Skip to main content

Table 4 Predictive Accuracies of the Prognostic Models

From: Inflammation-scores as prognostic markers of overall survival in lung cancer: a register-based study of 6,210 Danish lung cancer patients

 

NSCLC

 

SCLC

Model

AIC

C-index

AIC

C-index

NLRa

65,397

0.594

9,622

0.559

NLRb

65,286

0.612

9,600

0.575

mGPS

65,175

0.635

9,602

0.588

PLRc

65,722

0.556

9,664

0.506

PLRd

65,572

0.579

9,657

0.522

CNG

65,055

0.646

9,599

0.589

ACBS

64,952

0.651

9,604

0.594

HALP

65,917

0.502

9,665

0.501

LMR

65,529

0.576

9,647

0.547

MLR

65,630

0.574

9,653

0.543

SII

65,786

0.533

9,665

0.504

Stage + age + sex + PS + smoking

52,232

0.769

7,743

0.730

NLRa + stage + age + sex + PS + smoking

52,118

0.775

7,732

0.731

NLRb + stage + age + sex + PS + smoking

52,065

0.777

7,717

0.736

mGPS + stage + age + sex + PS + smoking

52,100

0.776

7,718

0.737

PLRc + stage + age + sex + PS + smoking

52,201

0.770

7,744

0.730

PLRd + stage + age + sex + PS + smoking

52,169

0.771

7,739

0.731

CNG + stage + age + sex + PS + smoking

52,075

0.778

7,717

0.737

ACBS + stage + age + sex + PS + smoking

52,019

0.779

7,720

0.737

HALP + stage + age + sex + PS + smoking

52,230

0.769

7,745

0.730

LMR + stage + age + sex + PS + smoking

52,105

0.771

7,737

0.732

MLR + stage + age + sex + PS + smoking

52,182

0.771

7,738

0.732

SII + stage + age + sex + PS + smoking

52,210

0.770

7,744

0.730

  1. ACBS Aarhus composite biomarker score, AIC Akaikes information criterion, CNG The Combined NLR and Glasgow prognostic score, HALP haemoglobin × albumin × lymphocytes/platelet count, LMR lymphocyte to monocyte ratio, mGPS modified Glasgow prognostic score, MLR Monocyte to lymphocyte ratio, NLR Neutrophil to lymphocyte ratio, NSCLC non-small cell lung cancer, PLR thrombocyte to lymphocyte ratio, SCLC Small cell lung cancer, SII The systemic inflammation indexa cut point ≥ 3; b cut point ≥ 4; c cut point ≥ 150; d cut point ≥ 150 AIC estimates the quality of each model relative to the other models. The values are arbitrary. The model with the minimum, AIC is the model with the optimal fit of data. ** c-index: Harrell's concordance index. C-index gives a measure of goodness of fit for the model. Values range between 0.5 – 1.0; 1.0 is the perfect fit