Skip to main content
. 2014 May 30;9(5):e97149. doi: 10.1371/journal.pone.0097149

Table 1. Summary of Confounding Variables, by Category, Measured Prior to Index Prescription and by which Propensity Scores were Adjusted.

Demographics Respiratory disease severity (past year) General health Co-medications associated previously with CAP (past year)
Age Asthma Healthcare use (in past year) Statins [34]
- Number of emergency hospitalizations
- Number of non-emergency hospitalizations
- Number of general practice visits
Gender COPD severity Charlson comorbidity chapters as defined by Khan et al [36] minus HIV (too infrequent) and respiratory disease (ever) ACE-inhibitors [35]
- Lung function less than FEV1 % predicted
- Moderate to severe dyspnea (Medical Research Council >2)
- Number of moderate and severe COPD exacerbations
Calendar year of index prescription Respiratory medications Additional comorbidities associated Immunosuppressants [37]
- Short-acting beta-agonists with pneumonia (ever)
- Oral corticosteroids - Depression
- Theophyllines - Anxiety
- Oxygen use - GERD
- Nebulized therapies
Smoking status Pneumonia (past year) GERD medications
Body mass index Vaccinations Benzodiazepines and non-
- Influenza (past year) benzodiazepine sedatives [38]
- Pneumococcal (past 5 years)
Socio-economic status
- Overall Social Deprivation Scores for England [39]
- Townsend Score
- Pneumonia in the baseline period (CAP and severe CAP)

ACE: angiotensin-converting-enzyme; CAP: community-acquired pneumonia; FEV1: forced expiratory volume in 1s; GERD: gastroesophageal reflux disease.