The PIRO concept: prediction of bacteremia in patients admitted to the intensive care unit with infection

Introduction

Bacteremia in critically ill patients has a significant attributable mortality. Routine use of blood cultures has been questioned because of low yield and association with anemia. Attempts to identify predictors of bacteremia in critically ill patients have had variable success. The PIRO concept based on factors related to predisposition (P), infection (I), response (R), and organ dysfunction (O) has been introduced as a hypothesis-generating model for research in sepsis.

Aim

The aim of this study is to assess the association of PIRO-related variables with the occurrence of bacteremia in patients admitted to the intensive care unit (ICU) with infection.

Methods

This retrospective, cohort study included 191 patients admitted to three ICUs of a tertiary care medical center, with an infection-related diagnosis between 1 July and 31 December 2002. Bacteremia was defined as the isolation of a bacterium in one or more blood cultures. The P-related variables include evidence of immunosuppression, dialysis-dependent renal failure, nursing home residence, and performance status. The I-related variables include site of infection and place of acquisition. The R-related variables include heart rate, respiratory rate and temperature. Sequential Organ failure Assessment (SOFA) scores were calculated to quantify organ dysfunction. We used Student's t test, the Mann–Whitney U test, the chi-square test and Fisher's exact test for comparison between bacteremia and nonbacteremia groups. To define which variables were significantly and independently associated with bacteremia, we performed logistic regression analysis including day 1 SOFA score, temperature > 38°C, heart rate, immunocompromised status and site of infection (urinary 1, other 0). P < 0.05 was considered significant.

Results

The mean (SD) age of the patients was 64 (16) years and 92% were Caucasian. Bacteremia was documented in 44 (23%) patients. In univariate analysis, only the SOFA score was associated with the occurrence of bacteremia. The mean (SD) SOFA score in the bacteremia group was 7.3 (4) vs 5.6 (3) in the nonbacteremia group (P = 0.012). The day 1 SOFA score was also found to be an independent predictor of bacteremia (P = 0.0057) along with day 1 temperature > 38°C (P = 0.016).

Conclusion

The SOFA score, used as a measure of organ dysfunction in the PIRO concept, may have a role in the prediction of bacteremia in patients admitted to the intensive care unit with infection.