Loading [MathJax]/jax/output/PreviewHTML/jax.js
Skip to main page content
U.S. flag

An official website of the United States government

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2015 Dec;71(4):1139-49.
doi: 10.1111/biom.12344. Epub 2015 Jul 20.

Assessing incremental value of biomarkers with multi-phase nested case-control studies

Qian M Zhou  1 , Yingye Zheng  2 , Lori B Chibnik  3 , Elizabeth W Karlson  3 , Tianxi Cai  4
Affiliations

Assessing incremental value of biomarkers with multi-phase nested case-control studies

Qian M Zhou et al. Biometrics. 2015 Dec.

Abstract

Accurate risk prediction models are needed to identify different risk groups for individualized prevention and treatment strategies. In the Nurses' Health Study, to examine the effects of several biomarkers and genetic markers on the risk of rheumatoid arthritis (RA), a three-phase nested case-control (NCC) design was conducted, in which two sequential NCC subcohorts were formed with one nested within the other, and one set of new markers measured on each of the subcohorts. One objective of the study is to evaluate clinical values of novel biomarkers in improving upon existing risk models because of potential cost associated with assaying biomarkers. In this paper, we develop robust statistical procedures for constructing risk prediction models for RA and estimating the incremental value (IncV) of new markers based on three-phase NCC studies. Our method also takes into account possible time-varying effects of biomarkers in risk modeling, which allows us to more robustly assess the biomarker utility and address the question of whether a marker is better suited for short-term or long-term risk prediction. The proposed procedures are shown to perform well in finite samples via simulation studies.

Keywords: Incremental value; Inverse probability weighting; Nested case-control study; Rheumatoid arthritis; Risk prediction; Time dependent ROC curve analysis.

PubMed Disclaimer

Figures

Figure 1
Figure 1
Sequential NCC sampling design in NHS.
Figure 2
Figure 2
Analysis of NHS RA Data: estimated regression coefficients (circles) and their 95% confidence intervals (vertical bars) of (a) age (b) pack-years of smoking (c) cumulative alcohol intake (d) sTNFRII (e) IL-6 (f) GRS under the t0-year GLM with (t0 = 5, 6, 8, 10, 12, 14, 15). Shown also are the estimates from the Cox PH model with ZG+B.
Figure 3
Figure 3
Analysis of the NHS RA data: cross-validated point estimates (dots) along with 95% confidence intervals (vertical bars) of the IncV by adding {B, G} on top of Z with respect to AUCt0, TPRt0, PPVt0 and NPVt0 at the cut-off values chosen to achieve FPRt0 = 0.1 for t0 = 5, 6, 8, 10, 12, 14, 15. The solid and dashed lines represent the IncV estimates corresponding to risk modeling with time-specific GLMs and PH models, respectively.
Figure 4
Figure 4
Simulation Results: average point estimates (circles) of the regression coefficients for marker Z, B and G in the full time-specific GLM along with their empirical 95% confidence intervals (solid vertical bars) at t0 = 5, 6, 7, 8, 9, 10. Shown also are the regression coefficient estimates along with their 95% confidence intervals from the PH model (diamond with dashed vertical bars) and from the CLR (triangle with dotted vertical bars). The dotdashed line represents the true values of the regression coefficients.
Figure 5
Figure 5
Simulation Results: average DIPW estimates (circles: time-specific GLM, diamonds: PH model) of the time-dependent AUC for the model with ZG+B and the IncV in AUC of G on top of ZB along with their empirical 95% confidence intervals (solid bars: time-specific GLM, dashed bars: PH model) for t0 = 5, 6, 7, 8, 9, 10. Shown also are the naive estimates obtained from using the CLR to estimate the log hazard ratios but treats the data as standard case-control study to perform ROC analyses.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

References

    1. Anderson WF, Rosenberg PS, Menashe I, Mitani A, Pfeiffer RM. Age-related crossover in breast cancer incidence rates between black and white ethnic groups. Journal of the National Cancer Institute. 2008;100:1804–1814. - PMC - PubMed
    1. Boekholdt SM, Hack CE, Sandhu MS, Luben R, Bingham SA, Wareham NJ, Peters RJ, Jukema JW, Day NE, Kastelein JJ, Khaw K-T. C-reactive protein levels and coronary artery disease incidence and mortality in apparently healthy men and women: the epic-norfolk prospective population study 1993-2003. Atherosclerosis. 2006;187:415–422. - PubMed
    1. Cai T, Zheng Y. Nonparametric evaluation of biomarker accuracy under nested case-control studies. Journal of the American Statistical Association. 2011;106:569–580. - PMC - PubMed
    1. Cai T, Zheng Y. Evaluating prognostic accuracy of biomarkers under nested case-control studies. Biostatistics. 2012;13:89–100. - PMC - PubMed
    1. Cai T, Zheng Y. Resampling procedures for making inference under nested case–control studies. Journal of the American Statistical Association. 2013;108:1532–1544. - PMC - PubMed

Publication types