Abstract
Objective: To identify low and high-risk sub-groups of patients admitted for Health-Care Associated Pneumonia (HCAP).
Methods: We conducted a retrospective study of patients hospitalized for HCAP. Demographic data, co-morbidity, clinical features, laboratory findings, and chest radiographic changes were used to construct logistic regression models that employed Pneumonia Severity Index (PSI) or CURB-65 to predict 30-day mortality and guide the creation of low- and high-risk sub-groups.
Results: Among 798 hospital episodes for HCAP, the median age was 80 years and 67. 9% of patients had pre-morbid ambulation impairment. The 30-day mortality was 35.3 %. For the subgroup (3.5%) of patients with PSI class II with no pre-morbid ambulation impairment, 30-day mortality was 0%. At the other end, CURB-65 score of 4 or 5 with pre-morbid ambulation impairment identified a subgroup (3.0%) of patients with 30-day mortality of 83.3%.
Conclusion: Combining either PSI or CURB-65 with pre-morbid ambulation status, we were able to identify very low and high-risk mortality subgroups among HCAP patients. In the appropriate clinical context, early discharge may be considered for the very low risk subgroup, while de-escalation of antibiotic therapy and symptom palliation may be considered for the very high-risk subgroup.
Keywords: Antibiotic, co-morbidity, pneumonia, severity.