Purpose: The purpose of this study was to determine the recent incidence of infection and to evaluate antimicrobial usage among adult leukemic patients undergoing chemotherapy for acute myelogenous leukemia (AML) and acute lymphoblastic leukemia (ALL) at Stanford University Hospital.
Patients and methods: The records of 142 adult patients from a consecutive series of 226 induction or consolidation/maintenance chemotherapy courses for AML or ALL between 1982 to 1986 were reviewed retrospectively. Data were analyzed to compare the infectious disease complications and antimicrobial usage for patients receiving identical chemotherapy for a specific phase of leukemia treatment. Evaluation for each chemotherapy course included assessments for the following: compliance with criteria for initiating antibiotics, incidence of infection that was documented by culture or clinical criteria, predictive value of surveillance cultures, incidence of superinfection, survival outcomes, antimicrobial usage, antibiotic-related adverse effects, and cost for antibiotics and diagnostic studies.
Results: Antimicrobials were employed in 190 (84%) of 226 chemotherapy courses. Broad-spectrum antibiotics were regularly begun within the first five days of admission and they were continued for an average of 3.5 weeks until the granulocyte count was greater than 1,000/microL after discontinuation of chemotherapy. There were no differences in the types of infection or outcomes among the patient groups. There was only a 37% rate of documented infections by culture or clinical signs among these patients during their entire hospital stay. Bacterial infections, especially those caused by coagulase-negative staphylococci in patients with Hickman catheters, accounted for 93% of the episodes. Viral and fungal infections accounted for 4% and 3% of documented cases, respectively, and occurred more than 10 days after the institution of broad-spectrum antibiotic therapy. A total of 922 different antimicrobials were employed in 190 courses (average 4.9 per course). The rationale for excessive usage and multiple changes was a persistent or intermittent fever, rather than documented infection(s). This practice led to usage of more broad-spectrum and expensive antibiotics. Further analyses indicate that the greater number of antibiotics employed correlated with apparent increased toxicity, especially renal and hepatic adverse reactions. These toxicities were associated with higher rates of fatal outcomes, i.e., 12 (39%) of 31 patients died with antibiotic-associated hepatic and/or renal insufficiency, compared with 12 (7.5%) of 159 patients who died without antibiotic-associated organ damage.
Conclusion: Excessive antibiotic usage and multiple antibiotic changes among adult leukemic patients undergoing chemotherapy appear to increase the risks of adverse hepatic and renal effects and death. Furthermore, this practice leads to use of more broad-spectrum and expensive antibiotics...