Outcome | Trend, % Change per Year (95% CI) | ||
---|---|---|---|
Before Enrollment | After Enrollmenta | Difference | |
Emergency department visit rate | 0.8 (0.7 to 0.9) | 1.5 (1.4 to 1.5) | 0.7 (0.6 to 0.8) |
Primary care continuity | –0.8 (–0.9 to –0.8) | 0.4 (0.3 to 0.4) | 1.2 (1.2 to 1.2) |
Primary care visit rate | 0.6 (0.6 to 0.7) | –2.2 (–2.1 to –2.3) | –2.8 (–2.8 to –2.8) |
Proportion of primary care visits with A888 code | 4.8 (4.6 to 5.0) | 8.5 (8.4 to 8.6) | 3.5 (3.4 to 3.7) |
Notes: Regression model included age (in 5-year categories), income quintile, comorbidity (adjusted diagnostic groups), and morbidity (resource utilization band) as time-varying covariates and sex as a stable variable. Fiscal year was included to account for secular trends. Table presents the percent change per year in the outcome based on the rate ratio after results of the original model performed on log(e) scale were taken to the exponent. Rate ratios were calculated using negative binomial regression analysis. Full model parameters on the log(e) scale are available in Supplemental Appendix 3 (http://www.annfammed.org/content/16/5/419/suppl/DC1).
↵a Calculated by adding the baseline trend and the change in trend in the original parameter estimates generated using log(e) scale. Estimates presented in this table are converted, so will not add up in the same way.