Incremental Net Benefit

The difference in expected net benefit between two competing interventions, found by converting the incremental health effect into money and subtracting incremental cost.

Source: Stinnett & Mullahy 1998

The incremental health effect is multiplied by the willingness-to-pay threshold to express it in money, and the incremental cost is subtracted, giving a single figure for the comparison between two options. A positive result indicates that the additional health is worth more than the additional resource at that threshold. Where the analysis is probabilistic, the calculation is performed for every simulation and the results averaged, which is what makes the measure an expectation rather than a point estimate. Because the result is a monetary quantity rather than a ratio, options can be ranked directly on it without any intermediate comparison step, which is what makes it convenient when more than two alternatives are being assessed.

Source: Stinnett & Mullahy 1998

Because it is linear in the underlying parameters, so it can be averaged across simulations, compared directly between options and analysed statistically in ways a ratio cannot. Ratios become unstable as the difference in effect approaches zero, since the denominator vanishes, and they cannot distinguish an option that is cheaper and less effective from one that is dearer and more effective. Net benefit returns a single figure whose sign carries the recommendation regardless of which quadrant the comparison falls in. It also handles the case where one option is both cheaper and more effective without special treatment, simply returning a larger positive figure, whereas a ratio in that situation is negative and requires interpretation.

Source: Stinnett & Mullahy 1998

Entirely, since the health gain is valued at it, so both the magnitude and potentially the sign change as the threshold moves. This is why results are reported across a range of thresholds rather than at a single figure, and why the threshold at which the sign changes is more informative than the value at any one point. That threshold is arithmetically identical to the incremental cost-effectiveness ratio, so the two measures carry the same information presented differently. Reporting the crossing point alongside the figure therefore communicates both the recommendation and the threshold at which it would change, which is more useful to a decision maker than either alone.

Source: Briggs, Claxton & Sculpher 2006

Yes. Dividing the incremental cost by the threshold converts it into the health that would be displaced elsewhere, which is then subtracted from the health gained to give incremental net health benefit. This avoids monetising health and states the result in the units the decision actually concerns, which many decision makers find more intuitive. The ordering of options is identical under both formulations, so the choice between them is presentational. The health formulation also has the practical advantage of expressing the result in the same units as the displaced activity, which makes the opportunity cost explicit rather than implicit in the threshold.

Source: Stinnett & Mullahy 1998

The threshold used and the range across which results were calculated, since the figure means nothing without them. The comparator, perspective, horizon and discount rate, as for any incremental measure. And the distribution across simulations rather than the mean alone, since two comparisons with the same expected net benefit can differ substantially in how much of that distribution falls below zero, which is what determines the confidence attaching to the recommendation. Where the analysis covers subgroups, results should be reported for each rather than aggregated, since a positive figure overall can conceal groups for whom the intervention is not worthwhile at the same threshold.

Source: Drummond et al. 2015