Cohort characteristics
| Subjects n | 169 |
| Age years | 68, 16–98 (18) |
| Males n (%) | 88 (52.0%) |
| BMI# kg·m2 | 26 (22–30) |
| Ethnicity | |
| White British | 166 (98.2%) |
| White other | 2 (1.2%) |
| Black African | 1 (0.6%) |
| Comorbidities | |
| COPD | 70 (41.0%) |
| Chronic lung disease other than COPD | 21 (12.4%) |
| Congestive cardiac failure | 23 (13.6%) |
| Dementia | 2 (1.2%) |
| Diabetes# | 28 (16.7%) |
| Hepatic disease | 5 (3.0%) |
| Renal disease | 14 (8.3%) |
| Lived in nursing/residential care | 8 (4.7%) |
| Smoking status# | |
| Active smoker | 63 (39%) |
| Ex-smoker | 66 (41%) |
| Never-smoker | 32 (20%) |
| Charlson comorbidity index | |
| 0 | 56 (33.1%) |
| 1 | 69 (40.8%) |
| 2 | 18 (10.7%) |
| 3 | 17 (10.1%) |
| 4 | 6 (3.6%) |
| 5 | 2 (1.2%) |
| 6 | 1 (0.6%) |
| >6 | 0 |
| Influenza infection# | 18 (16.8%) |
| CURB65 score | |
| 0–1 | 79 (46.7%) |
| 2 | 50 (29.6%) |
| 3–5 | 40 (23.7%) |
| Infection markers | |
| Pyrexial | 90 (53.0%) |
| Neutrophil count ×109 per L | 9.9 (7.1–14.8) |
| CRP mg·mL1 | 145 (61–248) |
| Pro-calcitonin# ng·mL1 | 0.70 (0.1–3.9) |
| >0.25 ng·mL1 | 98 (64.5%) |
| >0.5 ng·mL1 | 83 (54.6%) |
| Antibiotic regimen¶ | |
| Appropriate | 107/159 (67.3%) |
| Over treated | 41/159 (25.8%) |
| Under treated | 11/159 (6.9%) |
| Received macrolide | 133/159 (83.6%) |
| Outcome | |
| Length of stay days | 6, 0–58 (7.8) |
| Readmission within 30 days of discharge | 16/135 (11.8%) |
| In-hospital mortality | 13 (7.7%) |
| Death within 30 days of discharge | 1/135 (0.7%) |
| Death post discharge | 13/135 (9.6%) |
| Total 1-year mortality | 26 (15.4%) |
| Cause of in-hospital death | |
| CAP | 8 (61.5) |
| Sepsis | 2 (15.4) |
| Myocardial infarction | 1 (7.7) |
| Respiratory failure | 1 (7.7) |
| Unknown | 1 (7.7) |
| Cause of death post discharge | |
| CAP | 2 (15.4%) |
| HAP | 1 (7.7%) |
| Gastric cancer | 1 (7.7%) |
| Lung cancer | 3 (23.1%) |
| Interstitial lung disease | 1 (7.7%) |
| COPD | 2 (15.4%) |
| Unknown | 3 (23.1%) |
Data are presented as median, range (sd), or median (interquartile range), unless otherwise stated. BMI: body mass index; COPD: chronic obstructive pulmonary disease; CURB65: confusion, urea, respiratory rate, blood pressure, age >65 years; CRP: C-reactive protein; CAP: community-acquired pneumonia; HAP: hospital-acquired pneumonia. #: incomplete data for diabetes (n=168), BMI (n=126), smoking status (n=161) and pro-calcitonin (n=166). ¶: initial empirical antibiotic choice was deemed appropriate if it was consistent with that stated in the local guidelines. Local guidelines are based on the British Thoracic Society guidelines and are based around CURB65 score on admission. Over-treatment was therefore a treatment regime ordinarily reserved for a higher CURB65 score and under-treatment was a regime aimed at lower risk patients based on the CURB65 score. Here, the assessment of appropriateness does not take into account factors other than CURB65 score, such as treatment duration. Patients were recorded as having received a macrolide if at any point in their pneumonia treatment they received a macrolide of any sort and for any duration.