10 Practical guidelines for better health economic evaluations

By Prof. Dr Lieven Annemans, expert-trainer of the course Health Economics for Non-Health-Economists
 

For many non-health-economists, understanding what makes health economic evaluations robust and decision-relevant isn’t always straightforward. To support pharma professionals who work with health economic data, whether in Medical Affairs, Market Access, or Marketing, Prof. Dr Lieven Annemans has listed the below 10 key guidelines that reflect international best practices.
 

1. Clearly define the medical problem and patient population

Economic models must start with a well-defined patient population. Vague definitions like “patients at risk of stroke” don’t provide enough context for meaningful conclusions. A good example: “Spinal cord stimulation in patients with failed back surgery, average age 52, with predominantly back pain and an average pain duration of 4 years.”
 

2. Choose the most relevant comparator

Cost-effectiveness is always relative. The new treatment should be compared to the most likely alternative in clinical practice, not a placebo or outdated option. A thorough description of both treatment strategies is important, including known side effects and usage patterns.
 

3. Be clear about the evaluation perspective

Is the analysis conducted from a payer, healthcare system, or societal perspective? This matters. For instance, the societal perspective might include productivity losses, especially relevant in diseases like depression, while payer perspectives often do not.
 

4. Make the model transparent and validate its outcomes

The assumptions behind the model should be clearly explained. Where possible, model outputs should be validated against real-world data. For example: comparing predicted stroke rates under current care with those observed in practice.
 

5. Use realistic and well-sourced cost estimates

Costs should be broken down into all relevant components (e.g. acute care, follow-up, re-admissions) and linked to the defined patient population. All sources must be disclosed. Over-simplified assumptions can lead to misleading results.
 

6. Use reliable clinical inputs and appropriate outcomes

Data inputs should come from high-quality sources and reflect the population being studied. When it comes to outcomes, the QALY remains the standard across many countries. Using the most relevant data for the base case is key.
 

7. Choose a time horizon that reflects the full impact

The time horizon should be long enough to capture all significant costs and effects. Results should be tested for different horizons to show how conclusions might change over time.
 

8. Thoroughly analyse uncertainty

Every model involves uncertainty. High-quality evaluations include sensitivity analyses, both one-way (e.g. Tornado diagrams) and probabilistic (PSA), to show how results vary with changes in key inputs and assumptions.
 

9. Apply appropriate discounting

If the time horizon extends beyond one year, future costs and benefits should be discounted. Including alternative discount rates in sensitivity analysis is good practice.
 

10. Be specific and well-reasoned in your conclusions

A strong conclusion is based on the PICOS framework: Patient, Intervention, Comparator, Outcome, and Study Design. Rather than saying “the treatment is cost-effective,” clarify for whom, compared to what, and over what time frame.

These 10 guidelines are increasingly important as health economic evaluations influence funding, pricing, and access decisions in a growing number of countries. By applying or recognising these principles, you can better evaluate evidence and contribute to stronger, more credible value communications.