Nicolle M. Gatto

Redundant causation from a sufficient cause perspective

Sufficient causes of disease are redundant when an individual acquires the components of two or more sufficient causes. In this circumstance, the individual still would have become diseased even if one of the sufficient causes had not been acquired. In the context of a study, when any individuals acquire components of more than one sufficient cause over the observation period, the etiologic effect of the exposure (defined as the absolute or relative difference between the proportion of the exposed who develop the disease by the end of the study period and the proportion of those individuals who would have developed the disease at the moment they did even in the absence of the exposure) may be underestimated. Even in the absence of confounding and bias, the observed effect estimate represents only a subset of the etiologic effect. This underestimation occurs regardless of the measure of effect used. To some extent, redundancy of sufficient causes is always present, and under some circumstances, it may make a true cause of disease appear to be not causal. This problem is particularly relevant when the researchers goal is to characterize the universe of sufficient causes of the disease, identify risk factors for targeted interventions, or construct causal diagrams. In this paper, we use the sufficient component cause model and the disease response type framework to show how redundant causation arises and the factors that determine the extent of its impact on epidemiologic effect measures.

Causal identification: a charge of epidemiology in danger of marginalization

The requirement for framing all causal questions as well-defined interventions is being promoted in the causal inference literature within epidemiology. One can consider this perspective as an intervention on the field which requires a refocusing of epidemiologic questions and retooling of epidemiologic methods. Although this intervention has produced many positive results, we think that its underlying assumptions and the possibilities of unintended consequences warrant examination. In so doing, we argue that this approach can lead to the neglect of causal identification as a useful link between associations and the estimation of intervention effects.

Is the Large Simple Trial Design Used for Comparative, Post-Approval Safety Research?

Post-approval, observational drug safety studies face well known difficulties in controlling for confounding, particularly confounding by indication for drug use. A study design that addresses confounding by indication is the large simple trial (LST). LSTs are characterized by large sample sizes, often in the thousands; broad entry criteria consistent with the approved medication label; randomization based on equipoise, i.e. neither physician nor patient believes that one treatment option is superior; minimal, streamlined data collection requirements; objectively-measured endpoints (e.g. death, hospitalization); and follow-up that minimizes interventions or interference with normal clinical practice. In theory then, the LST is a preferred study design for drug and vaccine safety research because it controls for biases inherent to observational research while still providing results that are generalizable to ‘real-world’ use.To evaluate whether LSTs are used for comparative safety evaluation and if the design is, in fact, advantageous compared with other designs, we conducted a review of the published literature (1949 through 31 December 2010) and the ClinicalTrials.gov registry (2000 through 31 December 2010). Thirteen ongoing or completed safety LSTs were identified. The design has rarely been used in comparative drug safety research, which is due to the operational, financial and scientific hurdles of implementing the design. The studies that have been completed addressed important clinical questions and, in some cases, led to re-evaluation of medical practice. We conclude the design has demonstrated utility for comparative safety research of medicines and vaccines if the necessary scientific and operational conditions for its use are met.

Methodological comparison of marginal structural model, time-varying Cox regression, and propensity score methods: the example of antidepressant use and the risk of hip fracture.

Observational studies including time‐varying treatments are prone to confounding. We compared time‐varying Cox regression analysis, propensity score (PS) methods, and marginal structural models (MSMs) in a study of antidepressant [selective serotonin reuptake inhibitors (SSRIs)] use and the risk of hip fracture.

Extending the sufficient component cause model to describe the Stable Unit Treatment Value Assumption (SUTVA)

Causal inference requires an understanding of the conditions under which association equals causation. The exchangeability or no confounding assumption is well known and well understood as central to this task. More recently the epidemiologic literature has described additional assumptions related to the stability of causal effects. In this paper we extend the Sufficient Component Cause Model to represent one expression of this stability assumption--the Stable Unit Treatment Value Assumption. Approaching SUTVA from an SCC model helps clarify what SUTVA is and reinforces the connections between interaction and SUTVA.

Consensus of recommendations guiding comparative effectiveness research methods

Because of an increasing demand for quality comparative effectiveness research (CER), methods guidance documents have been published, such as those from the Agency for Healthcare Research and Quality (AHRQ) and the Patient‐Centered Outcomes Research Institute (PCORI). Our objective was to identify CER methods guidance documents and compare them to produce a summary of important recommendations which could serve as a consensus of CER method recommendations.

Author's response to Poole, C. Commentary: How Many Are Affected? A Real Limit of Epidemiology

We read with interest Charlie Poole’ sc ommentary [1] on our paper, “Redundant causation from a sufficient cause perspective,”[2] in which he questions the utility of the sufficient component cause (SCC) model for examining differences between etiologic and excess effects. Poole contends that the concept we term “redundant causation” is uncomplicated and (we presume), well understood. He questions whether “it needs to be explained in terms any deeper than those of potential outcomes” [1]. His critique of our paper focuses on our hypothetical and simplistic example of sufficient causes (SCs) of liver cancer. To be of value, Poole believes our example must be realistic and must bring “aspects of the potential outcome and sufficient cause models, and their interface, into sharp relief” [1]. His concerns raise larger issues about the roles of simplifications and the SCC model in methods research in general. We address each of these below. Simplifications