Concept Architecture
Concept
Theoretically, Heterogeneous Treatment Effect (HTE) refers to variation in the causal effect of an intervention across individuals or population subgroups. It recognises that treatment effects are rarely uniform because patient characteristics, disease severity, genetics, comorbidities or healthcare settings may modify the response to treatment. In health economics, heterogeneous treatment effects are fundamental to personalised medicine, subgroup cost-effectiveness analysis and resource allocation decisions.
Mathematically, heterogeneous treatment effects are represented by allowing the causal treatment effect to vary as a function of observed or unobserved individual characteristics. Within the potential outcomes framework, the individual treatment effect is defined as the difference between potential outcomes under treatment and control, while statistical models estimate how these effects differ across covariates using interaction terms, hierarchical models or machine learning methods.
In practice, heterogeneous treatment effects are estimated using subgroup analyses, regression models with treatment interactions, causal forests, Bayesian hierarchical models, propensity score methods and other causal inference techniques. Health economists use these estimates to evaluate whether interventions provide greater value in specific patient populations and to support stratified reimbursement, precision medicine and targeted healthcare policies.
Purpose
Used to estimate how treatment effects vary across individuals or population subgroups, supporting personalised healthcare, subgroup cost-effectiveness analysis and evidence-based resource allocation.
Mathematical Formulae
Primary Formula
HTE(x) = E[Y(1) ? Y(0) | X = x]
where:
- HTE(x) = conditional treatment effect for individuals with characteristics X = x
- Y(1) = potential outcome under treatment
- Y(0) = potential outcome under control
- E(�) = conditional expectation
Supporting Formulae
ITE? = Y?(1) ? Y?(0)
Y = ?? + ??Treatment + ??X + ??(Treatment ? X) + �
HTE = ??
Related Mathematical Methods
- Potential outcomes framework
- Conditional Average Treatment Effect (CATE)
- Average Treatment Effect (ATE)
- Regression interaction models
- Bayesian hierarchical modelling
- Causal forests
- Propensity score methods
Example
A new oncology treatment is evaluated in two patient subgroups defined by biomarker status. The estimated treatment effect is 0.45 additional QALYs for biomarker-positive patients and 0.08 additional QALYs for biomarker-negative patients.
Regression analysis includes an interaction between treatment and biomarker status. The statistically significant interaction coefficient indicates a heterogeneous treatment effect, suggesting that the intervention provides substantially greater benefit in biomarker-positive patients and may therefore be more cost-effective when targeted to this subgroup.
Excel Implementation
| Function | Example Formula | Health Economics Application |
|---|---|---|
| LINEST | =LINEST(Y2:Y501,X2:AA501,TRUE,TRUE) | Estimates regression models containing treatment-by-subgroup interaction terms |
| IF | =IF(B2="Positive",1,0) | Creates subgroup indicators for interaction analysis |
| AVERAGEIFS | =AVERAGEIFS(Outcome,Treatment,1,Subgroup,"Positive")-AVERAGEIFS(Outcome,Treatment,0,Subgroup,"Positive") | Estimates subgroup-specific treatment effects |
| SUMPRODUCT | =SUMPRODUCT(OutcomeRange,WeightRange)/SUM(WeightRange) | Calculates weighted subgroup treatment effects where appropriate |
VBA (Optional)
VBA can automate estimation of heterogeneous treatment effects across multiple predefined subgroups and generate comparative reports for economic evaluation.
Sources
- Hern�n MA, Robins JM. Causal Inference: What If.
- Rubin DB. Estimating Causal Effects of Treatments in Randomized and Nonrandomized Studies. Journal of Educational Psychology.
- Athey S, Imbens GW. Recursive Partitioning for Heterogeneous Causal Effects. Proceedings of the National Academy of Sciences.
- Wager S, Athey S. Estimation and Inference of Heterogeneous Treatment Effects Using Random Forests. Journal of the American Statistical Association.
- Drummond MF, Sculpher MJ, Claxton K, Stoddart GL, Torrance GW. Methods for the Economic Evaluation of Health Care Programmes.
- Briggs A, Claxton K, Sculpher M. Decision Modelling for Health Economic Evaluation.
Related Concepts (2)
Library
Publications
1
Interpreting Indirect Treatment Comparisons and Network Meta-Analysis for Health-Care Decision Making: ISPOR Task Force on Indirect Treatment Comparisons Good Research Practices, Part 1 — Jansen, Fleurence, Devine, Itzler, Barrett, Hawkins, Lee, Boersma, Annemans & Cappelleri, Vol. 14, No. 4 ed., 2011 (Value in Health)
The ISPOR good-practice guidance on interpreting indirect treatment comparisons, network and mixed treatment comparisons for decision making — terminology, assumptions, validity and how to critically appraise an ITC/NMA when head-to-head trial evidence is unavailable.
Journal ArticleView source →
Frequently Asked Questions (6)
What is a heterogeneous treatment effect?
Variation in the size or direction of a treatment's effect across individuals or subgroups, so the overall average may not fit any one patient.
Source: Kravitz, Duan & Braslow 2004
What does a heterogeneous treatment effect imply for applying trial results?
When a treatment's effect varies across patients, the single average a trial reports may fit few of them well, being a blend of larger benefits for some and little or none for others. Applying that average uniformly can therefore lead to treating patients who will not benefit and overlooking those who would benefit most. Recognising heterogeneity means looking beyond the average to who gains and who does not, so that treatment can be directed where it helps. The average conceals a distribution of effects. Kent and colleagues (2018) discuss this.
Source: Kent et al. 2018
Why do treatment effects vary between patients?
Treatment effects vary between patients because individuals differ in characteristics that influence how they respond, such as disease severity, biology, comorbidities, genetics, age, adherence, and competing risks, so the same treatment can produce different benefits or harms in different people. Baseline risk also affects the absolute benefit, with higher-risk patients often gaining more. These sources of variation mean the effect is not uniform across a population. Recognising why effects vary helps identify which patients are most likely to benefit, supporting more targeted and effective use of treatments.
Source: Kravitz, Duan & Braslow 2004
Why do heterogeneous treatment effects matter?
Heterogeneous treatment effects matter because the average effect from a trial may not apply to an individual patient, so treating everyone according to the average could give less benefit, or even harm, to some, while others benefit greatly. Understanding the variation allows treatments to be targeted to those most likely to benefit, improving outcomes and efficiency. For clinical decisions and personalised care, knowing how effects vary is important, since it moves beyond a single average to consider which patients should receive a treatment, making heterogeneity central to individualising care.
Source: Rubin 1974
How are heterogeneous treatment effects investigated?
Heterogeneous treatment effects are investigated through subgroup analyses examining the effect in groups defined by characteristics, analyses of effect modification, and, increasingly, methods that model how effects vary with patient features, including risk-based and machine-learning approaches. Prespecification and adequate power are important, since post hoc subgroup findings can be spurious. Predicting individual benefit from baseline risk and characteristics helps. So investigating heterogeneity uses careful analytic methods to identify genuine variation in treatment effect, distinguishing real effect modification from chance, to inform which patients benefit most.
Source: Kravitz, Duan & Braslow 2004
What are the risks of ignoring heterogeneous treatment effects?
The risks of ignoring heterogeneous treatment effects include applying an average effect to all patients when some benefit little or are harmed, leading to suboptimal or harmful treatment for those individuals, and missing opportunities to target treatment to those most likely to benefit. Relying solely on the average can also obscure important variation relevant to decisions. Conversely, over-interpreting spurious subgroup differences carries its own risk. So the challenge is to recognise genuine heterogeneity without being misled by chance, since both ignoring real variation and inventing false variation can lead to poor treatment decisions.
Source: Kravitz, Duan & Braslow 2004
Trust Record
Verified by Dr Darrin Baines
British health economist
Professional identity: darrinbaines.org
Verification date: 20 Nov 2025
Content version: 1.0.0
Canonical Identity
- Term code
- HE-ES-CER-016
Stable URI · Machine-readable · Resolvable · CC BY 4.0