Skip to main content

Table 4 Pre-defined clinical and functional outcome measures analyzed by different statistical approaches regarding the comparison “Triple always” versus “Triple not always”

From: Effects of triple therapy on disease burden in patients of GOLD groups C and D: results from the observational COPD cohort COSYCONET

Outcome measure

Univariate

Linear Regression

Optimal full matching

IPW

SGRQ activity (score)

Estimate

9.625

3.197

1.835

2.611

95%CI

4.009; 14.685

-0.778; 7.172

-4.811; 8.481

-3.253; 8.476

p value

< 0.001

0.116

0.589

0.384

SGRQ impact (score)

Estimate

1.946

-0.357

-1.811

-0.728

95%CI

-2.753; 6.644

-4.960 4.245

-7.583; 3.961

-5.943; 4.488

p value

0.416

0.879

0.539

0.785

SGRQ symptoms (score)

Estimate

1.291

4.233e-01

-1.933

0.106

95%CI

-2.664; 5.246

-3.589; 4.454

-6.559; 2.692

-4.197; -4.409

p value

0.521

0.833

0.414

0.962

IPAQ (score)

Estimate

-281.629

245.720

-44.08

161.7

95%CI

-1318.311; 755.072

-819.705; 1311.145

-1249.088; 1160.928

-901.404; 1224.804

p value

0.593

0.652

0.943

0.766

6-MWD (m)

Estimate

-40.633

-9.252

-4.144

-8.116

95%CI

68.553; 12.715

-31.672; 13.166

-38.246; 29.958

-40.022; 23.791

p value

0.0045

0.419

0.812

0.619

EQ-VAS (score)

Estimate

-2.787

0.075

1.252

0.590

95%CI

-6.739; 1.163

-3.504; -3.655

-3.948; 6.452

-4.116; 5.295

p value

0.165

0.967

0.637

0.806

FVC /%predicted)

Estimate

-6.811

2.04e-01

-0.262

-0.019

95%CI

-11.048; -2.573

-0.621; 1.029

-11.202; 10.679

-4.616; 4.579

p value

0.002

0.627

0.919

0.994

ITGV (% predicted)

Estimate

9.625

-3.840

-5.342

-5.342

95%CI

0.269; 18.981

-8.352; 0.672

-16.136; 5.452

-16.136; 5.452

p value

0.044

0.097

0.333

0.333

PHQ9 (score)

Estimate

-0.724

-1.123

-1.555

-1.252

95%CI

-1.892; 0.444

-2.267; 0.020

-2.932; -0.179

-2.647; 0.142

p value

0.223

0.055

0.028

0.0796

CAT1 cough (score)

Estimate

-0.133

-0.116

-0.126

-0.097

95%CI

-0.372; 0.106

-0.360; 0.129

-0.418; 0.165

-0.383; 0.189

p value

0.274

0.354

0.396

0.506

CAT2 phlegm (score)

Estimate

0.054

0.008

-0.111

0.017

95%CI

-0.214; 0.322

-0.265; 0.282

-0.430; 0.208

-0.303; 0.337

p value

0.691

0.951

0.495

0.917

CAT3 chest tightness (score)

Estimate

-0.009

-0.129

-0.224

-0.187

95%CI

-0.313; 0.295

-0.429; 0.172

-0.580; 0.132

-0.522; 0.149

p value

0.953

0.402

0.219

0.276

CAT4 breathlessness (score)

Estimate

0.083

-0.077

-0.092

-0.070

95%CI

-0.127; 0.294

-0.262; 0.109

-0.335; -0.151

-0.298; 0.158

p value

0.437

0.418

0.46

0.549

CAT5 activity (score)

Estimate

0.300

0.018

-0.062

0.003

95%CI

-0.007; 0.608

-0.242; 0.278

-0.398; 0.273

-0.309; 0.316

p value

0.055

0.891

0.716

0.984

CAT6 confidence (score)

Estimate

0.254

-0.022

-0.247

-0.104

95%CI

-0.066; 0.574

-0.322; 0.277

-0.665; 0.172

-0.488; 0.281

p value

0.119

0.884

0.249

0.598

CAT7 sleep (score)

Estimate

-0.157

-0.080

-0.261

-0.210

95%CI

-0.495; 0.181

-0.526; 0.154

-0.627; 0.105

-0.552; 0.133

p value

0.361

0.284

0.164

0.231

CAT8 energy (score)

Estimate

-0.092

-0.139

-0.269

-0.150

95%CI

-0.338; 0.153

-0.373; 0.095

-0.545; 0.006

-0.424; 0.125

p value

0.459

0.245

0.056

0.285

CAT Sum (score)

Estimate

0.300

-0.642

-1.393

0.797

95%CI

-1.264;1.864

-2.094; 0.810

-3.276; 0.490

-2.586; 0.992

p value

0.706

0.387

0.148

0.383

PaCO2 (mmHg)

Estimate

0.755

-0.335

-0.109

-0.252

95%CI

-0.422; 1.931

-1.406; 0.735

-1.383; 1.166

-1.456; 0.951

p value

0.207

0.540

0.867

0.681

PaO2 (mmHg)

Estimate

-1.331

-0.380

-0.537

0.359

95%CI

-3.401; 0.739

-2.415; 1.655

-3.067; 1.994

-2.544; 1.827

p value

0.207

0.715

0.678

0.748

SaO2 (%)

Estimate

-0.341

-0.052

0.073

-0.046

95%CI

-0.991; 0.308

-0.693; 0.588

-0.993; 0.846

-0.822; 0.729

p value

0.302

0.873

0.876

0.907

  1. IPW Inverse Probability Weighting. For abbreviations of variables see text. All variables shown were first tested in univariate analyses and then in the multivariable analyses shown in the other three columns, always using the set of predictors shown in Table 1