Demographic and clinical characteristics by exposure status at cohort entry in a restricted cohort of new users of NSAIDs
| PPI analysis | ||
|---|---|---|
| PPI exposed | Unexposed* | |
| Total number of patients | 96 870 | 4 141 634 |
| Database (%) | ||
| Alberta | 6.1 | 4.3 |
| Saskatchewan | 0 | 0 |
| Manitoba | 3.0 | 5.5 |
| Ontario | 34.2 | 22.8 |
| Quebec | 1.0 | 0.7 |
| Nova Scotia | 0.5 | 1.0 |
| MarketScan | 25.9 | 52.1 |
| GPRD | 29.4 | 13.7 |
| Age (%) (years)† | ||
| 40–54 | 24.9 | 41.6 |
| 55–64 | 20.0 | 23.9 |
| 65–74 | 30.5 | 19.5 |
| 75–84 | 19.8 | 12.2 |
| 85+ | 4.9 | 2.8 |
| Women (%) | 62.8 | 58.3 |
| Year of cohort entry (%) | ||
| 1997 | 0.2 | 1.3 |
| 1998 | 0.5 | 2.3 |
| 1999 | 0.6 | 2.9 |
| 2000 | 1.4 | 4.4 |
| 2001 | 2.2 | 4.6 |
| 2002 | 3.0 | 4.6 |
| 2003 | 4.4 | 4.8 |
| 2004 | 5.7 | 4.6 |
| 2005 | 7.2 | 3.7 |
| 2006 | 9.0 | 3.7 |
| 2007 | 17.1 | 18.5 |
| 2008 | 21.0 | 19.0 |
| 2009 | 20.9 | 18.0 |
| 2010 | 6.6 | 7.6 |
| 2011 | 0.0 | 0.0 |
| Nursing home (%)‡ | 0.7 | 0.8 |
| Urban (%)§ | 85.5 | 78.8 |
| Comorbidities (%)¶ | ||
| Asthma | 2.6 | 1.8 |
| Bronchiectasis | 0.1 | 0.1 |
| COPD | 3.1 | 2.3 |
| Diabetes | 9.9 | 9.7 |
| Non-hospitalised pneumonia | 2.2 | 2.2 |
| Hospitalisations in the year preceding cohort entry (%) | ||
| None | 91.7 | 93.2 |
| One | 6.8 | 5.6 |
| Two | 1.1 | 0.9 |
| Three | 0.3 | 0.2 |
| Four or more | 0.1 | 0.1 |
| Prescriptions in the year preceding cohort entry (%) | ||
| None | 4.0 | 5.2 |
| One | 6.1 | 8.2 |
| Two | 7.4 | 9.7 |
| Three | 8.4 | 10.2 |
| Four or more | 74.1 | 66.7 |
| Medications (%)¶ | ||
| PPI** | 25.3 | 3.0 |
| H2RA** | 2.6 | 1.6 |
| NSAIDs** | 30.7 | 28.5 |
| Immunosuppressive agents | 1.3 | 1.1 |
| Influenza vaccine†† | 20.9 | 5.6 |
| Inhaled bronchodilators | 11.4 | 8.4 |
| Inhaled corticosteroids | 11.2 | 13.6 |
| Pneumococcal vaccine†† | 11.9 | 2.4 |
| Non-topical antibiotics | 43.4 | 45.0 |
| Non-topical corticosteroids | 13.5 | 16.8 |
*Saskatchewan did not contribute to the PPI analysis. In addition, small differences in the unexposed groups exist in the PPI and H2RA analyses because of trimming during the estimation of high-dimensional propensity scores.
†Alberta, Ontario, Quebec and Nova Scotia only contributed data for those >66 years of age whereas MarketScan only contributed data for those aged 40–65 years.
‡The proportion of study participants living in nursing homes was only reported by Manitoba and Ontario.
§Data regarding urban/rural location were only available in Alberta, Manitoba, Nova Scotia and Ontario.
¶Comorbidities and medication use are defined using data in the year prior to and including the date of cohort entry.
**Denotes use in the 7–12 months before cohort entry. Use in the 6 months before cohort entry resulted in exclusion.
††Vaccination data were only available for the GPRD, Manitoba, MarketScan and Nova Scotia.
COPD, chronic obstructive pulmonary disease; GPRD, General Practice Research Database; H2RA, histamine-2 receptor antagonist; NSAIDs, non-steroidal anti-inflammatory drugs; PPI, proton pump inhibitor.