Comparison of proportional and differential odds models for mixed-effects analysis of categorical data

  • Maria C. KjellssonEmail author
  • Per-Henrik Zingmark
  • E. Niclas Jonsson
  • Mats O. Karlsson


In this work a model for analyzing categorical data is presented; the differential odds model. Unlike the commonly used proportional odds model, this model does not assume that a covariate affects all categories equally on the log odds scale. The differential odds model was compared to the proportional odds model, by assessing statistical significance and improvement of predictive performance when applying the differential odds model to data previously analyzed using the proportional odds model. Three clinical studies; 3-category T-cell receptor density data, 5-category diarrhea data and 6-category sedation data, were re-analyzed with the differential odds model. As expected, no improvements were seen with T-cell receptor density and diarrhea data. However, for the more complex measurement sedation, the differential odds model provided both statistical improvements and improvements in simulation properties. The estimated actual critical value was for all data lower than the nominal value, using the number of added parameters as the degree of freedom, i.e. the differential odds model is statistically indicated to a less extent than expected. The differential odds model had the desired property of not being indicated when not necessary, but it may provide improvements when the data does not represent a categorization of continuous data.


NONMEM Mixed-effects models Pharmacodynamics Categorical data Proportional odds model Differential odds model 


  1. 1.
    Agresti A (2002) Categorical data analysis. Wiley, Hoboken, NJGoogle Scholar
  2. 2.
    Sheiner LB (1994) A new approach to the analysis of analgesic drug trials, illustrated with bromfenac data. Clin Pharmacol Ther 56: 309–322PubMedCrossRefGoogle Scholar
  3. 3.
    Zingmark PH, Edenius C, Karlsson MO (2004) Pharmacokinetic/pharmacodynamic models for the depletion of Vbeta5.2/5.3 T-cells by monoclonal antibody ATM-027 in patients with multiple sclerosis as measured by FACS. Br J Clin Pharmacol 58: 378–389. doi: 10.1111/j.1365-2125.2004.02177.x PubMedCrossRefGoogle Scholar
  4. 4.
    Mandema JW, Stanski DR (1996) Population pharmacodynamic model for ketorolac analgesia. Clin Pharmacol Ther 60: 619–635. doi: 10.1016/S0009-9236(96)90210-6 PubMedCrossRefGoogle Scholar
  5. 5.
    Gupta SK, Sathyan G, Lindemulder EA, Ho PL, Sheiner LB, Aarons L (1999) Quantitative characterization of therapeutic index: application of mixed-effects modeling to evaluate oxybutynin dose-efficacy and dose-side effect relationship. Clin Pharmacol Ther 65: 672–685. doi: 10.1016/S0009-9236(99)90089-9 PubMedCrossRefGoogle Scholar
  6. 6.
    Gomeni R, Teneggi L, Iavarone L, Squassante L, Bye A (2001) Population pharmacokinetic-pharmacodynamic model of craving in an enforced smoking cessation population: indirect response and probabilistic modeling. Pharm Res 18: 537–543. doi: 10.1023/A:1011070814530 PubMedCrossRefGoogle Scholar
  7. 7.
    Lunn DJ, Wakefield J, Racine-Poon A (2001) Cumulative logit models for ordinal data: a case involving allergic rhinitis severity scores. Stat Med 20: 2261–2285. doi: 10.1002/sim.922 PubMedCrossRefGoogle Scholar
  8. 8.
    Mould DR, Chapelsky M, Aluri J, Swagdis J, Samuels R, Granett J (2001) A population pharmacokinetic-pharmacodynamic and logistic regression analysis of lotrafiban in patients. Clin Pharmacol Ther 69: 210–222. doi: 10.1067/mcp.2001.114925 PubMedCrossRefGoogle Scholar
  9. 9.
    Mould DR, Holford NHG, Schellens JHM, Beijner JH, Hutson PR, Rosing H et al (2002) Population pharmacokinetic and adverse event analysis of topotecan in patients with solid tumors. Clin Pharmacol Ther 71: 334–348. doi: 10.1067/mcp.2002.123553 PubMedCrossRefGoogle Scholar
  10. 10.
    Johnston SR, Hickish T, Ellis P, Houston S, Kelland L, Dowsett M et al (2003) Phase II study of the efficacy and tolerability of two dosing regimens of the farnesyl transferase inhibitor, R11577, in advanced breast cancer. J Clin Oncol 21: 2492–2499. doi: 10.1200/JCO.2003.10.064 PubMedCrossRefGoogle Scholar
  11. 11.
    Kowalski KG, McFadyen L, Hutmacher MM, Frame B, Miller R (2003) A two-part mixture model for longitudinal adverse event severity data. J Pharmacokinet Pharmacodyn 30: 315–336. doi: 10.1023/B:JOPA.0000008157.26321.3c PubMedCrossRefGoogle Scholar
  12. 12.
    Cullberg M, Eriksson UG, Wåhlander K, Eriksson H, Shulman S, Karlsson MO (2005) Pharmacokinetics of ximelagatran and relationships to clinical response in acute deep vein thrombosis. Clin Pharmacol Ther 77: 279–290. doi: 10.1016/j.clpt.2004.11.001 PubMedCrossRefGoogle Scholar
  13. 13.
    Xie R, Mathijssen RHJ, Sparreboom A, Verweij J, Karlsson MO (2002) Clinical pharmacokinetics of irinotecan and its metabolites in relation with diarrhea. Clin Pharmacol Ther 72: 265–275. doi: 10.1067/mcp.2002.126741 PubMedCrossRefGoogle Scholar
  14. 14.
    Zingmark PH, Ekblom M, Odergren T, Ashwood T, Lyden P, Karlsson MO et al (2003) Population pharmacokinetics of clomethiazole and its effect on the natural course of sedation in acute stroke patients. Br J Clin Pharmacol 56: 173–183. doi: 10.1046/j.0306-5251.2003.01850.x PubMedCrossRefGoogle Scholar
  15. 15.
    Knibbe CAJ, Zuideveld KP, DeJongh J, Kuks PFM, Aarts LPHJ, Danhof M (2002) Population pharmacokinetic and pharmacodynamic modeling of propofol for long-term sedation in critically ill patents: a comparison between propofol 6% and propofol 1%. Clin Pharmacol Ther 72: 670–684. doi: 10.1067/mcp.2002.129500 PubMedCrossRefGoogle Scholar
  16. 16.
    Olofsen E, Romberg R, Bijl H, Mooren R, Engbers F, Kest B et al (2005) Alfentanil and placebo analgesia. No sex difference detected in models of experimental pain. Anesthesiology 103: 130–139. doi: 10.1097/00000542-200507000-00020 Google Scholar
  17. 17.
    Gibiansky E, Gibiansky L (2007) Population PK/PD model of GPI 15715 and GPI 15715-derived propofol in sedation and comparison of PK/PD models for ordered categorical observations. In: Abstracts of the annual meeting of the population approach group in Europe. Accessed 10/03/07
  18. 18.
    Armstrong BG, Sloan M (1989) Ordinal regression models for epidemiologic data. Am J Epidemiol 129: 191–204PubMedGoogle Scholar
  19. 19.
    Gisleskog PO, Karlsson MO, Beal SL (2002) Use of prior information to stabilize a population data analysis. J Pharmacokinet Pharmacodyn 29: 473–505. doi: 10.1023/A:1022972420004 PubMedCrossRefGoogle Scholar
  20. 20.
    Beal SL, Sheiner LB (eds) (1989–1998) Users guides, NONMEM Icon Development Solutions, Elliot City, MDGoogle Scholar
  21. 21.
    Stram DO, Lee JW (1994) Variance components testing in the longitudinal mixed effects model. Biometrics 50: 1171–1177. doi: 10.2307/2533455 PubMedCrossRefGoogle Scholar
  22. 22.
    Pinheiro JC, Bates DM (2000) Mixed-effects models in S and S-plus. In: (eds) Statistics and computing. Springer, New YorkGoogle Scholar
  23. 23.
    Wählby U, Bouw MR, Jonsson EN, Karlsson MO (2002) Assessment of type I error rates for the statistical sub-model in NONMEM. J Pharmacokinet Pharmacodyn 29: 251–269. doi: 10.1023/A:1020254823597 PubMedCrossRefGoogle Scholar
  24. 24.
    Efron B (1979) Bootstrap method: another look at the Jackknife. Ann Stat 7: 1–26. doi: 10.1214/aos/1176344552 CrossRefGoogle Scholar
  25. 25.
    Scott SC, Goldberg MS, Mayo NE (1997) Statistical assessment of ordinal outcomes in comparative studies. J Clin Epidemiol 50: 45–55. doi: 10.1016/S0895-4356(96)00312-5 PubMedCrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media, LLC 2008

Authors and Affiliations

  • Maria C. Kjellsson
    • 1
    Email author
  • Per-Henrik Zingmark
    • 2
  • E. Niclas Jonsson
    • 1
    • 3
  • Mats O. Karlsson
    • 1
  1. 1.Division of Pharmacokinetics and Drug Therapy, Department of Pharmaceutical BiosciencesUppsala UniversityUppsalaSweden
  2. 2.Clinical Pharmacology & DMPKAstraZeneca R&DSödertäljeSweden
  3. 3.Exprimo NVMechelenBelgium

Personalised recommendations