Structural Uncertainty in Economic Models

Definition and scope

Uncertainty about the model's representation of the decision problem, including state structure, extrapolation, treatment-effect duration, and evidence links, distinguished from parameter uncertainty within a fixed structure. The target population, use, and analysis remain protocol decisions; this card explains the choices and their consequences rather than approving a study.

Operational specification

Identify plausible alternative causal/clinical structures before viewing favored results: states and history, progression-survival relationship, treatment waning, cure assumptions, subsequent therapies, resource pathways, and external mortality constraints. Keep structural choices separate from estimated parameters, and record which sources or expert judgments support each alternative.

Analysis and diagnostics

Run coherent scenario models with compatible parameters and uncertainty. Show which decisions change, which assumptions drive the change, and what additional evidence could discriminate among structures. Model averaging requires justified model weights and compatible targets; assigning equal probabilities by convenience does not convert a scenario set into an empirically validated posterior.

When NOT to use and common errors

A wide PSA interval within one structure does not automatically include uncertainty about other structures. Do not combine incompatible state transitions, apply lifetime treatment benefits without evidence, or present the most favorable extrapolation as the uniquely supported model.

Worked example and implementation boundary

Two synthetic, separately coherent structures yield incremental NMB of 1000 and -500 at the same threshold and perspective. Calculate their difference, while avoiding an unsupported probability-weighted average. The scenario difference is 1500 and the decision changes sign. No probability is assigned to either structure by this calculation. The code performs that explicitly stated calculation with synthetic inputs. It is not a complete fitted analysis and does not demonstrate the identification assumptions, estimate sampling uncertainty, or certify application in another population.

Data-source and reporting requirements

Preserve the provenance, dates, units, and observation rule for each input. Claims capture services and transactions under observable benefits; absence of a record is not proof that an event or resource use did not occur. EHR and registry measurements depend on the collection process and may miss care outside the system. Linked sources add reconciliation and selection questions. Primary collection can improve a missing domain but still requires valid measurement and follow-up. Report which of these limitations affects the stated use, which evidence supports the chosen inputs, and what remains uncertain.

Worked example

Scenario

Two synthetic, separately coherent structures yield incremental NMB of 1000 and -500 at the same threshold and perspective. Calculate their difference, while avoiding an unsupported probability-weighted average.

Dataset

Synthetic inputs with units defined in the scenario; no patient data.

Steps

Result

The scenario difference is 1500 and the decision changes sign. No probability is assigned to either structure by this calculation.

Trade-offs

Code example

scenario1_nmb = 1000
scenario2_nmb = -500
result = scenario1_nmb-scenario2_nmb
assert abs(result - 1500) < 1e-9
print(result)

scenario1_nmb <- 1000
scenario2_nmb <- -500
result <- scenario1_nmb-scenario2_nmb
stopifnot(abs(result - 1500) < 1e-9)
print(result)

Citations

1. [1]Caro JJ, et al. Modeling Good Research Practices—Overview: A Report of the ISPOR-SMDM Modeling Good Research Practices Task Force-1. Value in Health. 2012.
2. [2]Eddy DM, et al. Model Transparency and Validation: A Report of the ISPOR-SMDM Modeling Good Research Practices Task Force-7. Value in Health. 2012.
3. [3]Husereau D, et al. Consolidated Health Economic Evaluation Reporting Standards (CHEERS) 2022 Explanation and Elaboration: A Report of the ISPOR CHEERS II Good Practices Task Force. Value in Health. 2022.