Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Markov Models, part II Marcelo Coca Perraillon University of Colorado Anschutz Medical Campus Cost-Effectiveness Analysis HSMP 6609 2016 1 / 37 Outline More examples Markov model extensions 1 2 3 4 Incorporating time dependency Relaxing the Markov assumption (memoryless property) Patient-level simulation (microsimulations) Static and dynamic models Summary 2 / 37 Big picture In the last two classes we incorporated uncertainty into cost-effectiveness using decision analysis and Markov models Decision trees without a Markov model within are less common Markov models provide more flexibility; they incorporate uncertainty and also model disease progression Good setting for calculating life expectancy and costs Today we will see more examples and talk about extensions to Markov models 3 / 37 Recap We needed very few elements to build a Markov model 1 2 3 4 Health states Cycles Transition probabilities Rewards (costs and benefits) Even though there are few elements, Markov models can be fairly complex (in a good way) 4 / 37 Example 1: description Cost-effectiveness of options for the diagnosis of high blood pressure in primary care, Lovibond et al (2011), Lancet; 378:1219-30 Compare three diagnosis strategies. Further blood pressure measurement 1) in clinic, 2) at home, 3) ambulatory monitor Hypothetical primary care population aged 40 or older with a screening blood-pressure greater than 140/90 and risk-factor prevalence of the general population (UK) Cycle: 3 months; time-horizon: 60 years Rewards: costs and QALYs Stratified by age Data: meta-analyses, risk of events using Framingham risk equations 5 / 37 Example 1: Model 6 / 37 Example 1: Model Note that all paths lead to death Different types of states: hypertension, diagnosed, events Possible to go back to suspected hypertension in some cases Certain events last only one cycle (MI, angina, stroke, TIA) (tunnel states; more on this shortly) Even though patients stay in that state for one cycle, this cycle is expensive and lowers QALY (and it changes the probability of death) One Markov model by intervention; used Excel, stratified by sex and age group Transition probabilities were not fixed 7 / 37 Example 1: Conclusions Ambulatory monitoring most cost effective After an initial raised reading, it reduces the misdiagnosis and thus saves costs Additional costs of monitoring are compensated by cost savings from targeted treatment Recommended monitoring before starting antihypertensive drugs 8 / 37 Example 2: description Objective: to estimate the cost effectiveness of herpes zoster (shingles) vaccine versus no vaccination Herpes zoster: a reactivation of the chickenpox virus in the body (causes a painful rash). No cure so prevention is key Cohort entered the model healthy at 50 y/o Rewards: costs and QALYs; time horizon: 70 years (lifelong); cycle: 1 year 9 / 37 Example 2: Model 10 / 37 Example 2: Model 11 / 37 Example 2: features Model expressed as decision tree with Markov models The decision node shows the options: vaccine or no vaccine Sex is a chance node; similar to stratification in the hypertension example but in that example they did not show a tree Their model is not very clear. It is easier to understand Markov models using transition diagrams 12 / 37 Example 2: Conclusions Not cost effective: ICER = $500,754 per QALY The vaccine is expensive and the incidence of shingles is low at that age Efficacy of vaccine is very low after 10 to 12 years Better to use the vaccine in older patients 13 / 37 Example 3: description Cost effectiveness of alternative treatments for breast cancer One-year adjuvant trastuzumab (AT) therapy, with or without anthracyclines Evidence comes from clinical trial data; they wanted to model lifespan outcomes 49 y/o women with early-stage breast cancer Cycle: 1 month; time-horizon: could not find it in paper, probably around 50 years 14 / 37 Example 3: Model 15 / 37 Example 3: Conclusions AT ICER is $39,982 per QALY Of course, results sensitivy to medication costs A straightforward model with a lot of assumptions about parameter inputs They needed to make assumptions because clinical trials are short term 16 / 37 Extending Markov models: time-dependency So far we have only dealt with Markov models that have the same transitions probabilities each cycle The transition probabilities did not depend on the time spent in one state or just time (i.e. cycle) Patients getting older would not have an accompanying increase in the probability of dying (but this doesn’t mean that it matters; remember, always a comparison...) (This is often called competing risks) 17 / 37 Probabilities vary according to time in model Some transition probabilities change as people get old In other words, transition probabilities can be a function of cycle In the HIV example, the probability of state D (death) should be higher in latter cycles Straightforward to implement 18 / 37 Probabilities vary according to time in model Two absorbing states: death due to HIV and death due to aging (non-HIV) Could have kept one death state while increasing the probability of dying 19 / 37 Probabilities vary according to time in a state Different situation: transition probabilities depend on time in state Example: Probability of dying increases according to how long a person has had AIDS This is a harder situation to handle with the type of Markov models we have covered We do not have a way to keep track of people moving from state to state (in the HIV example, we can only keep track of people in state A) We need to relax the Markov assumption (memoryless property) 20 / 37 Relaxing Markov assumption Example (Chapter 2 of Briggs et al, 2006): probability of dying of cancer after a cancer recurrence Patients can have a local or regional recurrence and then remission But after remission, the probability of death should depend on whether the recurrence was local or regional and the time in remission We need to find a way to keep track of time in remission 21 / 37 Partial transition diagram No memory of time and the type of remission once in remission 22 / 37 Relaxing Markov assumption We can follow a standard trick: adding states Remember that we “store” costs and benefits in health states and we can add as many states as we want (although it increases the complexity of the model) Instead of one remission state, we could have remission states for each type of recurrence 23 / 37 Adding remission states Now the transition probability from remission to death could be different depending on the type of recurrence Costs and benefits can also be different 24 / 37 We now need to solve the time issue We have made remission dependent of the type of recurrence But we wanted to also consider the time in remission because it affects the probability of dying We follow the same trick: add more states that reflect time 25 / 37 Health states that reflect time 26 / 37 Why does it work? Patients cannot stay in the new remission states for more than one cycle (no arrow to the same health state). These type of health states are called tunnel states We have added memory to the Markov model. We now know that patients in, say, “Remission after LR, Year 2” have been cancer free for two years (no other way to get to this state) We now can make the probability of death different for each tunnel state, which means making the probability different according to time in state Of course we could add more tunnel states. The only cost is the added complexity of programming the model Knowledge of the disease (and data availability) guides choices Note that the time in a tunnel state is guided by cycle length 27 / 37 Tunnel states Tunnel states have a very good name; think of them as tunnels will tolls A tunnel is the only way to go from state A to B They impose costs and benefits that last only one cycle If you are in B, that means that you came from A one cycle ago Markov models with some sort of memory are sometimes called semi-Markov processes or models 28 / 37 Big picture The extensions we have covered could be called “games you can play by adding health states” They are relatively easy ways to make Markov models more realistic The methods we have covered so far are extensively used But for many types of problems, we need still need different tools 29 / 37 Patient-level simulations Patient-level simulations (or microsimulations) model the transitions of an individual patient rather than a cohort Very useful to make transition probabilities dependent on patient history, not just cycle time We could take into account that the probability of an event depends on age, sex, disease severity, time with the disease, and so on These models are all about memory, so we do not have to worry about the Markov assumption Some of these models take into account how people interact with each other (key aspect of infections diseases) 30 / 37 Discrete event simulation (DES) Origins in operations research (design and optimization of industrial processes). As a result, they use different terminology Like Markov models, events happen in time (discrete periods, like cycles) It models entities (e.g. patients, hospitals) that have attributes (e.g. age, sex, disease history) Attributes can change in the model and they determine how entities interact with environment and events Events are “things” that could happen to an entity in an environment (e.g. death, infection, MI, stroke) Costs and benefits can be added and DES models capture time spent in events 31 / 37 Example Cost effectiveness of improving ambulance and thrombolysis response times after myocardial infarction (Chase et al, 2006) “Interventions: Improving the ambulance response time to 75% of calls reached within 8 minutes and the hospital arrival to thrombolysis time interval (door-to-needle time) to 75% receiving it within 30 minutes and 20 minutes, compared to best estimates of response times in the mid-1990s” 32 / 37 DES model 33 / 37 Disadvantages It may not be obvious but more realism is not always better. Remember Einstein: Everything should be made as simple as possible, but not simpler More realism means a lot more data is needed They easily turn into “black boxes” (a note on model calibration) How do we get transition probabilities that depend on age, sex, disease history? Many times the type of model follows the data. Given the data we have, what is the best model we can possibly build? Sensitivity analyses are more difficult (specially probabilistic sensitivity analyses) Computational burden could be a problem 34 / 37 Infectious diseases and vaccinations (static versus dynamic models) Microsimulation models have a clear advantage when considering infectious diseases and vaccinations studies The likelihood of contracting an infection depends on the number of people already infected Similarly, vaccines can be effective even if not all people are vaccinated (herd immunity) In static models, the rate of infection is fixed. In dynamic models, rate of infection is not fixed and may depend on the number of people already infected (i.e. possible to model epidemics) No easy way to add these features in Markov models Discrete event simulations are useful in these situations; agent-based models are a variation 35 / 37 More on types of models Adapted from Hay (2004) A note on Net Logo, Arena, and @Risk 36 / 37 Summary Markov models are extensively used in CEA Easy to make transition probabilities depend on time (cycle) Relatively easy to add more health states to incorporate memory into Markov models –but it gets complicated with too many health states In certain problems, we need to make probabilities conditional on characteristics and history of patients Microsimulation models are needed for infectious diseases and vaccination problems Field is quickly evolving (classification and naming of models will become clearer) 37 / 37