Appendix Acost-Effectiveness Calculation

Appendix ACost-effectiveness calculation

The cost-effectiveness ratio () for each country,is equal to the average of the yearlycountry-specific CE-ratios (). Formally, was calculated as follows:

(1)

The numerator of equation (1) captured the change in healthcare spending, where equals the average yearly change (%) in per capita healthcare expenditure for country c and equals per capita health expenditures for year t-1 and country c. The denominator of equation (1) contained the standardized change in avoidable mortality, i.e. the decline in mortality corrected for confounders and dynamic effects, expressed in terms of life-years won. reflects this standardized yearly change in avoidable mortality for country c(see equation (2)); and is the avoidable mortality rate per 100,000 inhabitants for year t-1 and country c. was equal to the number of life years won per unit decline in the avoidable mortality rate for year t and country c. We calculated this gain in life years by taking the difference between the average age for avoidable deaths (around 60 for all countries and years) and the life expectancyat 60 by country and year.The standardized change in avoidable mortality was calculated as follows:

(2)

In equation (2) the impact of the confounders and dynamic effects is eliminated by subtracting the change in avoidable mortality as predicted by all confounding factors from its actual change. In other words, reflects thepredicted avoidable mortality while keeping healthcare expenditures constant. As a result these confounders and time effects items did not influence the cost-effectiveness ratio.

Appendix B

Regression results of the fixed-effects models(coefficients and p-values in parentheses)

Model
(1) / Model
(2) / Model
(3) / Model
(4a) / Model
(4b) / Model
(4c) / Model
(4d) / Model
(5a) / Model
(5b) / Model
(5c) / Model
(3)
RE GLS / Model
(4d)
RE GLS
Healthcare spending(LRP)1 / -0.71 (0.00) / -0.50 (0.00) / -0.37
(0.00) / -0.32 (0.00) / -0.24 (0.00) / -0.36
(0.00) / -0.27
(0.00) / -0.33
(0.00) / -0.33
(0.00) / -0.26
(0.00) / -0.34
(0.00) / -0.24
(0.00)
Time-trend (Year) / -0.01 (0.00) / -0.01
(0.00) / -0.02 (0.00) / -0.01 (0.00) / -0.01
(0.00) / -0.01
(0.00) / -0.01
(0.00) / -0.01
(0.00) / -0.02
(0.00) / -0.02
(0.00) / -0.02
(0.00)
Age structure (% > 60 yr) / 0.01
(0.08) / 0.01 (0.18) / 0.01 (0.11) / 0.02
(0.01) / 0.01
(0.03) / 0.02
(0.01) / 0.02
(0.01) / 0.02
(0.01) / 0.01
(0.09) / 0.01
(0.03)
Residual mortality / 0.67
(0.00) / 0.56 (0.00) / 0.63 (0.00) / 0.66
(0.00) / 0.63
(0.00) / 0.63
(0.00) / 0.65
(0.00) / 0.65
(0.00) / 0.68
(0.00) / 0.64
(0.00)
Education
(% low educated) / 0.00
(0.67)
Other expenditure / -0.23
(0.00) / -0.17
(0.00) / -0.11
(0.04) / -0.10
(0.12) / -0.10
(0.10) / -0.18
(0.00)
Unemployment rate / 0.01
(0.00) / 0.004
(0.02) / 0.01
(0.00) / 0.01
(0.00) / 0.01
(0.02) / 0.004
(0.03)
Tobacco consumption (t) / 0.01
(0.65)
Tobacco consumption
(t-15) / -0.05
(0.30)
Alcohol consumption (t)2 / -0.13
(0.01)
Constant term / 8.72 (0.00) / 26.5 (0.00) / 30.96
(0.00) / 35.63 (0.00) / 33.75 (0.00) / 30.88
(0.00) / 32.97
(0.00) / 29.34
(0.00) / 31.93
(0.00) / 35.78
(0.00) / 33.19
(0.00) / 34.43
(0.00)
N / 119 / 119 / 119 / 107 / 119 / 118 / 118 / 104 / 110 / 115 / 119 / 118
R2 (within) / 0.898 / 0.907 / 0.956 / 0.954 / 0.964 / 0.964 / 0.967 / 0.974 / 0.970 / 0.969 / 0.956 / 0.967
F-statistic (Prob > F) / 267.7 (0.00) / 221.5 (0.00) / 310.8
(0.00) / 285.5 (0.00) / 357.3 (0.00) / 363.1 (0.00) / 380.2
(0.00) / 412.3
(0.00) / 391.6
(0.00) / 397.5
(0.00) / - / -
BIC-criterion / -494.8 / -500.6 / -580.4 / -521.8 / -597.6 / -593.7 / -599.6 / -544.0 / -560.0 / -587.8 / - / -
Highest VIF-score / 2.7 / 2.7 / 2.9 / 2.9 / 2.7 / 3.0 / 3.5 / 3.9 / 3.3 / 4.1 / 2.9 / 3.5
Ramsey RESET test
(Ho=no omitted variables) / 0.01 / 0.83 / 0.78 / 0.50 / 0.67 / 0.76 / 0.34 / 0.69 / 0.22 / 0.37 / 0.78 / 0.40

1LRP = Long Run Propensity, i.e. the combined effect of current, one-year lagged, and two-year lagged healthcare spending.

2Alcohol consumption (t-15) wasnot significant in the univariable regression and was not included in the multiple regression models.

Regression results of the fixed-effects models by disease group(coefficients and p-values in parentheses)

Diseases of the circulatory system1 / Neoplasms1
Model
(1) / Model
(2) / Model
(3) / Model
(4d) / Model
(1) / Model
(2) / Model
(3) / Model
(4d)
Healthcare spending(LRP) / -1.10 (0.00) / -0.71 (0.00) / -0.55
(0.00) / -0.37
(0.00) / -0.30 (0.00) / -0.38 (0.00) / -0.27
(0.00) / -0.26
(0.00)
Time-trend (Year) / -0.02 (0.00) / -0.02
(0.00) / -0.02
(0.00) / 0.004 (0.37) / -0.002
(0.53) / -0.002
(0.54)
Age structure (% > 60 yr) / 0.01
(0.26) / 0.01
(0.45) / 0.02
(0.02) / 0.02
(0.04)
Residual mortality / 0.87
(0.00) / 0.81
(0.00) / 0.71
(0.00) / 0.71
(0.00)
Education
(% low educated)
Other expenditure / -0.33
(0.00) / -0.01
(0.90)
Unemployment rate / -0.003
(0.31) / -0.001
(0.69)
Constant term / 10.32 (0.00) / 42.58 (0.00) / 47.87
(0.00) / 52.18
(0.00) / 5.10 (0.00) / -1.53 (0.84) / 4.10
(0.49) / 4.21
(0.50)
N / 119 / 119 / 119 / 118 / 119 / 119 / 119 / 118
R2 (within) / 0.920 / 0.933 / 0.967 / 0.975 / 0.520 / 0.524 / 0.705 / 0.702
F-statistic (Prob > F) / 328.2 (0.00) / 295.8 (0.00) / 310.8
(0.00) / 565.3
(0.00) / 152.2 (0.00) / 147.2 (0.00) / 199.9
(0.00) / 141.3
(0.00)
BIC / -422.4 / -437.1 / -512.29 / -530.8 / -447.7 / -443.9 / -491.0 / -477.0
Highest VIF-score / 2.6 / 2.7 / 2.9 / 3.5 / 2.7 / 2.7 / 2.9 / 3.5
Ramsey RESET test(Prob > F)
(Ho=no omitted variables) / 0.00 / 0.97 / 0.93 / 0.76 / 0.04 / 0.23 / 0.02 / 0.00

1In these models the independent variable entailed all avoidable mortality from circulatory system diseases or neoplasms (Table 1).

Regression results of the growth-rate models(coefficients and p-values in parentheses)

Model
(1) / Model
(2) / Model
(3) / Model
(4a) / Model
(4b) / Model
(4c) / Model
(4d) / Model
(5a) / Model
(5b)
Health care spending (t) / -0.12
(0.22) / 0.13
(0.17) / 0.04
(0.59) / 0.10
(0.25) / 0.07
(0.36) / 0.06
(0.45) / 0.08
(0.32) / 0.05
(0.60) / 0.11
(0.17)
Health care spending (t-1) / -0.56
(0.00) / -0.28
(0.00) / -0.20
(0.02) / -0.19
(0.03) / -0.21
(0.01) / -0.20
(0.03) / -0.21
(0.01) / -0.25
(0.01) / -0.22
(0.02)
Joint-significance / (0.00) / (0.01) / (0.08) / (0.07) / (0.04) / (0.05) / (0.03) / (0.17) / (0.05)
Time-trend (Constant term) / -0.03
(0.00) / -0.02
(0.00) / -0.02
(0.00) / -0.02
(0.00) / -0.02
(0.00) / -0.02
(0.00) / -0.02
(0.00) / -0.02
(0.00)
Age structure (% > 60 yr) / -0.04
(0.54) / -0.05
(0.44) / -0.03
(0.62) / -0.04
(0.56) / -0.03
(0.61) / -0.04
(0.59) / -0.03
(0.63)
Residual mortality / 0.64
(0.00) / 0.60
(0.00) / -0.61
(0.00) / 0.64
(0.00) / 0.62
(0.00) / 0.65
(0.00) / 0.66
(0.00)
Education (% low educated) / 0.02
(0.42)
Other expenditure / -0.11
(0.10) / -0.09
(0.22)
Unemployment rate / 0.02
(0.14) / 0.02
(0.32)
Tobacco consumption (t) / 0.004
(0.91)
Tobacco consumption (t-15) / 0.006
(0.90)
N / 105 / 105 / 105 / 93 / 105 / 104 / 104 / 93 / 97
Adjusted R2 / 0.540 / 0.061 / 0.403 / 0.367 / 0.413 / 0.407 / 0.41 / 0.384 / 0.389
F-statistic (Prob > F) / 62.5
(0.00) / 4.4
(0.01) / 18.6
(0.00) / 11.7
(0.00) / 15.66
(0.00) / 15.15
(0.00) / 12.95
(0.00) / 12.45
(0.00) / 13.24
(0.00)
BIC-criterion / -451.4 / -470.7 / -511.0 / -442.8 / -509.2 / -502.9 / -499.9 / -444.5 / -461.0
Highest VIF-score / NA1 / 1.1 / 1.1 / 1.1 / 1.1 / 1.1 / 1.2 / 1.2 / 1.1
Ramsey RESET test
(Ho=no omitted variables) / NA1 / 0.08 / 0.21 / 0.23 / 0.30 / 0.10 / 0.20 / 0.41 / 0.34

1These tests could not be conducted in a model without a constant term