summary_plot_psa: Function to do deterministic sensitivity analysis

In sheejamk/MarkovModel: Decision Analysis Markov Model

Description

Function to do deterministic sensitivity analysis

Usage

summary_plot_psa(result_psa_params_control,
  result_psa_params_treat = NULL, threshold = NULL,
  comparator = NULL)

Arguments

result_psa_params_control	result from determnistic sensitivity analysis for first or control model
result_psa_params_treat	result from determnistic sensitivity analysis for the compartive markov mdoel
threshold	threshold value of WTP
comparator	the strategy to be compared with

Value

plot of sensitivity analysis

Examples

param_list<-define_parameters(cost_zido = 2278,cost_direct_med_A = 1701,
cost_comm_care_A = 1055,cost_direct_med_B = 1774,cost_comm_care_B = 1278,
cost_direct_med_C = 6948,cost_comm_care_C = 2059,tpAtoA = 1251/(1251+483),
tpAtoB = 350/(350+1384),tpAtoC = 116/(116+1618),tpAtoD = 17/(17+1717),
tpBtoB = 731/(731+527),tpBtoC = 512/(512+746),tpBtoD = 15/(15+1243),
tpCtoC = 1312/(1312+437), tpCtoD = 437/(437+1312),tpDtoD = 1,
cost_health_A = "cost_direct_med_A+ cost_comm_care_A",
cost_health_B = "cost_direct_med_B+ cost_comm_care_B",
cost_health_C = "cost_direct_med_C+ cost_comm_care_C",
cost_drug = "cost_zido")
A <- health_state("A", cost="cost_health_A+ cost_drug ",utility=1)
B <- health_state("B", cost="cost_health_B + cost_drug",utility=1)
C <- health_state("C", cost="cost_health_C + cost_drug",utility=1)
D <- health_state("D", cost=0,utility=0)
tmat <- rbind(c(1, 2,3,4), c(NA, 5,6,7),c(NA, NA, 8,9), c(NA,NA,NA,10))
colnames(tmat) <- rownames(tmat) <- c("A","B" ,"C","D")
tm <- transition_matrix(4, tmat, c("tpAtoA","tpAtoB","tpAtoC","tpAtoD",
"tpBtoB", "tpBtoC", "tpBtoD","tpCtoC","tpCtoD","tpDtoD" ), colnames(tmat) )
health_states <- combine_state(A,B,C,D)
mono_strategy <- strategy(tm, health_states, "mono")
mono_markov <-markov_model(mono_strategy, 20, c(1, 0,0,0),c(0,0,0,0),discount=c(0.06,0),param_list)
sample_list <-define_parameters(cost_zido ="gamma(mean = 2756, sd = sqrt(2756))")
param_table <- define_parameters_psa(param_list, sample_list)
result<-do_psa(mono_markov,param_table,10)
summary_plot_psa(result, NULL,NULL,NULL)

sheejamk/MarkovModel documentation built on Jan. 23, 2020, 2:44 a.m.