Skip to main content

Table 5 Pooled positive predictive value and negative predictive value according to prevalence

From: DNA methylation as a triage marker for colposcopy referral in HPV-based cervical cancer screening: a systematic review and meta-analysis

 

N datasets

PPV (mean ± SD)

Set prevalence

Pooled prevalencee

5%

PPV (mean ± SD)

10%

PPV (mean ± SD)

20%

PPV (mean ± SD)

30%

PPV (mean ± SD)

40%

PPV (mean ± SD)

50%

PPV (mean ± SD)

60%

PPV (mean ± SD)

CIN2 + detection

All studies with all markers a,b [12,13,14,15,16,17,18,19,20,21,22, 24, 25, 27,28,29, 31,32,33,34]

56

0.620

(0.601–0.639)

0.138

(0.128–0.148)

0.252

(0.237–0.267)

0.432

(0.412–0.452)

0.565

(0.545–0.585)

0.669

(0.651–0.687)

0.752

(0.737–0.767)

0.820

(0.808–0.832)

All studies with best markersa,c [12,13,14,15,16,17,18,19,20,21,22, 24, 25, 27,28,29, 31,32,33,34]

25

0.514

(0.491–0.537)

0.126

(0.116–0.136)

0.233

(0.216–0.250)

0.405

(0.382–0.428)

0.538

(0.515–0.561)

0.644

(0.622–0.666)

0.731

(0.713–0.749)

0.803

(0.788–0.818)

Studies with CADM1, FAM19A4, MAL and/or miR124-2a [12,13,14, 17,18,19,20,21,22, 28, 33]

15

0.446

(0.424–0.468)

0.113

(0.104–0.122)

0.212

(0.197–0.227)

0.376

(0.355–0.397)

0.508

(0.486–0.530)

0.616

(0.595–0.637)

0.706

(0.688–0.724)

0.783

(0.768–0.798)

Set threshold to achieve 70% specificity [12, 14, 17,18,19,20, 22, 33]

9

0.478

(461–0.494)

0.106

(0.099–0.113)

0.200

(0.189–0.211)

0.359

(0.343–0.375)

0.490

(0.473.0.507)

0.599

(0.582–0.616)

0.691

(0.676–0.706)

0.770

(0.758–0.782)

CIN3 + detection

All studies with all markers a,b [12,13,14,15,16,17,18,19,20,21,22,23,24, 26, 27, 29,30,31,32, 34]

55

0.515

(0.497–0.533)

0.151

(0.142–0.160)

0.273

(0.258–0.288)

0.457

(0.439–0.475)

0.591

(0.573–0.609)

0.692

(0.676–0.708)

0.771

(0758–0.784)

0.835

(0.825–0.845)

All studies with best markersa,d [12,13,14,15,16,17,18,19,20,21,22,23,24, 26, 27, 29,30,31,32, 34]

25

0.392

(0.371–0.413)

0.134

(0.123–0.145)

0.246

(0.229–0.263)

0.423

(0.401–0.445)

0.557

(0.513–0.579)

0.661

(0.641–0.681)

0.745

(0.734–0.756)

0.814

(0.800–0.828)

Studies with CADM1, FAM19A4, MAL and/or miR124-2a [12,13,14, 17,18,19,20,21,22,23]

14

0.294

(0.274–0.314)

0.119

(0.109–0.129)

0.221

(0.204–0.238)

0.389

(0.365–0.413)

0.522

(0.497–0.547)

0.629

(0.606–0.652)

0.718

(0.698–0.738)

0.792

0.776–0.808)

Set threshold to achieve 70% specificity [12, 14, 17,18,19,20, 22, 33]

9

0.347

(0.326–0.368)

0.120

(0.110–0.130)

0.224

(0.208–0.240)

0.393

(0.371–0.415)

0.526

(0.503–0.549)

0.633

(0.612–0.654)

0.721

(0.703–0.739)

0.795

(0.780–0.810)

 

N studies

NPV (mean ± SD)

Set prevalence

Pooled prevalencee

5%

NPV (mean ± SD)

10%

NPV (mean ± SD)

20%

NPV (mean ± SD)

30%

NPV (mean ± SD)

40%

NPV (mean ± SD)

50%

NPV (mean ± SD)

60%

NPV (mean ± SD)

CIN2 + detection

All studies with all markers a,b[12,13,14,15,16,17,18,19,20,21,22, 24, 25, 27,28,29, 31,32,33,34]

56

0.880

(0.875–0885)

0.978

(0.977–079)

0.954

(0.952–0.956)

0.902

(0.898–0.906)

0.844

(838–850)

0.776

(0.768–0.784)

0.698

(0.688–0.708)

0.607

(0.596–0.618)

All studies with best markersa,c[12,13,14,15,16,17,18,19,20,21,22, 24, 25, 27,28,29, 31,32,33,34]

25

0.857

(0.848–0.866)

0.978

(0.976–0.980)

0.954

(0.951–0.957)

0.903

(0.897–0.909)

0.844

(0.834–0.854)

0.777

(0.764–790)

0.699

(0.684–0.714)

0.608

(0.591–0.625)

Studies with CADM1, FAM19A4, MAL and miR124-2a[12,13,14, 17,18,19,20,21,22, 28, 33]

15

0.858

(0.859–0.866)

0.975

(0.973–0.977)

0.948

(0.945–0.951)

0.890

(0.883–0.897)

0.825

(0.815–0.835)

0.752

(0.739–0.765)

0.669

(0.654–0.684)

0.574

(0.557–0.591)

Set threshold to achieve 70% specificity[12, 14, 17,18,19,20, 22, 33]

9

0.815

(0.804–0.826)

0.972

(0.970–0.974)

0.942

(0.938–0.946)

0.878

(0.871–0.885)

0.808

(0.797–0.819)

0.730

(0.716–0.744)

0.643

(0.627–0.659)

0.546

(0.529–0.563)

CIN3 + detection

All studies with all markersa,b[12,13,14,15,16,17,18,19,20,21,22,23,24, 26, 27, 29,30,31,32, 34]

55

0.916

(0.911–0.921)

0.985

(0.984–0.986)

0.969

(0.967–0.971)

0.933

(0.929–0.937)

0.890

(0.884–0.896)

0.839

(0.830–0.847)

0.776

(0.765–0.787)

0.698

(0.684–0.712)

All studies with best markersa,d[12,13,14,15,16,17,18,19,20,21,22,23,24, 26, 27, 29,30,31,32, 34]

25

0.938

(0.932–0.006

0.984

(0.982–0.986)

0.967

(0.964–0.970)

0.929

(0.922–0.936)

0.885

(0.875–0-895)

0.832

(0.818–8.46)

0.768

(0.750–0.783)

0.688

(0.666–0.710)

Studies with CADM1, FAM19A4, MAL and miR124-2a[12,13,14, 17,18,19,20,21,22,23]

14

0.939

(0.934–0.944)

0.979

(0.977–0.981)

0.958

(0.954–0.962)

0.910

(0.903–0.917)

0.855

(0.844–0.866)

0.791

(0.777–0.805)

0.716

(0.698–0.734)

0.627

(0.607–0.647)

Set threshold to achieve 70% specificity[12, 14, 17,18,19,20, 22, 33]

9

0.919

(0.912–0.926)

0.978

(0.976–0.980)

0.955

(0.951–0.959)

0.903

(0.895–0.911)

0.845

(0.833–0.857)

0.778

(0.762–0.794)

0.701

(0.681–0.721)

0.610

(0.588–0.632)

  1. aThe sensitivity and specificity were estimated as reported by authors. When multiple thresholds reported, 70% specificity was selected. bPooled together all the genes and gene combination reported in each study. For studies [16, 23, 25, 26, 29, 32], more than one entrance was considered. cOnly one entrance per study was considered. The best combination reported by the authors was selected. For [16] and [32] was considered JAM3; for [29] and [27] was considered C13orf18/EPB41L3/JAM3; for [25] was considered S5 classifier (EPB41L3/HPV16L1/HPV16L2/HPV18L2/HPV31L1/HPV33L2). dOnly one entrance for study was considered. The best combination reported by the authors was selected. For [30] was considered PAX1/ZNF582; for [16] was considered JAM3; for [29] was considered C13orf18/EPB41L3/JAM3; for [25] was considered S5 classifier; for [27] was considered SOX1/ZSCAN1; for [32] was considered SOX1. ePresented in Table 2
Back to article page

Clinical Epigenetics

ISSN: 1868-7083

Contact us