Access

You are not currently logged in.

Access your personal account or get JSTOR access through your library or other institution:

login

Log in to your personal account or through your institution.

If you need an accessible version of this item please contact JSTOR User Support

Adjusting for Nonignorable Drop-Out Using Semiparametric Nonresponse Models

Daniel O. Scharfstein, Andrea Rotnitzky and James M. Robins
Journal of the American Statistical Association
Vol. 94, No. 448 (Dec., 1999), pp. 1096-1120
DOI: 10.2307/2669923
Stable URL: http://www.jstor.org/stable/2669923
Page Count: 25
  • Download ($14.00)
  • Cite this Item
If you need an accessible version of this item please contact JSTOR User Support
Adjusting for Nonignorable Drop-Out Using Semiparametric Nonresponse Models
Preview not available

Abstract

Consider a study whose design calls for the study subjects to be followed from enrollment (time t = 0) to time t = T, at which point a primary endpoint of interest Y is to be measured. The design of the study also calls for measurements on a vector V(t) of covariates to be made at one or more times t during the interval [0, T). We are interested in making inferences about the marginal mean μ0 of Y when some subjects drop out of the study at random times Q prior to the common fixed end of follow-up time T. The purpose of this article is to show how to make inferences about μ0 when the continuous drop-out time Q is modeled semiparametrically and no restrictions are placed on the joint distribution of the outcome and other measured variables. In particular, we consider two models for the conditional hazard of drop-out given (V̄(T), Y), where V̄(t) denotes the history of the process V(t) through time t, t ∈ [0, T). In the first model, we assume that $\lambda_Q(t\mid\bar V(T), Y) = \lambda_0(t\mid\bar V(t))$ exp(α0 Y), where α0 is a scalar parameter and $\lambda_0(t\mid\bar V(t))$ is an unrestricted positive function of t and the process $\bar V(t)$. When the process $\bar V(t)$ is high dimensional, estimation in this model is not feasible with moderate sample sizes, due to the curse of dimensionality. For such situations, we consider a second model that imposes the additional restriction that $\lambda_0(t\mid\bar V(t)) = \lambda_0(t)$ exp(γ'0 W(t)), where λ0(t) is an unspecified baseline hazard function, W(t) = w(t, V̄(t)), w(·, ·) is a known function that maps (t, V̄(t)) to Rq, and γ0 is a q × 1 unknown parameter vector. When α0 ≠ 0, then drop-out is nonignorable. On account of identifiability problems, joint estimation of the mean μ0 of Y and the selection bias parameter α0 may be difficult or impossible. Therefore, we propose regarding the selection bias parameter α0 as known, rather than estimating it from the data. We then perform a sensitivity analysis to see how inference about μ0 changes as we vary α0 over a plausible range of values. We apply our approach to the analysis of ACTG 175, an AIDS clinical trial.

Page Thumbnails

  • Thumbnail: Page 
1096
    1096
  • Thumbnail: Page 
1097
    1097
  • Thumbnail: Page 
1098
    1098
  • Thumbnail: Page 
1099
    1099
  • Thumbnail: Page 
1100
    1100
  • Thumbnail: Page 
1101
    1101
  • Thumbnail: Page 
1102
    1102
  • Thumbnail: Page 
1103
    1103
  • Thumbnail: Page 
1104
    1104
  • Thumbnail: Page 
1105
    1105
  • Thumbnail: Page 
1106
    1106
  • Thumbnail: Page 
1107
    1107
  • Thumbnail: Page 
1108
    1108
  • Thumbnail: Page 
1109
    1109
  • Thumbnail: Page 
1110
    1110
  • Thumbnail: Page 
1111
    1111
  • Thumbnail: Page 
1112
    1112
  • Thumbnail: Page 
1113
    1113
  • Thumbnail: Page 
1114
    1114
  • Thumbnail: Page 
1115
    1115
  • Thumbnail: Page 
1116
    1116
  • Thumbnail: Page 
1117
    1117
  • Thumbnail: Page 
1118
    1118
  • Thumbnail: Page 
1119
    1119
  • Thumbnail: Page 
1120
    1120