Initial severity‐dependent longitudinal model with application to a randomized controlled trial of women with depression Academic Article uri icon


  • The assessment of initial severity of a disease is arguably one of the most important factors in identifying appropriate therapies. In this paper, we propose an initial severity-dependent longitudinal model to account for the influence of the initial severity of a disease on the posttreatment severity and the efficacy of medical treatments. The proposed model has the flexibility of nonparametric modeling, as it allows coefficients to vary with the initial severity of the disease. It also provides attractive and practical patient-specific interpretation of initial severity-dependent coefficients. As a result, the proposed model enables patient-specific modeling and treatment recommendations consistent with the assessment of the patient's initial severity, and thus, it can be used as a decision support tool for clinicians. A new empirical likelihood approach is employed for efficient estimation and statistical inference about the initial severity-dependent coefficients. In contrast to the literature on marginal regression models, the proposed estimation procedure allows nuisance parameters associated with the working correlation matrix and the error variances to vary smoothly with the initial severity. The effectiveness of the proposed procedure is demonstrated via simulation studies. We further apply the proposed method by analyzing a data set arising from a randomized controlled trial of women with depression and discover an interesting phenomenon; antidepressant medication intervention is effective for patients with moderate or severe depression, whereas psychotherapy intervention using manual-guided cognitive behavior therapy is effective for patients with a severe case of depression.

author list (cited authors)

  • Kim, S., Cho, H. R., & Zhang, X.

citation count

  • 1

complete list of authors

  • Kim, Seonjin||Cho, Hyunkeun Ryan||Zhang, Xianyang

publication date

  • April 2019