Abstract
Dear Editor:
Infections in patients with cancer in palliative care have high morbidity and mortality rates. 1 The appropriate prescription of antibiotics for subjects in this setting is a complex decision. Antibiotics may relieve symptoms but may also prolong the dying process and increase the symptoms producing interventions, such as laboratory testing or venous access.
The prevalence of the use of antibiotics ranged from 36% to 84% in a few published reports.2–4 Possibly, the values, beliefs, and practices of healthcare teams working in developing countries are different from those described in these published studies.
We developed a retrospective study at the palliative care unit (PCU) of the National Cancer Institute of Brazil. The PCU is a 56-bed hospital and is the only public cancer PCU located in the city of Rio de Janeiro, Brazil.
This study included 870 subjects referred to the PCU during 2010. Mean age was 62 years, 52% were female, and head and neck cancer was the most common cancer (28%). Twelve percent were cancer therapy naive, 38% had at least one device, and 54% had Karnofsky Performance Status (KPS) ≥50%. Three hundred fifty-four participants (41%) received at least one antimicrobial course. Two hundred five subjects used a single course, whereas 149 subjects received more than one course, totalizing 646 courses. Mean time of antimicrobial use was 8.1 days and they were maintained until death in 15% of the courses. The enteral route was used in 60% of the courses. By protocol definition, the reason for antimicrobial initiation included two categories: an infectious disease diagnosis reported by the prescribing physician or the condition in which only symptoms were mentioned in the records (e.g., for respiratory symptoms). Table 1 describes the reasons for antimicrobial use. The most prescribed antibiotic was amoxicillin + clavulanate (41%). Independent predictors of antimicrobial use were the presence of head and neck cancer, previous cancer treatment, and any device use.
Antibiotic use was highly prevalent in our cohort. The decision to start an antimicrobial was made without the citation of an infectious disease diagnosis in 311 (48%) of the courses. In 235 (36%) of the courses, physicians justified the initiation with the presence of new signs and symptoms. In both categories, there was a predominance of involvement of the respiratory tract.
It is possible that subjects have been over treated as symptoms of cancer progression may have mimicked infections. In addition, no standardized definition of infection was available in the charts.
In conclusion, the decision to start or not antimicrobials in palliative care is a difficult one. Some ethical issues should always be considered, such as the prolongation of the dying process and the uncertainty about symptom relief. In addition, the impact of antimicrobial use in drug resistance in the community and the increase in healthcare-associated costs should also be considered. In a palliative care setting, antibiotics should be prescribed to improve the quality of life and not the survival time, because the underlying disease defines the latter.