Table 4 Model performance of three regression methods for mapping the PHQ-8 to the EQ-5D-5L utility scores

1

BETAMIX M1a

1 component without truncation

Probability mass at full health

0.0354

0.1374

0.1720

18

18

18

2

BETAMIX M1b

2 components without truncation

Probability mass at full health

0.0010

0.1296

0.1708

15

17

16

3

BETAMIX M1c

2 components with truncation

Probability mass at full health

−0.0007

0.1274

0.1687

11

14

12.5

4

BETAMIX M1d

2 components with truncation

Probability mass at full health and truncation point

0.0031

0.1293

0.1696

14

16

15

5

BETAMIX M2a

1 component without truncation

Probability mass at full health

0.0355

0.1338

0.1691

17

15

16

6

BETAMIX M2b

2 components without truncation

Probability mass at full health

−0.0018

0.1254

0.1670

7

11

9

7

BETAMIX M2c

2 components with truncation

Probability mass at full health

0.0002

0.1243

0.1656

5

8

6.5

8

BETAMIX M2d

2 components with truncation

Probability mass at full health and truncation point

0.0051

0.1258

0.1657

10

9

9.5

9

BETAMIX M3a

1 component without truncation

Probability mass at full health

0.0357

0.1297

0.1663

16

10

13

10

BETAMIX M3b

2 components without truncation

Probability mass at full health

−0.0011

0.1213

0.1631

3

2

2.5

11

BETAMIX M3c

2 components with truncation

Probability mass at full health

0.0031

0.1212

0.1632

2

4

3

12

BETAMIX M3d

2 components with truncation

Probability mass at full health and truncation point

0.0023

0.1208

0.1620

1

1

1

13

OLS M1

0.0000

0.1279

0.1673

13

12

12.5

14

OLS M2

0.0000

0.1255

0.1642

8

5

6.5

15

OLS M3

0.0000

0.1238

0.1631

4

3

3.5

16

TOBIT M1

−0.0098

0.1278

0.1681

12

13

12.5

17

TOBIT M2

−0.0101

0.1256

0.1653

9

7

8

18

TOBIT M3

−0.0098

0.1245

0.1644

6

6

6