Statistics and Computing

, Volume 21, Issue 4, pp 483–499 | Cite as

ROC curve and covariates: extending induced methodology to the non-parametric framework

  • María Xosé Rodríguez-Álvarez
  • Javier Roca-Pardiñas
  • Carmen Cadarso-Suárez
Article

Abstract

Continuous diagnostic tests are often used to discriminate between diseased and healthy populations. The receiver operating characteristic (ROC) curve is a widely used tool that provides a graphical visualisation of the effectiveness of such tests. The potential performance of the tests in terms of distinguishing diseased from healthy people may be strongly influenced by covariates, and a variety of regression methods for adjusting ROC curves has been developed. Until now, these methodologies have assumed that covariate effects have parametric forms, but in this paper we extend the induced methodology by allowing for arbitrary non-parametric effects of a continuous covariate. To this end, local polynomial kernel smoothers are used in the estimation procedure. Our method allows for covariate effect not only on the mean, but also on the variance of the diagnostic test. We also present a bootstrap-based method for testing for a significant covariate effect on the ROC curve. To illustrate the method, endocrine data were analysed with the aim of assessing the performance of anthropometry for predicting clusters of cardiovascular risk factors in an adult population in Galicia (NW Spain), duly adjusted for age. The proposed methodology has proved useful for providing age-specific thresholds for anthropometric measures in the Galician community.

Keywords

ROC curve Non-parametric regression Bootstrap Cardiovascular risk factors Anthropometric measures 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. Alonzo, T.A., Pepe, M.S.: Distribution-free ROC analysis using binary regression techniques. Biostatistics 3, 421–432 (2002) MATHCrossRefGoogle Scholar
  2. Cai, T.: Semi-parametric ROC regression analysis with placement values. Biostatistics 5, 45–60 (2004) MATHCrossRefGoogle Scholar
  3. Cai, T., Pepe, M.S.: Semiparametric receiver operating characteristic analysis to evaluate biomarkers for disease. J. Am. Stat. Assoc. 97, 1099–1107 (2002) MATHCrossRefMathSciNetGoogle Scholar
  4. Carey, V.J., Walters, E.E., Colditz, G.A., Solomon, C.G., Willet, W.C., Rosner, B.A., Speizer, F.E., Manson, J.E.: Body fat distribution and risk of noninsulin-dependent diabetes in women: the Nurses’ Health Study. Am. J. Epidemiol. 145, 614–619 (1997) Google Scholar
  5. Chumlea, W.C., Baumgartner, R.N., Garry, P.J., Rhyne, R.L., Nicholson, C., Wayne, S.: Fat distribution and blood lipids in a sample of healthy elderly people. Int. J. Obes. 16, 125–133 (1992) Google Scholar
  6. de Boor, C.A.: A Practical Guide to Splines. Springer, New York (2001). Revised Edition MATHGoogle Scholar
  7. Efron, B., Tibshirani, R.J.: An Introduction to the Bootstrap. Chapman & Hall, New York (1993) MATHGoogle Scholar
  8. Eilers, P.H.C., Marx, B.D.: Flexible smoothing with B-splines and penalties. Stat. Sci. 11, 89–121 (1996) MATHCrossRefMathSciNetGoogle Scholar
  9. Fan, J., Gijbels, I.: Local Polynomial Modelling and Its Applications. Chapman & Hall, CRC (1996) Google Scholar
  10. Fan, J., Marron, J.S.: Fast implementation of non-parametric curve estimators. J. Comput. Graph. Stat. 3, 35–56 (1994) CrossRefGoogle Scholar
  11. Fan, J., Yao, Q.: Efficient estimation of conditional variance functions in stochastic regression. Biometrika 85, 645–660 (1998) MATHCrossRefMathSciNetGoogle Scholar
  12. Faraggi, D.: Adjusting receiver operating characteristic curves and related indices for covariates. Statistician 52, 179–192 (2003) MathSciNetGoogle Scholar
  13. González-Manteiga, W., Pardo Fernández, J.C., Van Keilegom, I.: ROC curves in nonparametric location-scale regression models. Scand. J. Stat. (2010). doi:10.1111/j.1467-9469.2010.00693x Google Scholar
  14. Haslam, D.W., James, W.P.T.: Obesity. Lancet 366, 1197–1209 (2005) CrossRefGoogle Scholar
  15. Hastie, T.J., Tibshirani, R.J.: Generalized Additive Models. Chapman & Hall, London (1990) MATHGoogle Scholar
  16. Hsieh, F., Turnbull, B.W.: Nonparametric and semiparametric estimation of the receiver operating characteristic curve. Ann. Stat. 24, 24–40 (1996) MathSciNetGoogle Scholar
  17. Hu, G., Qiao, Q., Tuomilehto, J., Eliasson, M., Feskens, E.J., Pyörälä, K.: Plasma insulin and cardiovascular mortality in non-diabetic European men and women: a meta-analysis of data from eleven prospective studies. The DECODE Insulin Study Group. Diabetologia 47, 1245–1256 (2004) Google Scholar
  18. International Diabetes Federation: The IDF consensus worldwide definition of the metabolic syndrome. http://www.idf.org/webdata/docs/IDF_Meta_def_final.pdf (2008). Accessed 7 September 2009
  19. Janes, H., Pepe, M.S.: Adjusting for covariate effects on classification accuracy using the covariate-adjusted ROC curve. Biometrika 96, 371–382 (2009) MATHCrossRefMathSciNetGoogle Scholar
  20. Karelis, A.D., St-Pierre, D.H., Conus, F., Rabasa-Lhoret, R., Poehlman, E.T.: Metabolic and body composition factors in subgroups of obesity: What do we know? J. Clin. Endocrinol. Metab. 89, 2569–2575 (2004) CrossRefGoogle Scholar
  21. Lloyd, C.J.: Using smooth receiver operating characteristic curves to summarize and compare diagnostic systems. J. Am. Stat. Assoc. 93, 1356–1364 (1998) CrossRefGoogle Scholar
  22. López-de-Ullibarri, I., Cao, R., Cadarso-Suárez, C., Lado, M.J.: Nonparametric estimation of conditional ROC curves: application to discrimination tasks in computerized detection of early breast cancer. Comput. Stat. Data Anal. 52, 2623–2631 (2008) MATHCrossRefGoogle Scholar
  23. MacIntosh, M.W., Pepe, M.S.: Combining several screening test: optimality of the risk score. Biometrics 58, 657–664 (2002) CrossRefMathSciNetGoogle Scholar
  24. Metz, C.E.: Basic principles of ROC analysis. Semin. Nucl. Med. 8, 183–298 (1978) CrossRefGoogle Scholar
  25. Mirmiran, P., Esmaillzadeh, A., Azizi, F.: Detection of cardiovascular risk factors by anthropometric measures in Tehranian adults: receiver operating characteristic (ROC) curve analysis. Eur. J. Clin. Nutr. 58, 1110–1118 (2004) CrossRefGoogle Scholar
  26. Nadaraya, E.A.: On estimating regression. Theory Probab. Appl. 9, 141–142 (1964) CrossRefGoogle Scholar
  27. Peden, M., McGee, K., Krug, E. (eds.): Injury: A Leading Cause of the Global Burden of Disease, 2000. World Health Organization, Geneva (2006) Google Scholar
  28. Peng, L., Zhou, X.H.: Local linear smoothing of receiver operating characteristic (ROC) curves. J. Stat. Plan. Inference 118, 129–143 (2004) MATHCrossRefMathSciNetGoogle Scholar
  29. Pepe, M.S.: Three approaches to regression analysis of receiver operating characteristic curves for continuous test results. Biometrics 54, 124–135 (1998) MATHCrossRefGoogle Scholar
  30. Pepe, M.S.: The Statistical Evaluation of Medical Tests for Classification and Prediction. Oxford University Press, New York (2003) MATHGoogle Scholar
  31. Ren, H., Zhou, X.H., Liang, H.: A flexible method for estimating the ROC curve. J. Appl. Stat. 31, 773–784 (2004) MATHCrossRefMathSciNetGoogle Scholar
  32. Ruppert, D., Wand, M.P., Holst, U., Hössjer, O.: Local polinomial variance function estimation. Technometrics 39, 262–273 (1997) MATHCrossRefMathSciNetGoogle Scholar
  33. Schisterman, E.F., Faraggi, D., Reiser, B.: Adjusting the generalized ROC curve for covariates. Stat. Med. 23, 3319–3331 (2004) CrossRefGoogle Scholar
  34. Su, J.Q., Liu, J.S.: Linear combinations of multiple diagnostic markers. J. Am. Stat. Assoc. 88, 1350–1355 (1993) MATHCrossRefMathSciNetGoogle Scholar
  35. Swets, J.A., Pickett, R.M.: Evaluation of Diagnostic Systems: Methods from Signal Detection Theory. Academic Press, New York (1982) Google Scholar
  36. Tomé, M.A., Botana, M.A., Cadarso-Suárez, C., Rego-Iraeta, A., Fernández-Mariño, A., Mato, J.A., Solache, I., Perez-Fernandez, R.: Prevalence of metabolic syndrome in Galicia (NW Spain) on four alternative definitions and association with insulin resistance. J. Endocrinol. Invest. 32, 505–511 (2008) Google Scholar
  37. Tosteson, A.N., Begg, C.B.: A general regression methodology for ROC curve estimation. Med. Decis. Mak. 8, 204–215 (1988) CrossRefGoogle Scholar
  38. Wand, M.P, Jones, M.C.: Kernel Smoothing. Chapman & Hall, London (1995) MATHGoogle Scholar
  39. Watson, G.S.: Smooth regression analysis. Sankhyā, Ser. A 26, 359–372 (1964) MATHGoogle Scholar
  40. Zheng, Y., Heagerty, P.J.: Semiparametric estimation of time-dependent ROC curves for longitudinal marker data. Biostatistics 4, 615–632 (2004) CrossRefGoogle Scholar
  41. Zhou, X.H., Harezlak, J.: Comparison of bandwidth selection methods for kernel smoothing of ROC curves. Stat. Med. 21, 2045–2055 (2002) CrossRefGoogle Scholar
  42. Zhou, X.H., Obuchowski, N.A., McClish, D.K.: Statistical Methods in Diagnostic Medicine. Wiley, New York (2002) MATHCrossRefGoogle Scholar
  43. Zimmet, P., Alberti, K.G.M.M., Shaw, J.: Global and societal implications of the diabetes epidemic. Nature. 414, 782–787 (2001) CrossRefGoogle Scholar
  44. Zou, K.H., Hall, W.J., Shapiro, D.E.: Smooth nonparametric receiver operating characteristic (ROC) curves for continuous diagnostic test. Stat. Med. 16, 2143–2165 (1997) CrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media, LLC 2010

Authors and Affiliations

  • María Xosé Rodríguez-Álvarez
    • 1
    • 2
  • Javier Roca-Pardiñas
    • 3
  • Carmen Cadarso-Suárez
    • 1
    • 2
  1. 1.Unit of Biostatistics, Dept. of Statistics and Operations Research, Faculty of MedicineUniversity of Santiago de CompostelaSantiago de CompostelaSpain
  2. 2.Instituto de Investigación Sanitaria de Santiago (IDIS)Santiago de CompostelaSpain
  3. 3.Dept. of Statistics and Operational ResearchUniversity of VigoVigoSpain

Personalised recommendations