AE-NNT (Adverse Event adjusted Number Needed to Treat)

1. Description

AE-NNT (Adverse Event Adjusted-NNT) estimates the number of patients to be treated to observe one patient in whom treatment was successful without inducing treatment-related adverse events. This is described as the "unqualified success" in the original publication.[1] Therefore, AE-NNT penalises NNT for the occurrence of AEs in the same patient.

2. Evaluation

2.1 Principle

• A "successful" treatment (benefit) and the adverse events (risk) need to be clearly defined a priori.

• AE-NNT only handles one benefit and one risk criteria.

• However the selected risk must not occur in the same patient when the benefit occurs.

2.2 Features

• AE-NNT integrates benefit and risk into one measure but only estimates the NNT in the best case scenario.

• An opposite of AE-NNT is the "unmitigated failure" concept, which estimates the NNH in the worst case scenario i.e. only risk occurred but not benefit.

2.3 Visualisation

• Visualisations similar to the one used in NNT may be used.

2.4 Assessability and accessibility

• Patient-level data are required.

• Use with summary data requires strong assumptions about mutually exclusive events have to be made.

• AE-NNT is undefined when there is no difference between treatment and comparator group.

• The confidence intervals also suffer from the shortcomings of NNT.

3. References

[1] Schulzer M, Mancini GB. 'Unqualified success' and 'unmitigated failure': number-needed -to-treat-related concepts for assessing treatment efficacy in the presence of treatment-induced adverse events. Int J Epidemiol 1996 Aug;25(4):704-12.