ICER, INB, and ACER in Cost-Effectiveness Analysis

Research updated on July 29, 2025
Cite: Biopharma Foundry. (2026, Month Day). Article title in italics. Article link
Author: Santhosh Ramaraj

When evaluating treatments, healthcare decision-makers need to compare both costs and health benefits. It is not just about which drug works better, but also about whether the improvement in patient outcomes is worth the additional cost. In pharmaceutical economics, three key measures are often used: Incremental Cost-Effectiveness Ratio (ICER), Incremental Net Benefit (INB), and Average Cost-Effectiveness Ratio (ACER). These measures provide structured ways to compare treatments, especially in randomized controlled trials (RCTs).

What is ICER?

The Incremental Cost-Effectiveness Ratio (ICER) measures the extra cost required to achieve one additional unit of health outcome (such as life-years gained or quality-adjusted life years) when using one treatment compared to another.

If we have two treatments, let:

$C_1$ and $C_2$ be the average costs of treatment 1 and treatment 2.
$E_1$ and $E_2$ be the average effectiveness measures for treatment 1 and treatment 2.

The formula for ICER is:

$ICER = \frac{\mu_{C1} - \mu_{C2}}{\mu_{E1} - \mu_{E2}}$

This ratio tells you how much additional cost is required for each unit of effectiveness gained by choosing treatment 1 over treatment 2.

Example of ICER

Imagine two migraine treatments:

Treatment A costs $5,000 and improves patient outcomes by 3 quality-adjusted life years (QALYs).
Treatment B costs $3,000 and improves outcomes by 2 QALYs.

The ICER is:

$ICER = \frac{5000 - 3000}{3 - 2} = \frac{2000}{1} = 2000$

This means Treatment A costs $2,000 for each additional QALY compared to Treatment B.

Limitations of ICER

ICER, while widely used, has some limitations:

If $\mu_{E1} - \mu_{E2}$ (the difference in effectiveness) is very small, the denominator approaches zero, which can cause misleading results.
ICER cannot distinguish between a treatment that is both cheaper and more effective (dominant) and one that is more expensive and less effective (dominated).
Statistical inference for a ratio like ICER is complex.

Because of these issues, other measures like Incremental Net Benefit (INB) are often preferred in certain scenarios.

What is INB?

The Incremental Net Benefit (INB) introduces the concept of willingness-to-pay (WTP), represented by $\lambda$ . This is the maximum amount a policy-maker is willing to pay for one additional unit of effectiveness. INB is calculated as:

$INB(\lambda) = \lambda (\mu_{E1} - \mu_{E2}) - (\mu_{C1} - \mu_{C2})$

If INB(λ) > 0, treatment 1 is considered cost-effective compared to treatment 2 because the health gain (valued at $\lambda$ ) outweighs the additional cost.
If INB(λ) < 0, treatment 1 is not cost-effective.

Example of INB

Continuing with the migraine example:

$\mu_{E1} - \mu_{E2} = 1 , \text{QALY}$
$\mu_{C1} - \mu_{C2} = 2000$

Suppose $\lambda = 50000$ (meaning the decision-maker is willing to pay $50,000 per QALY).

The INB is:
$INB(50000) = 50000 \times 1 - 2000 = 48000$

Since INB is positive, Treatment A is cost-effective under this WTP threshold.

Relationship Between ICER and INB

ICER and INB are mathematically linked. If INB(λ) > 0, it means ICER < λ. In simpler terms, if the cost per additional QALY (ICER) is less than the willingness-to-pay threshold, the treatment is considered worth it.

What is ACER?

The Average Cost-Effectiveness Ratio (ACER) looks at the average cost per unit of effectiveness for a single treatment, not the difference between two treatments. It is defined as:

$ACER = \frac{\mu_C}{\mu_E}$

where:

$\mu_C$ is the mean cost of a treatment.
$\mu_E$ is the mean effectiveness of that treatment.

You can also calculate the incremental ACER (ΔACER) between two treatments:

$\Delta ACER = \frac{\mu_{C1}}{\mu_{E1}} - \frac{\mu_{C2}}{\mu_{E2}}$

Example of ACER

For Treatment A:
$ACER_A = \frac{5000}{3} \approx 1666.67$

For Treatment B:
$ACER_B = \frac{3000}{2} = 1500$

The incremental ACER is:
$\Delta ACER = 1666.67 - 1500 = 166.67$

This suggests that Treatment A costs an extra $166.67 per unit of effectiveness on average compared to Treatment B.

Debate Around ACER

While ACER can provide useful information, it does not always guide decision-making as effectively as ICER. For example, it is possible for a treatment to have a lower ACER but still be less cost-effective when directly compared to another treatment. Some researchers argue ACER can mislead healthcare decisions, which is why ICER remains the primary metric.

Why These Measures Matter in Pharma

For pharma executives, understanding ICER, INB, and ACER is crucial when assessing new treatments or negotiating with payers. These metrics inform whether a new drug is worth its price compared to existing alternatives.

Example in Practice:
When introducing a biologic drug for rheumatoid arthritis, pharma teams compare its cost and QALY gains against cheaper generic drugs. If ICER is well below the WTP threshold (e.g., $50,000/QALY in the US), the drug has a higher chance of getting reimbursement approval.

Steps to Calculate and Interpret

Collect Data: Use RCT data to obtain average costs and effectiveness measures for both treatment groups.
Compute ICER: Apply the ICER formula to understand cost per additional unit of health gain.
Define WTP: Decide the maximum acceptable cost per unit of effectiveness.
Calculate INB: Check if INB(λ) is positive.
Compare ACER: Use ACER for single-treatment analysis, but do not rely on it alone for decisions.
Interpret Results: A treatment with a low ICER or high positive INB is generally preferred.

Relevant Reads