Economic Modelling and Transferring Economic Evaluations Flashcards
(17 cards)
what is alongside
- Evaluations where the economic data are prospectively collected from patients within an effectiveness evaluation (sometimes referred to as ‘EEACTs’)
- The economic and clinical evaluations run concurrently
- Costs and outcomes are estimated for each individual patient
- Low bias (if the underlying evaluation is unbiased) and high ‘internal validity’
Issues with alongside
- Comparators are limited to those in the study
- How many HIV medications are there?
- Length of follow-up can be too short
- How long to assess the effectiveness of HIV meds?
- Low ‘external validity’?
- Are patients, doctors and hospitals typical?
- Inadequate sample size for rare events?
Role of models today
- It provides a method for combining data in a structured manner; data are chosen to meet the decision maker’s needs…all relevant comparators, appropriate time horizon and representing the relevant patient population
- Best estimates of parameters are combined in a ‘base case’ analysis, then altered in sensitivity analyses
what is a model
• A model is a simplification of the real world. The world is described by a set of parameters which represent our knowledge of that world and which can be changed to to assess different scenarios
list at least 4 model types
- Decision trees
- State transition (“Markov”) models
- Transmission models
- Discrete event simulation models
- Hybrid models, e.g. trees with Markov terminal nods
- All are different types of decision analytic model
- Unstructured calculations are also sometimes used…are they models? Technically, yes
what types of data do models use
patient data
what type of patient data does models use
- Cohort
- Probabilities relate to averages for the patient population
- Model is run once using the mean value for each probability
- Mean costs and outcomes are produced
- Individual patient
- Probabilities are specific to each individual, based on baseline characteristics and subsequent events
- Model is run for each individual in the patient population
- Mean costs and outcomes are calculated from the distributions produced by each run of the model
- Sometimes referred to as individual patient simulations
When to use a decision tree
For simple models or problems with special characteristics (e.g., very short time horizons, very few outcomes), a decision tree may be appropriate.
when not to use a decision tree
Not useful when:
• States ‘recycle’ or there are recurrences of same illness
• Modelling complex problems over a long time
• Timing of events is important
• Need to model interaction
what is the State-transition models
model type
Disease is defined by health states (ovals)
Disease progression is modelled by transitions between states (lines)
Transitions occur at end of ‘cycles’ and are ‘memoryless’
when do you use ‘Markov’ models?-
more suitable for long term conditions
what are ‘Markov’ models?-
- The conditional probability distribution of future states of the process depends only upon the present state, not on the sequence of events that preceded it.
- Markov models are memoryless – transition probabilities do not depend on history
- Markov assumption required in cohort-level state transition models
When to use a state-transition model
If the conceptualization involves representing the disease or treatment process as a series of health states, state-transition models are appropriate.
Their primary disadvantage, the Markovian assumption that transition probabilities do not depend on history, can be addressed by increasing the number of states.
Individual state-transition models, which do not require this assumption, are an alternative when the number of states grows too large
when not to use a state-transition model
when
• Many events could occur in one time unit
• Transitions are depend on patient history, e.g. number of previous episodes of ill-health or ‘time to….’
• Modelling interactions
what do you need to consider when Choosing a modelling method
- Representation of disease as states or events
- Recurring events
- Time horizon
- Influence of past events
- Interactions between individuals
Rationale for transferability
- Economic evaluations are time consuming and expensive.
- Lack of technical expertise and data in some countries
- Effectiveness is generally generalisable…most big drug trials are multinational
- Existing studies done elsewhere can be transferred to other contexts
- Just change the currency?
Why is transferability an issue?
- Countries differ in many different ways:
- Treatment patterns
- Effectiveness….genetics, case mix, adherence, etc.
- Costs…price levels, quality of care, etc.
- Values placed on outcomes….health state utilities
- Acceptable methods*…discount rates, outcome measures
- Funding thresholds
- Some of these can be solved with simple adjustments (e.g unit costs), but some may require additional data collection (e.g. new comparators)
