Future Cases as Present Controls to Adjust for Exposure Trend Bias in Case-only Studies

Department of Community Health-Epidemiology, Brown University, Providence, RI, USA.
Epidemiology (Cambridge, Mass.) (Impact Factor: 6.2). 07/2011; 22(4):568-74. DOI: 10.1097/EDE.0b013e31821d09cd
Source: PubMed


Self-matched case-only studies (such as the case-crossover or self-controlled case-series method) control by design for time-invariant confounders (measured or unmeasured), but they do not control for confounders that vary with time. A bidirectional case-crossover design can be used to adjust for exposure-time trends. In pharmacoepidemiology, however, illness often influences future use of medications, making a bidirectional design problematic. Suissa's case-time-control design combines a case-crossover and case-control design, and adjusts for exposure-trend bias in the cases' self-controlled odds ratio by dividing that ratio by the corresponding self-controlled odds ratio in a concurrent matched control group. However, if not well matched, the control group may reintroduce selection bias. We propose a "case-case-time-control" that involves crossover analyses in cases and future-case controls. This person-time sampling strategy improves matching by restricting controls to future cases. We evaluate the proposed study design through simulations and analysis of a theoretically null relationship using Veterans Administration (VA) data. Simulation studies show that the case-case-time-control can adjust for exposure trends while controlling for time-invariant confounders. Use of an inappropriate control group left case-time-control analyses biased by exposure-time trends. When analyzing the relationship between vitamin exposure and stroke, using data on 3192 patients in the VA system, a case-crossover odds ratio of 1.5 (95% confidence interval = 1.3-1.7) was reduced to 1.1 (0.9-1.3) when divided by the concurrent exposure trend odds ratio (1.4) in matched future cases. This applied example demonstrates how our approach can adjust for exposure trends observed across time axes.

Download full-text


Available from: David Dore
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: The conditional logistic regression model is the standard tool for the analysis of epidemiological studies in which one or more cases (the event of interest), are individually matched with one or more controls (not showing the event). These situations arise, for example, in matched case–control studies and self–matched case–only studies (such as the case–crossover [1], the case–time–control [2] or the case–case–time–control [3] designs). Usually, odds ratios are estimated by maximizing the conditional log–likelihood function and variable selection is performed by conventional manual or automatic selection procedures, such as stepwise. These techniques are, however, unsatisfactory in sparse, high-dimensional settings in which penalized methods, such as the lasso (least absolute shrinkage and selection operator) [4], have emerged as an alternative. In particular, the lasso and related methods have recently been adapted to conditional logistic regression [5]. The R package clogitLasso implements, for small to moderate sized samples (less than 3, 000 observations), the algorithms discussed in [5], based on the stratified discrete-time Cox propor-tional hazards model and depending on the penalized package [6]. For large datasets, clogit-Lasso computes the highly efficient procedures proposed in [7, 8], based on an IRLS (iteratively reweighted least squares) algorithm [9] and depending on the lassoshooting package [10]. The most common situations that involve 1:1, 1:M and N:M matching are available.
    Preview · Article · Jul 2012
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: To assess case-only designs for surveillance with administrative databases. We reviewed literature on two designs that are observational analogs to crossover experiments: the self-controlled case series (SCCS) and the case-crossover (CCO) design. SCCS views the 'experiment' prospectively, comparing outcome risks in windows with different exposures. CCO retrospectively compares exposure frequencies in case and control windows. The main strength of case-only designs is they entail self-controlled analyses that eliminate confounding and selection bias by time-invariant characteristics not recorded in healthcare databases. They also protect privacy and are computationally efficient, as they require fewer subjects and variables. They are better than cohort designs for investigating transient effects of accurately recorded preventive agents, for example, vaccines. They are problematic if timing of self-administration is sporadic and dissociated from dispensing times, for example, analgesics. They tend to have less exposure misclassification bias and time-varying confounding if exposures are brief. Standard SCCS designs are bidirectional (using time both before and after the first exposure event), so they are more susceptible than CCOs to reverse-causality bias, including immortal-time bias. This is true also for sequence symmetry analysis, a simplified SCCS. Unidirectional CCOs use only time before the outcome, so they are less affected by reverse causality but susceptible to exposure-trend bias. Modifications of SCCS and CCO partially deal with these biases. The head-to-head comparison of multiple products helps to control residual biases. The case-only analyses of intermittent users complement the cohort analyses of prolonged users because their different biases compensate for one another.
    Full-text · Article · Jan 2012 · Pharmacoepidemiology and Drug Safety
  • [Show abstract] [Hide abstract]
    ABSTRACT: Time-dependent effects of antipsychotics on risk of stroke and potential effect modification by age have not been fully investigated. A case-case-time-control design uses within- and between-case comparisons to evaluate short-term effects at the same time as adjusting for unmeasured time-invariant confounders and exposure-time trends. We conducted a case-case-time-control design study using data from the Veterans Health Administration. Veterans with inpatient hospitalizations for ischemic stroke between 2002 and 2007 were included. For every stroke case, the "current" exposure period was defined as 1 to 30 days before hospitalization and the "reference" period as 91 to 120 days before hospitalization. Exposure during the current period was compared with exposure during the reference period within cases. Exposure-time trend-adjusted estimates of the effect of antipsychotic exposure on risk of stroke were obtained by dividing exposure odds for antipsychotic exposure by average exposure odds for other medications over the same time period among the same cases. After adjusting for exposure-time trends, odds of stroke were 1.8 (95% CI, (1.7-1.9) times higher when exposed to antipsychotics than when unexposed. Age-stratified estimates suggest a greater triggering effect of antipsychotics among older patients. Exposure to antipsychotics may be a proximal trigger for stroke. Elevation in risk is apparent after brief exposure to antipsychotics.
    No preview · Article · Jan 2012 · Stroke
Show more