Skip to main content
. 2024 Aug 1;14(8):e082977. doi: 10.1136/bmjopen-2023-082977

Table 3. Summary of parameters and modelling approaches subject to sensitivity analysis.

Cause of uncertainty Parameters and modelling approaches subject to sensitivity analysis Reason
Data or sampling variations Missing cost data Some cost data collected retrospectively, so missing values will be assumed based on literature or comparable costs.
Intervention delivery ‘mode’ costs Unexpected pivot to telephonic sessions during COVID-19 pandemic meant that there is a wide variation in the ratio of in-person:telephonic sessions per participant.
Generalisability of results Discount rates applied to costs and consequences Key to explaining how costs and effects vary across different settings.
Methods used to measure and value costs and effects
  1. Discounting (or not discounting) consequences

  2. Costing time horizon

  1. Subject to disagreement among researchers, and greatly influence total effects.

  2. The costing time horizon will be adjusted accordingly so that the time horizons for costs and benefits match. This is in keeping with best practice.4