Adjusting for nonignorable drop-out using semiparametric nonresponse models
Adjusting for nonignorable drop-out using semiparametric nonresponse models
复制标题
DOI：
10.2307/2669923
复制
发表时间：
1999-12-01
影响因子：
3.7
通讯作者：
Robins, JM
中科院分区：
文献类型：
--
作者：
Scharfstein, DO;Rotnitzky, A;Robins, JM
考虑一项研究，其设计要求从入组（时间t = 0）到时间t = T对研究受试者进行随访，在时间t = T时测量关注的主要终点Y。研究设计还要求在间隔（0，T）内的一个或多个时间t测量协变量向量V（t）。我们感兴趣的是，当一些受试者在常见的固定随访结束时间T之前的随机时间Q退出研究时，对Y的边际平均mu（0）进行推断。这篇文章的目的是展示如何进行推断mu（0）时，连续辍学时间Q是半参数建模，没有限制的结果和其他测量变量的联合分布。特别地，我们考虑了两个模型的条件风险下降我们给定的（（V）在酒吧（T），Y），其中（V）在酒吧（t）表示的历史过程V（t）通过时间t，t是（0，T）的一个元素。在第一个模型中，我们假设lambda（Q）（t\（V）over bar（T），Y）= lambda（0）（t\（V）over bar（t））exp（alpha（0）Y），其中alpha（0）是标量参数，lambda（0）（t\（V）over bar（t））是t和过程（V）over bar（t）的无限制正函数。当过程（V）在bar（t）上是高维的时，由于维数灾难，该模型中的估计在中等样本量下是不可行的。对于这种情况，我们考虑第二个模型，该模型施加了额外的限制，即lambda（0）（t\（V）over bar（t））= lambda（0）（t）exp（gamma（0）'W（t），其中lambda（0）（t）是未指定的基线风险函数，W（t）= w（t，（V）over bar（t）），w（.，.）是将（bar（t）上的ti（V））映射到R-q的已知函数，并且gamma（0）'是q × 1未知参数向量。当alpha（0）不等于0时，drop-out是不可忽略的。由于可识别性问题，Y的平均值mu（0）和选择偏差参数ao的联合估计可能是困难的或不可能的。因此，我们建议将选择偏差参数α（0）视为已知，而不是从数据中估计。然后，我们进行敏感性分析，以了解当我们在一个合理的值范围内改变alpha（0）时，关于mu（0）的推断如何变化。我们应用我们的方法来分析ACTG 175，艾滋病临床试验。
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 mu(0) of Y when some subjects drop out of the study at random times Q prior to the common fixed end of follow-up rime T. The purpose of this article is to show how to make inferences about mu(0) 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-our given ((V) over bar(T), Y), where (V) over bar(t) denotes the history of the process V(t) through time t, t is an element of (0,T). In the first model, we assume that lambda(Q)(t\(V) over bar(T), Y) = lambda(0)(t\(V) over bar(t)) exp(alpha(0)Y), where alpha(0) is a scalar parameter and lambda(0)(t\(V) over bar(t)) is an unrestricted positive function of t and the process (V) over bar(t). When the process (V) over bar(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\(V) over bar(t)) = lambda(0)(t) exp(gamma(0)'W(t)), where lambda(0)(t) is an unspecified baseline hazard function, W(t) = w(t, (V) over bar(t)), w(.,.) is a known function that maps (t, (V) over bar(t)) to R-q, and gamma(0)' is a q x 1 unknown parameter vector. When alpha(0) not equal 0, then drop-out is nonignorable. On account of identifiability problems, joint estimation of the mean mu(0) Of Y and the selection bias parameter ao may be difficult or impossible. Therefore, we propose regarding the selection bias parameter alpha(0) as known, rather than estimating it from the data. We then perform a sensitivity analysis to see how inference about mu(0) changes as we vary alpha(0) over a plausible range of values. We apply our approach to the analysis of ACTG 175, an AIDS clinical trial.