Pseudomonas aeruginosa is responsible for an increasing proportion of infections acquired in the modern hospital setting.1 Of all the infections caused by this organism, bacteremia is one of the most severe. According to data collected between 1990 and 1996 by the Centers for Disease Control and Prevention National Nosocomial Infections Surveillance System, P. aeruginosa ranks seventh in the United States in terms of the frequency of isolation from the bloodstream.2 Therefore, identification of mortality risk factors and definitive, optimized, empirical treatment options have been the focus of study for the past few years.3-8
Despite the recent antibiotic advances, P. aeruginosa continues to be associated with considerable mortality, with this rate ranging from 18% to 61%.3-9 One study has shown that bacteremia caused by P. aeruginosa incurs an even greater risk of hospital mortality than Staphylococcus aureus-related analogs.10 Given the unfavorable outcome for this form of bacteremia, the effect of timely and adequate antipseudomonal therapy in terms of the associated mortality and relationship with factors influencing prognosis is of profound concern. Hence, the principal aims of this retrospective observation study were (1) to determine which factors are important predictors of mortality in P. aeruginosa bacteremia and (2) to evaluate the influence of appropriate antimicrobial therapy on patient outcome.
MATERIALS AND METHODS
This study was conducted at Hsin-Chu Hospital, a district teaching facility with 679 beds and 38 intensive care units located in northern Taiwan. All cases involving P. aeruginosa isolation from blood cultures during the period July 1, 2002, to June 30, 2004, were enrolled in this analysis. The following information was obtained from a review of the medical records: age, sex, underlying disease type, comorbid condition, source of the bacteremia, details of the antibiotic regimen, duration of antibiotic use, and patient outcome, as well as the dates on admission and bacteremia diagnosis, institution of appropriate antibiotics, and eventual discharge. Only the first bacteremic episode for each patient was included in the analysis.
Any concomitant bacteremia other than that caused by P. aeruginosa was regarded as polymicrobial. Nosocomial bloodstream infections were diagnosed according to the relevant Centers for Disease Control and Prevention definition (1988).11 Appropriate antibiotic treatment was defined where the isolate was susceptible to the agent in vitro, with adequate dosage prescribed. The source of the bacteremia was decided on the basis of clinical judgment or by growth of the same organism from blood and another kind of specimen. No patient ever received combination therapy with antipseudomonal β-lactam and a quinolone for bacteremia. Therefore, combination therapy was defined where fluoroquinolone, anti-Pseudomonas cephalosporin, imipenem, or ureidopenicillin was used in combination with an aminoglycoside, as determined by isolate susceptibility to both. Delayed treatment was defined as institution of appropriate treatment more than 72 hours after the blood culture samples were obtained or where there was complete absence of appropriate treatment. The mortality rate was determined from deaths occurring within 30 days of positive blood culture.
All statistical analyses were performed using SPSS 8.0 software (SPSS Inc, Chicago, Ill), with χ2 test (likelihood ratio method) or 2-tailed Fisher exact test used for contingency table analysis, as appropriate. The t test or Mann-Whitney U test was used for continuous or interval variable analysis. A P value of less than 0.05 was considered statistically significant for all tests.
Fifty-six cases of P. aeruginosa bacteremia were identified during the study period. All the patients had fever while their blood cultures were drawn. Subject ages ranged from 1 to 92 years (mean, 67.0 ± 17.9 years). Of all the bacteremia episodes, 76.8% (43/56) involved men. Only one patient had no underlying disease. The most common underlying diseases were malignancy (n = 24; 42.9%), followed by diabetes mellitus (n = 12; 21.4%)), existing cerebrovascular accident (n = 11; 19.6%), uremia (n = 3; 5.4%), and liver cirrhosis (n = 1; 1.8%), with 39.3% (n = 22) of the cases also involving other comorbidities. Of 24 patients with malignancy, 19 had solid tumor, and 5 had hematologic malignancy. Polymicrobial bacteremia was acquired in 23% (13/56) of the cases, with 55% (31/56) of the infections nosocomial. The mean latency of bacteremia occurrence was 14.1 ± 21.5 days (range, 0-137 days) after admission.
Twenty subjects (35.7%) died within 30 days of bacteremia occurrence, 35 survived, and one was lost to follow-up. Of the 20 fatalities, 80% (16/20) died within the first 1 week. The median duration from the occurrence of bacteremia to the day of death was 4.40 ± 4.52 days (range, 0-16 days). The risk factors predicting 30-day mortality from P. aeruginosa bacteremia are presented in Table 1. Relative to the nonsurvivors, a higher proportion of the cured patients had received appropriate antibiotic treatment (odds ratio, 7.25; P = 0.002). No other risk factors were statistically significant comparing mortality rates between the 2 groups. Urinary and respiratory tract infections were the 2 main origins of the bacteremia (Table 2). Of the 6 catheter-related bacteremic episodes, 3 were caused by Port-a-Cath infection, and 2 originated from double-lumen hemodialysis catheters.
No patients received antipseudomonal prophylaxis before developing bacteremia. A total of 19 patients did not receive appropriate antibiotic treatment, whereas 37 patients were treated. The proportions of the subgroup (n = 37) stratified by type of appropriate antibiotic regimens for bacteremia are as follows: piperacillin, 48.6% (n = 18); ceftazidime, 16.2% (n = 6); cefepime, 16.2% (n = 6); fluoroquinolones, 16.2% (n = 6); and cefperazone, 2.7% (n = 1). A comparison of the 37 treated cases is presented in Table 3. Of these, 29 patients were cured, and 8 died, of which 11 and 3, respectively, received a combination antibiotic therapy. No significant differences were demonstrated comparing the 2 groups in terms of the proportion of combined therapy. Patients who received appropriate antibiotic treatment but subsequently died had 1.4 ± 1.8 days' delay before institution of appropriate antibiotic treatment, typically succumbing in the first week of bacteremia, with antibiotic therapy consequently limited to only 4.1 ± 2.1 days in these cases. In contrast, where a cure was achieved, the duration of the antibiotic treatment was 13.2 ± 7.7 days, whereas the delay for initiation of appropriate antibiotics was 3.3 ± 3.0 days.
P. aeruginosa bacteremia comprises 10%12 and 10% to 25.6% of all and nosocomially acquired gram-negative bacteremias, respectively.4,7,13 The infection tends to occur in hospitalized patients4,5 or immunocompromised hosts,14,15 including those with HIV infection.16,17 Malignancy, especially leukemia, is a well-known concomitant disease. In recent years, the frequency of P. aeruginosa bacteremia has decreased in patients with solid tumors but has remained unchanged in those with acute leukemia.18 Furthermore, 11% of gram-negative bacilli from bloodstream samples of patients with hematologic malignancies are P. aeruginosa.15 In this study, all but one subject had at least one underlying disease. Malignancy, representing 42.9% of all the underlying diseases, was the most common of these. Subgroup analysis showed that solid tumors were seen more frequently because of scarce hematologic patients in Hsin-Chu hospital.
In this report, 23% of the patients acquired polymicrobial bacteremia. In their comparison of polymicrobial and monomicrobial bacteremias involving P. aeruginosa, Aliaga et al19 concluded that polymicrobial variants had a higher crude mortality rate. Our results were not comparable, however, with urinary and respiratory tracts, the 2 main bacteremia origins in this study. Furthermore, the origin did not affect the mortality rate for P. aeruginosa bacteremia in our sample population.
Our 30-day mortality rate of 35.7% is comparable to analogous results from other investigations.3-9 It has been demonstrated that central nervous system involvement,3 previous thromboembolism,3 fatal underlying disease,3,7 surgery,7 pneumonia,5-7 shock,4,5,7 low granulocyte count,4 inappropriate antibiotic therapy,4,5 development of septic metastasis,4 and high APACHE II score5 are poor prognostic factors in cases of P. aeruginosa bacteremia. In the present study, however, only appropriate antimicrobial therapy was a key determinant of clinical outcome. Vidal et al7 arrived at a similar conclusion after exclusion of their subset of patients with intravenous catheter-associated bacteremia.
Because the standard of practice for hospitalized patients with new-onset unknown fever in Hsin-Chu hospital includes drawing 2 sets of blood cultures promptly and instituting empirical antibiotics, there was no significant difference between survivor and nonsurvivor groups about the time from symptom onset to diagnosing P. aeruginosa bacteremia. After P. aeruginosa bacteremias were diagnosed, 80% (16/20) of the fatalities died in the first 1 week. The median duration from occurrence of bacteremia to death was 4.40 ± 4.52 days. Given the rapid onset and devastating nature of the often-fatal disease, emphasis must be placed on the importance of early detection of P. aeruginosa bacteremia and timely institution of appropriate antibiotic therapy. Gransden et al20 identified 7 independent predictive factors for P. aeruginosa bacteremia onset in their 20-year study. Kang et al5 also emphasized the importance of timely administration of effective antimicrobial therapy. Given the findings of the above studies and our own investigation, therefore, it appears prudent to advise and urge prompt prescription of antipseudomonal antibiotics when confronting possible P. aeruginosa bacteremia because of its rapidly deteriorating clinical course.
The issue of combined therapy for P. aeruginosa bacteremia has long been a topic of debate6-8,21; however, no differences in mortality rate were demonstrated between compared with monotherapy in this study. Micek et al22 have proposed that the risk of inadequate antimicrobial treatment for P. aeruginosa bloodstream infection may be minimized through increased use of empirical combinations of antimicrobial treatment until susceptibility results are known. The specific choice of these agents should be based on local susceptibility patterns. In this study, all classes of antipseudomonal antibiotics, where P. aeruginosa susceptibility had been demonstrated in vitro, were regarded as appropriate, except aminoglycoside monotherapy.3 Paradoxically, commencement of appropriate antibiotic treatment was more delayed in the cured group (Table 3). A reasonable assumption for this apparent anomaly is that, as diseases among the fatalities were more severe, physicians were more likely to prescribe antipseudomonal therapy earlier.
This investigation has 2 main limitations. First, clinical data used to conclusively define disease severity for all the patients were inadequate because of the retrospective nature of the study. Second, as the sample was drawn from a regional hospital, the number of enrolled cases was quite modest in terms of the general applicability of the results. Our findings were in line with those of other studies, however, and we confirm that early institution of antipseudomonal antibiotics is vital to clinical outcome when encountering possible P. aeruginosa bacteremia.
1. Morrison AJ Jr, Wenzel RP. Epidemiology of infections due to Pseudomonas aeruginosa. Rev Infect Dis. 1984;6:S627-S642.
2. Centers for Disease Control and Prevention National Nosocomial Infections Surveillance (NNIS) report, October 1986 to April 1996. Am J Infect Control. 1996;24:380-388.
3. Kuikka A, Valtonen VV. Factors associated with improved outcome of Pseudomonas aeruginosa bacteremia in a Finnish university hospital. Eur J Clin Microbiol Infect Dis. 1998;17:701-708.
4. Bisbe J, Gatell JM, Puig J, et al. Pseudomonas aeruginosa bacteremia: univariate and multivariate analyses of factors influencing the prognosis in 133 episodes. Rev Infect Dis. 1988;10:629-635.
5. Kang CI, Kim SH, Kim HB, et al. Pseudomonas aeruginosa bacteremia: risk factors for mortality and influence of delayed receipt of effective antimicrobial therapy on clinical outcome. Clin Infect Dis. 2003;37:745-751.
6. Hilf M, Yu VL, Sharp J, et al. Antibiotic therapy for Pseudomonas aeruginosa bacteremia: outcome correlations in a prospective study of 200 patients. Am J Med. 1989;87:540-546.
7. Vidal F, Mensa J, Almela M, et al. Epidemiology and outcome of Pseudomonas aeruginosa bacteremia, with special emphasis on the influence of antibiotic treatment. Analysis of 189 episodes. Arch Intern Med. 1996;156:2121-2126.
8. Siegman-Igra Y, Ravona R, Primerman H, et al. Pseudomonas aeruginosa bacteremia: an analysis of 123 episodes, with particular emphasis on the effect of antibiotic therapy. Int J Infect Dis. 1998;2:211-215.
9. Gallagher PG, Watanakunakom C. Pseudomonas bacteremia in a community teaching hospital, 1980-1984. Rev Infect Dis. 1989;11:846-852.
10. Osmon S, Ward S, Fraser VJ, et al. Hospital mortality for patients with bacteremia due to Staphylococcus aureus or Pseudomonas aeruginosa. Chest. 2004;125:607-616.
11. Garner JS, Jarvis WR, Emori TG, et al. CDC definitions for nosocomial infections, 1988. Am J Infect Control. 1988;16:128-140.
12. Cermak P, Kolar M, Latal T. Frequency of gram-negative bacterial pathogens in bloodstream infections and their resistance to antibiotics in the Czech Republic. Int J Antimicrob Agents. 2004;23:401-404.
13. Gordon SM, Serkey JM, Keys TF, et al. Secular trends in nosocomial bloodstream infections in a 55-bed cardiothoracic intensive care unit. Ann Thorac Surg. 1998;65:95-100.
14. Grisaru-Soen G, Lerner-Geva L, Keller N, et al. Pseudomonas aeruginosa bacteremia in children: analysis of trends in prevalence, antibiotic resistance and prognostic factors. Pediatr Infect Dis J. 2000;19:959-963.
15. Wang FD, Lin ML, Liu CY. Bacteremia in patients with hematological malignancies. Chemotherapy. 2005;51:147-153.
16. Rongkavilit C, Rodriguez ZM, Gomez-Marin O, et al. Gram-negative bacillary bacteremia in human immunodeficiency virus type 1-infected children. Pediatr Infect Dis J. 2000;19:122-128.
17. Vidal F, Mensa J, Martinez JA, et al. Pseudomonas aeruginosa bacteremia in patients infected with human immunodeficiency virus type 1. Eur J Clin Microbiol Infect Dis. 1999;18:473-477.
18. Chatzinikolaou I, Abi-Said D, Bodey GP, et al. Recent experience with Pseudomonas aeruginosa bacteremia in patients with cancer: retrospective analysis of 245 episodes. Arch Intern Med. 2000;160:501-509.
19. Aliaga L, Mediavilla JD, Llosa J, et al. Clinical significance of polymicrobial versus monomicrobial bacteremia involving Pseudomonas aeruginosa. Eur J Clin Microbiol Infect Dis. 2000;19:871-874.
20. Gransden WR, Leibovici L, Eykyn SJ, et al. Risk factors and a clinical index for diagnosis of Pseudomonas aeruginosa bacteremia. Clin Microbial Infect. 1995;1:119-123.
21. Chamot E, Boffi El Amari E, Rohner P, et al. Effectiveness of combination antimicrobial therapy for Pseudomonas aeruginosa bacteremia. Antimicrob Agents Chemother. 2003;47:2756-2764.
22. Micek ST, Lloyd AE, Ritchie DJ, et al. Pseudomonas aeruginosa bloodstream infections: importance of appropriate initial antimicrobial treatment. Antimicrob Agents Chemother. 2005;49:1306-1311.
© 2006 Lippincott Williams & Wilkins, Inc.