Using Causal Inference for Decision Making
In clinical research, decision makers need to choose among different courses of action every day. Whether it is choosing one treatment over another or deciding to treat now or later. These decisions can be informed by doing causal inference, which in this case indicates a complex scientific task that relies on triangulating evidence from multiple sources and on the application of a variety of methodological approaches. It helps to compare the outcomes under these said different courses of action as well as compare the effectiveness and safety. We call this approach head-to-head comparisons using real world data. It allows clients to compare treatments without resorting to expensive Phase 4 trials.
One way to arrive at the causal effects is by conducting randomized clinical trials (RCTs) which, in principle, can answer each causal question about comparative effectiveness and safety. But randomized trials are not always feasible as they are expensive, sometimes unethical, and often need time to plan and execute. What do we do when we do not have a randomized trial at the time of decision making? We try to emulate the RCT using real world data. A Target trial is the hypothetical randomized trial that we would like to conduct to answer a causal question. Any causal inference of observational data that we conduct can be viewed as an attempt to emulate some target trial.
However, there are inherent differences between real world data collection and RCT (e.g., some variables or outcomes not routinely or uniformly recorded, missing). Observational analyses are often criticized for lack of randomization and many of them have a fundamental problem of failing to choose a correct time zero. Time zero of follow up in the target trial is the time when eligibility is met, treatment strategies are assigned and study outcomes begin to be counted, for each person. It is one of the key components of emulating target trials. Hence, time zero must be synchronized with the determination of eligibility and assignment of treatment strategies.
Although, we cannot say that observational studies are as good as randomized experiments, we can do better by using observational data to explicitly emulate randomized trials. Target trials are typically a middle ground between an ideal trial we would like to conduct and the trial we can reasonably emulate using the available data. The target trial may also have a longer follow-up and therefore more clinically relevant outcomes, or a richer set of comparisons.
Join Cytel and industry experts for the Head to Head Comparisons Using Real World Data five part webinar series. Hear from our team of experts on two pilot projects on head-to-head comparisons using real world data, in oncology and cardiovascular disease. Click below to register.
About the Author of Blog:
Mansha Sachdev specializes in content creation and knowledge management. She holds an MBA degree and has 11 years of experience in handling various facets of marketing, across industries. At Cytel, Mansha is a Content Marketing Manager and is responsible for producing informative content that is related to the pharmaceutical and medical devices industries.