Table 3.

Key points of target trial emulation

1. For all observational studies that aim to estimate the causal effect of a treatment, one can imagine the randomized trial that could have been conducted instead (=“target trial”).

2. Target trial emulation is a framework for the design and analysis of observational studies and involves precisely specifying the protocol of the target trial and then emulating each component of the protocol with observational data.

3. The design of randomized trials should be explicitly emulated by carefully aligning eligibility criteria, treatment assignment, and start of follow-up; this prevents avoidable biases, such as immortal time bias and depletion of susceptibles bias.

4. Target trial emulation is not a single study design or analysis; different target trial protocols may need different observational emulation designs and analyses.

5. Target trial emulation does not solve the problem of confounding in observational studies; this requires measuring and appropriately adjusting for all confounders.

6. Increased use of target trial emulation has great potential to improve the quality of causal observational studies; such high-quality studies can be used to complement evidence from randomized controlled trial, e.g., by investigating rare side effects or underrepresented populations.