Retrospective analysis of antibiotic use and survival in advanced cancer patients with infections.

Infection is common among advanced cancer patients. This study was undertaken to review the pattern of use of antibiotics and to identify potential factors that could affect outcomes after infection. The medical records of all patients with advanced cancer who were enrolled into the palliative care service of a district hospital during the period January, 2002 to July, 2002 were retrospectively reviewed for infections and the use of antibiotics. Among the eligible 87 patients, 17 did not have any infective episode and 70 had at least one infective episode and accounted for a total of 120 episodes. Sixty-eight episodes were associated with survival for >14 days, and 52 episodes were associated with survival of < or =14 days. The most frequent sites of infection were chest (n=63, 52.5%), urinary tract (n=35, 29.2%), and skin/wound (n=6, 5%). Antibiotics were prescribed for 97.5% (n=117) episodes. The use of second-line antibiotics was 16.2% (n=19). By multivariate logistic regression analysis, dyspnea [odds ratio (OR)=2.6, 95% confidence interval (CI)=1.1-6.3], antibiotic utilization pattern [empirical therapy (OR=4.8, 95% CI=1.7-13.2) vs. therapy according to antibiotic sensitivity], and route of administration [parenteral route (OR=3.3, 95% CI=1.3-8.2) vs. oral route] were identified as independent determinants affecting survival after infection. Dyspnea was possibly associated with poor prognosis during the treatment of infections in patients with advanced cancer, and antibiotic therapy according to sensitivity was associated with better prognosis. Further studies are encouraged to verify this. The bioethical principles on the use of antibiotics as a life-sustaining treatment should always be followed.