Skip to main content

Table 2 Parameters applied in the Markov model, their distributions and the values used to estimate the distributions

From: Pharmacy-based screening to detect persons at elevated risk of type 2 diabetes: a cost-utility analysis

Parameter Value (variation) Distribution Distribution values used in PSA
Mean (SE)
Source
Hazard ratios
 Effect of intervention (95% CI) 0.74 (0.53–1.03) Lognormal 0.74 (0.17) [26]
 T2D-specific mortality risk (95% CI) HR 2.47 (2.42–3.06) (women) Lognormal 2.47 (0.04) [27]
HR 1.93 (1.79–2.07) (men)   1.93 (0.05)  
 Mortality risk associated with T2D with complications HR 2.36 (1.70–3.29) Lognormal 2.36 (0.41) [28]
 All-cause mortalitya Based on age and gender [29]
Utilities
 Baseline utilities (EQ-5D-3L) Women
(Age, Utility, SE)
Beta Alpha
(Age, value)
Women
Beta
(Age, value)
Women
[30]
30–44 0.906 (0.003)   30–44 8573 30–44, 889  
45–54 0.865 (0.005)   45–54 4040 45–54, 631  
55–64 0.810 (0.006)   55–64 3463 55–64, 812  
65+  0.770 (0.008)   65+  2130 65+  636  
Men
(Age, Utility, SE)
  Men Men
30–44 0.917 (0.003)   30–44 7755 30–44, 702  
45–54 0.876 (0.005)   45–54 3806 45–54, 539  
55–64 0.821 (0.006)   55–64 3351 55–64, 731  
65+  0.781 (0.008)   65+  2087 65+  585
 Disutility of T2D (EQ-5D-3L) (SE) 0.041 (0.012) Beta Alpha
11.19
Beta
261.9
[31]
 Weighted disutility of T2D complications (EQ-5D-3L)b 0.119 (0.089–0.149) Beta 0.119 (0.015) Disutility values of individual complications [32,33,34,35,36]
Proportion of complications [37]
  1. aAll-cause mortality data is shown in additional file 7
  2. bFor variables without available confidence interval, a variation of ±25% has been used as an estimate. In these cases, SE has been calculated as:
  3. \( SE=\frac{Mean\ast 1.25- Mean\ast 0.75}{2\ast 1.96} \)