Table 1 Model parameters for screening accuracy and treatment pathway

  Whooley positive

0.0909

906

0.085–0.097

Howard et  al., 2018^a [15]

Beta

–

  Whooley negative

0.9091

9057

0.903–0.915

Howard et al., 2018^a [15]

Beta

–

  Whooley positive - true positive

0.4530

410.4

0.420–0.485

Howard et al., 2018^a [15]

Beta

–

  Whooley positive - false positive

0.5470

495.6

0.515–0.580

Howard et al., 2018^a [15]

Beta

–

  Whooley negative - true negative

0.9341

8460

0.929–0.939

Howard et al., 2018^a [15]

Beta

–

  Whooley negative - false negative

0.0659

596.8

0.061–0.071

Howard et al., 2018^a [15]

Beta

–

  EPDS positive

0.1144

1138

0.108–0.121

Howard et al., 2018^a [15]

Beta

–

  EPDS negative

0.8856

8809

0.879–0.892

Howard et al., 2018^a [15]

Beta

–

  EPDS positive - true positive

0.5188

590.6

0.490–0.548

Howard et al., 2018^a [15]

Beta

–

  EPDS positive - false positive

0.4813

547.9

0.452–0.510

Howard et al., 2018^a [15]

Beta

–

  EPDS negative - true negative

0.9534

8398

0.949–0.958

Howard et al., 2018^a [15]

Beta

–

  EPDS negative - false negative

0.0466

410.3

0.042–0.051

Howard et al., 2018^a [15]

Beta

–

  Whooley positive

0.0895

890.2

0.084–0.095

Howard et al., 2018^a [15]

Beta

–

  Whooley negative

0.9105

9057

0.905–0.916

Howard et al., 2018^a [15]

Beta

–

  Whooley positive, EPDS positive

0.4114

366.2

0.379–0.444

Howard et al., 2018^a [15]

Beta

–

  Whooley positive, EPDS negative

0.5886

524

0.556–0.621

Howard et al., 2018^a [15]

Beta

–

  Whooley positive, EPDS positive - true positive

0.7500

8460

0.741–0.759

Howard et al., 2018^a [15]

Beta

–

  Whooley positive, EPDS positive - false positive

0.2500

596.8

0.241–0.259

Howard et al., 2018^a [15]

Beta

–

  Whooley positive, EPDS negative - true negative

0.7531

274.6

0.708–0.797

Howard et al., 2018^a [15]

Beta

–

  Whooley positive, EPDS negative - false negative

0.2469

91.55

0.203–0.291

Howard et al., 2018^a [15]

Beta

–

  Whooley negative - true negative

0.9341

394.6

0.910–0.953

Howard et al., 2018^a [15]

Beta

–

Whooley negative - false negative

0.0659

129.4

0.047–0.090

Howard et al., 2018^a [15]

Beta

–

  No-screen positive

0.0438

6

0.016–0.084

Hearn et al., 1998 [28]

Beta

–

  No-screen negative

0.9562

131

0.916–0.984

Hearn et al., 1998 [28]

Beta

–

  No-screen positive - true positive

0.6667

4

0.284–0.947

Hearn et al., 1998 [28]

Beta

–

  No-screen positive - false positive

0.3333

2

0.053–0.716

Hearn et al., 1998 [28]

Beta

–

  No-screen negative - true negative

0.8855

116

0.826–0.934

Hearn et al., 1998 [28]

Beta

–

  No-screen negative - false negative

0.1145

15

0.066–0.174

Hearn et al., 1998 [28]

Beta

–

  Facilitated self help for mild/moderate depression

0.7921

79.21

0.705–0.864

Howard et al., 2018 [15]

Beta

Assuming 50% of women with moderate depression receive this treatment

  High intensity psychological therapy for moderate/severe depression

0.2079

20.79

0.136–0.295

Howard et al., 2018 [15]

Beta

Assuming 50% of women with moderate depression receive this treatment

  Spontaneous recovery

0.3300

33

0.242–0.425

Dennis et al., 2009 [29]

Beta

Midpoint of spontaneous recovery rate (25–40% = 33%).

  No spontaneous recovery

0.6700

67

0.575–0.758

Dennis et al., 2009 [29]

Beta

One minus midpoint of spontaneous recovery rate.

  Identified as depressed following first antenatal appointment

0.1025

10.25

0.050–0.166

Kessler et al., 2002 [30]

Beta

Based on 41% of misdiagnoses identified over the following 3 years.

  Not identified as depressed following first antenatal appointment

0.8975

89.75

0.834–0.950

Kessler et al., 2002 [30]

Beta

One minus rate of identification.

  Respond to facilitated self help

0.5109

51.09

0.413–0.607

NICE 2014 [17]

Beta

One minus probability of not responding.

  No response to facilitated self help

0.4891

48.91

0.393–0.587

NICE 2014 [17]

Beta

Relative risk of no improvement (0.73) reported in NICE (2014) [17] multiplied by absolute risk of no improvement (0.67) reported by Dennis et al. (2009) [29] reported above.

  Respond to high intensity psychological therapy

0.6784

67.84

0.586–0.767

NICE 2014 [17]

Beta

One minus probability of not responding.

  No response to high intensity psychological therapy

0.3216

32.16

0.233–0.414

NICE 2014 [17]

Beta

Relative risk of no improvement (0.48) reported in NICE (2014) [17] multiplied by absolute risk of no improvement (0.67) reported by Dennis et al. (2009) [29] reported above.