Ideally, in HTA we would perform randomized controlled trials (RCTs). However, this is not always possible, so we have to make do with non-randomized studies. In that case, we have to be careful: We need to take care of confounders!

 

What is a confounder?

A confounder is a factor that is associated with both an intervention and the outcome criterion of a study. It comes from the Latin word "confundere", which means "confuse" or "mix".

 

Challenges

⚠️When comparing groups which were not chosen randomly, we can't assume that these groups are homogenous. There may be systematic differences between them.

⚠️Ignoring the confounders can lead to false conclusions about the causal relation between the intervention and the outcome criterion.

 

Statistical Method

📊 For non-randomized studies Propensity-Score-Methods are widely used. It is important to adjust the model for all confounders by incorporating them into the model.

 

How to find confounders?

🔎A new systematic approach was presented 2022 by Pufulete et al. (https://www.jclinepi.com/article/S0895-4356(22)00079-8/abstract). In April 2025, the IQWiG published a report evaluating and extending this method (https://www.iqwig.de/download/ga23-02_systematische-confounderidentifikation-indikation-rrms_arbeitspapier_v1-0.pdf).

 

 

📢 Stay tuned and continue Building Statistical Confidence For Market Access