34.6 Within-Cycle Correction (WCC)

Similar to the prior section Half-Cycle Correction (HCC) , Within-Cycle Correction is an adjustment to reduce overcounting of values accumulated over time.

What does WCC look like?

WCC accounts for values/rewards accumulated within a cycle based on the average percentage of the cohort in a health state at the beginning and at the end of each cycle.

The format of a Markov Model is usually in the Traditional format for HCC but to implement WCC you first change the Markov configuration in Tree Preferences. This adjusts the type of reward entries and how they are accumulated. The reward types are:

  • HCC: State rewards are init, incr and final.

  • WCC: State rewards are startup, cycle and event.

WCC Markov models calculate the rewards that would be accrued at the start of the cycle and the end of the cycle. Then it applies the average of those two calculations. In steps, the method does the following:

  • Calculate the reward at the current stage (t) based on the percentage of the cohort in the state. (Cycle reward at time t)

  • Calculate the reward at the next stage (t+1) based on the percentage of the cohort who are going to that state. (Cycle reward at time t+1)

  • The total reward at the current stage (t) is then: [ (Cycle reward at time t) + (Cycle reward at time t+1) ]/2

In this section we will use the model Markov Conversion Model WCC.trex to show the rewards after conversion.

To perform within-cycle correction:

  • Open Tree > Tree Preferences and in Calculation > Markov/DES select the option in the WCC section for "Within-Cycle Correction".

    • If your model is already a HCC model, review the warnings about converting the model to WCC.

    • Click "Apply and Close".

  • Select a Markov state node (for example Local Cancer).

    • The rewards should now show Startup, Cycle and Event rewards.

Review the next section on converting from HCC to WCC to see what is done to existing entries in the model.

Rewards for all payoff sets (cost, effectiveness, etc.) will be adjusted in the same way.

NOTE: Be careful when applying WCC using the steps above. Clicking the Half-cycle Correction button will delete any entries in the Initial and Final rewards and replace them with half the Incremental Reward. If you have any additional entries in those fields (particularly up-front costs in the initial reward), be sure to add those back into the model after applying WCC.

Reporting with Within-Cycle Correction is covered in the section Markov Cohort Analysis Output.