Time-homogeneous cohort state update

Updates the cohort distribution when the same transition probability matrix is applied in each cycle.

Signature

s_(t+1) = s_t P
Inputs
InputsDefinitionUnit
s_tA row vector containing the proportion of the cohort in each health state at the start of cycle tproportion of cohort
PA square matrix in which each entry gives the probability of moving from one state to another during one cycleprobability per cycle
Output
s_(t+1)A row vector containing the proportion of the cohort in each health state at the start of the next cycleproportion of cohort
  • t The numbered model cycle (cycle)

Function

Cohort state-transition function

Maps a cohort's current state distribution and transition probabilities to its state distribution in the next model cycle.

Try this function

Implementations

  • Excel

    Update the cohort distribution

    Excel uses matrix multiplication to return the next-cycle state vector. CurrentStateVector and TransitionMatrix are named ranges.

    =MMULT(CurrentStateVector,TransitionMatrix)

Assumptions

  • Mutually exclusive and exhaustive states

    Every cohort member occupies one and only one modelled health state in each cycle.

  • Markov property

    Transition probabilities depend on the current state unless relevant history is represented through additional states or another model structure.

  • Fixed cycle length

    Every update represents the same stated length of time.

  • Time-homogeneous probabilities

    The transition matrix P remains constant across cycles for this formula variant.

  • Valid transition matrix

    Each transition probability is between 0 and 1, and every row of P sums to 1.

Worked examples

  • Three-state cohort after one cycle

    A cohort begins entirely in Stable disease. After one cycle, 80% remain Stable, 15% are Progressed and 5% are Dead.

    s_0 = [1,0,0]; P = [[0.80,0.15,0.05],[0,0.85,0.15],[0,0,1]]; s_1 = [0.80,0.15,0.05]

Common errors

  • Using probabilities from the wrong time interval

    Applying an annual transition probability directly in a monthly-cycle model changes the intended transition process. Convert or estimate probabilities for the model's actual cycle length first.

Sources

  • State-transition modelling good-practice report

    Siebert U, Alagoz O, Bayoumi AM, et al. State-transition modeling: a report of the ISPOR-SMDM Modeling Good Research Practices Task Force-3. Value in Health. 2012;15(6):812–820.

    View source

  • Practical guide to Markov models

    Sonnenberg FA, Beck JR. Markov models in medical decision making: a practical guide. Medical Decision Making. 1993;13(4):322–338.

    View source

Canonical Identity

Stable URI · Machine-readable · Resolvable · CC BY 4.0