Can C-reactive protein-to-albumin ratio be a predictor of short-term mortality in community-acquired pneumonia?