Table 3. Association between Usual Provider Continuity Index (UPI) and statin adherence (Medication Possession Ratio MPR ≥ 80) using propensity score matching * .
| MPR ≥ 80 | RR (95% CI) | |||||
|---|---|---|---|---|---|---|
| Propensity match | UPI (min-max) | No | Yes | Crude | Model 1 | Model 2 |
| Match 1 | ||||||
| Low tertile | 9.0–66.7 | 2529 (22.4) | 8788 (77.6) | 1.00 | 1.00 | 1.00 |
| Medium tertile | 66.8–88.9 | 2185 (19.3) | 9132 (80.7) | 1.04 (1.01-1.07) | 1.04 (1.01-1.07) | 1.04 (1.01-1.07) |
| Match 2 † | ||||||
| Low tertile | 9.0–66.7 | 2181 (22.7) | 7444 (77.3) | 1.00 | 1.00 | 1.00 |
| High tertile | 89.0–100 | 1781 (18.5) | 7844 (81.5) | 1.05 (1.02-1.09) | 1.05 (1.02-1.09) | 1.05 (1.02-1.09) |
| Match 3 | ||||||
| UPI <0.75 | 9.0–74.9 | 3297 (21.9) | 11742 (78.1) | 1.00 | 1.00 | 1.00 |
| UPI ≥0.75 | 75–100 | 2818 (18.7) | 12221 (81.3) | 1.04 (1.02-1.07) | 1.04 (1.01-1.07) | 1.04 (1.01-1.07) |
* Separate propensity matches were performed for the usual provider continuity index between cohorts of (1) low and medium tertiles; (2) low and high tertiles; (3) having a usual provider of care (UPI ≥0.75) or not. Propensity matching was performed using all covariates described in Table 1, including age, gender, highest education qualification, Aboriginal or Torres Strait Islander status, language other than English spoken at home, partnership status, private health insurance, employment status, annual household income, remoteness of residence, body mass index, current smoking status, alcohol consumption, fruit and vegetable consumption, physical exercise, self-rated health, self-reported heart disease, number of comorbidities, functional limitation, psychological distress, and new to statin status.
† There were no significant differences (Chi-square p-value < 0.05) between matched cohorts in the distribution of variables used for propensity matching, with the exception of physical activity and new to statin status within the cohort from Match 2.