The clinical impact of bacteremia in complicated acute pyelonephritis

• Published in: The American journal of the medical sciences Vol. 332; no. 4; p. 175
• Main Authors: Hsu, Chih-Yang, Fang, Hua-Chang, Chou, Kang-Ju, Chen, Chien-Liang, Lee, Po-Tsang, Chung, Hsiao-Min
• Format: Journal Article
• Language: English
• Published: United States 01.10.2006
• Subjects: Acute Disease; Adult; Aged; Anti-Bacterial Agents - administration & dosage; Bacteremia - blood; Bacteremia - drug therapy; Bacteremia - physiopathology; Case-Control Studies; Colony Count, Microbial - methods; Diabetes Mellitus - microbiology; Female; Humans; Leukocyte Count - methods; Male; Middle Aged; Monitoring, Physiologic - methods; Pyelonephritis - blood; Pyelonephritis - complications; Pyelonephritis - drug therapy; Pyelonephritis - microbiology; Pyelonephritis - physiopathology; Retrospective Studies; Serum Albumin - analysis; Severity of Illness Index; Shock, Septic - blood; Shock, Septic - drug therapy; Shock, Septic - physiopathology; Time Factors; Urinary Tract Infections - complications; Urinary Tract Infections - microbiology
• ISSN: 0002-9629
• DOI: 10.1097/00000441-200610000-00004

Bacteremia has been considered as a surrogate marker of severe infection in several infectious diseases. However, it remains uncertain whether the presence of bacteremia correlates with severe infection in patients with complicated acute pyelonephritis (APN). We performed a retrospective study to investigate the relationship between the presence of bacteremia and disease severity in complicated APN. To do this, we reviewed medical records from 128 patients diagnosed with complicated APN admitted to Kaohsiung Veterans General Hospital, Taiwan between January, 2003 and December, 2003. In our analysis, we compared clinical presentation, treatment response, and outcome in patients with and without bacteremia. Fifty-four of 128 patients (42%) were bacteremic. This group of patients presented more frequently with severe sepsis or septic shock (P < 0.001), compared with nonbacteremic patients. Other factors that correlated with the presence of bacteremia were older age, diabetes mellitus, more band forms in neutrophil cell counts, impaired renal function, and a lower level of serum albumin. Using a multivariate logistic regression analysis, we show that lower levels of serum albumin (odds ratio, 0.18; 95% CI, 0.05-0.65; P = 0.008) and presence of severe sepsis (odds ratio, 4.76; 95% CI, 1.43-15.84; P = 0.011) were independent factors associated with bacteremia. Following treatment, the bacteremic group took a longer time to become defervescent than the nonbacteremic group (5.1 +/- 2.3 vs. 4.2 +/- 1.6 days, P = 0.023). Also, the bacteremic group had a greater mean duration of intravenous antibiotics administration and longer hospital stays (P < 0.001). Multiple logistic regression analysis shows that non-Escherichia coli bacteremia, presence of urolithiasis or hydronephrosis, shorter duration of antibiotics administration, and being male were significantly associated with recurrence of urinary tract infection within 6 months. Bacteremia in cases of complicated APN indicates a severe disease, which is more likely to recur in patients with non-E coli bacteremia. Our study showed that bacteremia is indeed a useful clinical indicator of severe disease and, if found, should influence patient management. Therefore, we recommend that blood culture samples should be taken in all patients with complicated APN.