Table 3. Pooled changes in hospital admission rates for chronic obstructive pulmonary disease (COPD) immediately after and at the one-year mark of the implementation of the 2006 partial and 2011 comprehensive smoking bans by characteristics of the fourteen largest Spanish provinces, 2003–2012*.
| 25th and 75th percentiles of province-level characteristic | 2006 partial smoking ban† | 2011 comprehensive smoking ban‡ | p-value for effect modification§ | I2 for residual heterogeneity¶ (%) | ||
|---|---|---|---|---|---|---|
| Immediate percent change (95% CI) | One-year percent change (95% CI) | Immediate percent change (95% CI) | One-year percent change (95% CI) | |||
| Latitude (degrees north) | 0.31 | 98.1 | ||||
| 37.6 | -15.5 (-24.8, -5.0) | -15.3 (-25.3, -4.0) | 1.9 (-2.0, 6.0) | 6.6 (-2.0, 15.9) | ||
| 41.7 | -12.9 (-23.4, -1.0) | -10.6 (-22.0, 2.5) | -3.3 (-6.6, 0.2) | -1.2 (-9.8, 8.2) | ||
| Human development index | 0.04 | 97.4 | ||||
| 93.1 | -16.6 (-27.0, -4.8) | -16.6 (-27.6, -4.0) | 0.7 (-4.1, 5.6) | 10.9 (1.9, 20.6) | ||
| 96.1 | -12.6 (-22.9, -0.9) | -10.3 (-21.5, 2.4) | -2.0 (-5.7, 1.9) | -3.2 (-10.5, 4.6) | ||
| Population unemployed (%) | 0.05 | 97.6 | ||||
| 10.8 | -11.0 (-22.4, 2.0) | -7.1 (-19.2, 6.9) | -3.0 (-7.1, 1.2) | -6.6 (-13.2, 0.5) | ||
| 18.4 | -18.3 (-29.3, -5.5) | -20.1 (-31.2, -7.1) | 2.2 (-3.4, 8.1) | 15.5 (6.5, 25.4) | ||
| Rural population (%) | 0.02 | 98.2 | ||||
| 8.3 | -8.0 (-18.8, 4.1) | -6.5 (-18.2, 7.0) | 0.9 (-3.6, 5.6) | 6.6 (-4.2, 18.6) | ||
| 18.0 | -19.2 (-27.8, -9.5) | -18.4 (-27.7, -7.9) | -2.2 (-6.2, 2.0) | 0.5 (-8.8, 10.7) | ||
| # of hospital beds per 1,000 population | 0.32 | 98.1 | ||||
| 3.1 | -12.0 (-22.8, 0.2) | -13.5 (-25.1, -0.1) | 0.0 (-4.5, 4.6) | 8.8 (-0.8, 19.3) | ||
| 3.8 | -16.6 (-25.9, -6.0) | -13.7 (-24.3, -1.5) | -1.4 (-5.2, 2.5) | -1.0 (-9.0, 7.7) | ||
| # of physicians per 1,000 population | 0.55 | 98.2 | ||||
| 4.0 | -15.4 (-26.2, -3.0) | -14.2 (-26.0, -0.5) | -0.7 (-5.4, 4.3) | 7.3 (-3.1, 18.8) | ||
| 5.2 | -13.9 (-24.3, -2.0) | -12.8 (-24.1, 0.2) | -1.0 (-5.1, 3.1) | -0.5 (-9.5, 9.5) | ||
* Pooled changes by province-level characteristics were obtained from separate random-effects multivariate meta-regression models relating province-specific estimates of segmented regression coefficients with their values of the corresponding characteristic as a single continuous covariate.
† Pooled percent changes in hospital admission rates and their 95% confidence intervals (CIs)at the implementation and one year following the partial ban, compared with the projected trend from the pre-ban period, for the 25th and 75th percentiles of the province-level characteristic.
‡ Pooled percent changes in hospital admission rates and their 95% CIs at the implementation and one year following the comprehensive ban, compared with the projected trend from the previous partial ban period, for the 25th and 75th percentiles of the province-level characteristic.
§ Overall P value for modification of pooled ban effects by the corresponding province-level characteristic.
¶ Overall I2 statistic for residual heterogeneity in province-specific ban effects not explained by the corresponding province-level characteristic.