Table 3

Immediate and long-term effects of the events under study based on segmented logistic regression models (OR and 95% CI)*

Poor population†Low/middle-income pensionersLow-income non-pensionersMiddle-income non-pensionersHigh-income population†
OR95% CIOR95% CIOR95% CIOR95% CIOR95% CI
Baseline level of IMNA (intercept, Embedded Image ) 0.311 0.297 to 0.325 0.241 0.235 to 0.246 0.243 0.240 to 0.247 0.243 0.238 to 0.249 0.251 0.207 to 0.303
Baseline trend per week of IMNA (preintervention slope, Embedded Image )‡1.0010.999 to 1.002 0.997 0.997 to 0.998 1.0000.999 to 1.0011.0000.999 to 1.0011.0030.996 to 1.010
News release
 Immediate effect of the event (level change in IMNA rate from
previous week, Embedded Image
0.862 0.819 to 0.907 0.9920.966 to 1.019 0.963 0.947 to 0.980 0.9860.957 to 1.0151.1020.850 to 1.428
 Long-term effect of the event (gradual change in slope, per week, Embedded Image 0.9990.997 to 1.001 0.992 0.991 to 0.993 0.998 0.997 to 0.998 0.997 0.996 to 0.998 0.984 0.975 to 0.993
Introduction of the fixed copayment**
 Immediate effect of the event (level change in IMNA rate from
previous week, Embedded Image
1.315 1.216 to 1.422 1.370 1.319 to 1.423 1.047 1.015 to 1.079 1.0380.985 to 1.0931.2310.804 to 1.882
 Long-term effect of the event (gradual change in slope, per week, Embedded Image 0.962 0.948 to 0.976 1.058 1.052 to 1.064 1.023 1.015 to 1.030 1.028 1.015 to 1.041 1.104 1.020 to 1.194
Change in coinsurance scheme**
 Immediate effect of the event (level change in IMNA rate from
previous week, Embedded Image
0.9570.882 to 1.039 0.949 0.907 to 0.993 0.941 0.901 to 0.983 0.9940.922 to 1.0730.6590.415 to 1.048
 Long-term effect of the event (gradual change in slope, per week, Embedded Image 1.029 1.025 to 1.032 0.982 0.980 to 0.985 0.9990.997 to 0.999 0.996 0.994 to 0.998 0.9880.969 to 1.008
Suspension of fixed copayment
 Immediate effect of the event (level change in IMNA rate from
previous week, Embedded Image
0.676 0.642 to 0.711 0.883 0.865 to 0.912 0.927 0.912 to 0.942 0.856 0.833 to 0.880 0.9220.720 to 1.180
 Long-term effect of the event (gradual change in slope, per week, Embedded Image 0.995 0.994 to 0.995 0.999 0.998 to 0.999 0.999 0.999 to 0.999 1.0001.000 to 1.0011.0021.000 to 1.005
  • *Models were adjusted for the characteristics of patients, FPs and PC centres as well as for seasonality (week of the month and month of the year).

  • †Includes pensioners and non-pensioners.

  • ‡Bold numbers indicate statistically significant difference from zero.

  • §This row reports the level change in IMNA prevalence (OR)—that is, the increase/decrease in IMNA prevalence—immediately after the event week; bold numbers indicate statistically significant differences (95% confidence probability).

  • ¶This row reports the period trend in OR per 1-week increase. Bold numbers indicate statistically significant differences between each period slope and initial period slope.

  • **The end of the third period and the beginning of the fourth period are not the same for all population profiles. For non-pensioner only profiles (40% and 50%) the period change was in the 83rd week, for pensioner only profiles (40%) the period change was in the 88th week, and for pensioner and non-pensioner profile groups the period change was in the 85th week.

  • FP, family physician; IMNA, initial medication non-adherence; PC, primary care.