Machine learning methods for estimating individual treatment effect from real world data for use in Health Technology Assessment: Separating hype from reality

 

There is however a great deal of confusion about when and how these models can be used, what data we would need and importantly whether their output(s) meet the input requirements that health economics decision models typically have. 

 

In this talk, we review and taxonomize existing ML methods that estimate subgroup-specific and individualised treatment effects from real-world data under specific types of outcome variables, treatment exposure and data structure. The purpose is to help HTA practitioners identify the most appropriate method to use, depending on: the available data (cross-sectional or longitudinal); the outcome of interest (continuous, binary or time-to-event); whether the method can handle observed or unobserved confounders; if the method explicitly quantifies measure(s) of uncertainty; the software (R, Python or Stata) used to implement it.  

 

By contrasting the taxonomy against the information required to conduct CEA for HTA, the paper highlights the gaps that methods developers need to address for ML to become integral part of the next-generation toolbox used in HTA. 

 

Join Zoom Meeting
https://york-ac-uk.zoom.us/j/93319055715?pwd=RlA0R1h6Z0p0b0U2d3J2K0RaTVRhQT09

Meeting ID: 933 1905 5715
Passcode: 707822