as adequate initial antibiotic treatment [13
], rapid removal of catheters responsible for perpetuating the infection [14
], and optimization of hemodynamic status with fluid resuscitation or vasopressors [15
] may determine the evolution of this process. If we had calculated the prognostic scores on our patients at the time of hospital admission, given that one half of the patients acquired BSI after 21 days of hospitalization, the prediction of mortality would likely be totally different and would not take into account the impact of the nosocomial bloodstream infection.
Many intensive care units receive trauma and surgical patients and they are able to stratify acutely ill patients prognostically with success by using APACHE II [4
], SAPS II [6
] or SOFA scores [7
]. However the proportion of infected patients in these studies of scoring systems was less than 30%. To our knowledge, until this point there has been no study that compares the severity of illness scoring systems for patients with nosocomial bloodstream infection. In patients with P. aeruginosa
BSI, calculation of the area under the ROC curve in our study confirmed that assessing scores at day -1 was best for SAPS II and APACHE II, while SOFA was best assessed on day +1 of BSI. Multivariate analysis was performed to control for underlying disease using the Charlson weighted comorbidity index. This confirmed that SAPS II and APACHE II on day -1 were the best predictors for mortality (Table ).
In the APACHE II and SAPS II studies, calculations were done only on the day of admission [4
]. A previous report showed that the APACHE II score at admission was not useful as a prognostic factor, whereas the progression of organ dysfunction after the onset of pneumonia due to P. aeruginosa
in intubated patients was predictive [5
]. However, other studies showed increased APACHE II scores as a risk factor for mortality [3
]. On the other hand, in the SOFA study the scores were calculated every day [7
]. Other studies comparing scoring systems in cirrhotic patients with renal failure found that the discriminatory power of SOFA for predicting mortality was superior to the other scoring systems [17
The limitations of our study should be acknowledged. First, we performed a retrospective study. Second, we applied the APACHE II and SAPS II scores daily although these scores were originally intended to be calculated after the first 24 hours of ICU care. And lastly, it is a single center study which limits its generalizability.