Research & Regulatory News

Target Trial Emulation: Making Observational Data Behave

UKPWritten and published by UK Peptides2 min readUpdated 2 cited sources

The short answer

Target trial emulation is a method for analysing observational data by first specifying the randomised trial you would have run, then reproducing its design as closely as the data allows. It removes several classic biases and cannot remove confounding by unmeasured factors.

Key facts

Applies to
Routine records, claims, registries
Core step
Specify the hypothetical trial first
Fixes
Immortal time bias, prevalent user bias
Does not fix
Unmeasured confounding
GLP-1 example
Tang 2025 (PMID 40874398)
Tirzepatide example
Lin 2025 (PMID 40368924)

The problem it addresses

Observational analyses of health records are cheap, large and fast, and they are riddled with biases that arise from how the analysis is set up rather than from the data itself. The most common is immortal time bias, where a period during which an outcome could not have occurred gets assigned to the treatment group, making the treatment look protective for arithmetical reasons.

How the method works

Write down the trial you would run (eligibility, treatment strategies, assignment, follow-up start, outcome, analysis plan) before touching the data. Then construct the analysis so each element is matched as closely as the records permit. Specifying the design in advance is what prevents the analytical choices from being made after seeing which ones produce a result.

Research material referenced

Retatrutide 10mg, third-party HPLC tested

Buy Retatrutide · £59.99

Why the alignment of time zero is the crux

In a real trial, eligibility, assignment and the start of follow-up happen at one moment. In records they are scattered, and analysts routinely misalign them without noticing. Forcing them to coincide is the single change that eliminates most of the bias, and it is the discipline the framework really imposes.

What it still cannot do

Randomise. Emulation controls for what was recorded; it cannot balance what was not. If people prescribed a drug differ systematically from those who were not in some way absent from the records, such as motivation, unrecorded severity or access, the comparison remains confounded no matter how carefully the design is emulated.

Why it matters for obesity medicine specifically

The cardiovascular and kidney questions being asked about incretins would require enormous, decade-long randomised trials. Emulation offers earlier answers from data that already exists, which is valuable. It also means a growing share of what gets reported as GLP-1 outcome evidence is not from randomised trials at all, and readers should know which they are looking at.

How to read one

Check whether the target trial is actually specified in the paper. Check where time zero sits. Check which confounders were available and, above all, which were not. A well-conducted emulation states its own limitations clearly; the reporting that follows it usually does not.

Frequently asked questions

Is target trial emulation as good as a randomised trial?

No. It removes several design-induced biases but cannot balance unmeasured factors, which is what randomisation does.

What is immortal time bias?

A period during which the outcome could not have occurred being assigned to the treatment group, making the treatment appear protective for arithmetical reasons.

Why is it being used for GLP-1 questions?

Cardiovascular and kidney outcome questions would need enormous, decade-long trials. Emulation offers earlier answers from existing data.

Continue researching

Related guides and reference material.

Primary sources & clinical trials

Published papers, trial registrations, regulator pages and company announcements. Links open the original.

UKP

Written and published by UK Peptides, the trading name of AURYXRESEARCH LTD (company 17388727, registered in England and Wales).

Questions or a correction? Email hello@ukpeptides.org or call +44 7426 987848.

Corrections log

More Research & Regulatory News articles

Research use only. The information above is provided for scientific and educational reference. Compounds referenced are not approved for human use and are supplied for in vitro research or reference-material purposes only. No efficacy, safety, or therapeutic claims are made.