Model: use past and present costs and outcomes to predict future costs and outcomes and better define the “value” of an intervention

Value: what I want vs what I get

Budget Impact Models (BIMs) are developed by: drug manufacturers to aid a decision-maker in resource planning

BIM: accounting models used to calculate how a healthcare budget will be impacted by new drugs or a change in utilization of existing drugs (for example, an indication being expanded to both COPD and asthma, now there is a larger pt population)

How a BIM works: enter population characteristics (population size, target population size, current intervention, new intervention, drug costs and other costs) → model calculates projected future expenditures for current AND future scenario

Model Calculation is expressed as: PMPM (per member per month) OR PMPY (per member per year)

3 types of cost: 1. direct (quantifiable - meds, labs, etc) 2. nonmedical (nonquantifiable - transportation, babysitting, etc. 3. indirect (loss of productivity)*****

BIM advantage: simplest model

BIM disadvantage: assumed outcomes are EQUAL; developed from manufacturers - may lack credibility

BIM assesses: financial consequences in a short-term time frame to estimate cost

CEM assesses: value of a new drug in a long-term time frame to estimate cost per life year or quality-adjusted life years (QALY)

Calculation in CEM: ICER (evaluates cost and effectiveness)

ICER: incremental cost-effectiveness and ratio

Quadrant 1: high cost, high efficacy (use lambda to evaluate)

Quadrant 2: high cost, low efficacy (bad)

Quadrant 3: low cost, low efficacy (use lambda to evaluate)

Quadrant 4: low cost, high efficacy (good)

Name of the HTA agency in the UK: NICE

CEM advantages: established quantitative approach, allows for consideration of costs and outcomes

CEM disadvantages: complicated to explain to clinicians and healthcare decision-makers, transparency might be questioned, dependent on the quality of data

Key Parameters for Model Development: Type (BIM or CEM), Perspective, what costs to include, time horizon, relevant population, comparators

Good models: model inputs and managing with assumptions

One-Way Sensitivity Analysis: change one input at a time of assumptions

Probabilistic Sensitivity Analysis: change all variables at the same time of assumptions