Pharmacoeconomic Study On The Three Chemotherapies For Advanced Pancreatic Cancer In China Based On Markov Model

Background:The incidence and mortality of Pancreatic cancer in recent years showed an increasing trend in the global scope.In the rankings of China’s new cancer cases and deaths,the incidence of pancreatic cancer ranked ninth,but the mortality rate was second in 2015.Because of the low diagnosis rate of pancreatic cancer,it often comes to an advanced stage,which brings huge disease burden and economic burden to the people.In the face of the increasing medical expenditure budget and the increase of the demand for drugs,pharmacoeconomic evaluation can promote the rational allocation of medical resources.Purpose:Although advanced non-small cell lung cancer,gastric cancer,breast cancer,ovarian cancer and other malignant tumors have been carried out pharmacoeconomic evaluation(PE)in recent years in China.However,there is no PE of the first line chemotherapy for advanced pancreatic cancer(APC).Therefore,this study based on the principle of the Markov model and pharmacoeconomics,to carry out the PE of three schemes for the treatment of advanced pancreatic cancer:gemcitabine(G),gemcitabine plus oxaliplatin(GM)and gemcitabine + S-1(GS),in order to provide decision-making basis for domestic patients with APC.Research contents and methods:(1)From the patient’s point of view,determine the cost of drug for three treatment options according to the price standard of the Jiangsu Province People’s Hospital.The costs of adverse reactions and second-line treatment come from the literature.Then the direct cost of the G,GM,GS per treatment cycle.Collect the data of clinical randomized trials published in core journals at home and abroad.By the principle of DE ALE method,the transition probability is estimated based on the overall time(OS),progression free survival(PFS),duration of response(DOR).Utility weights used in the model were assigned to each health states based on the Romanus D’s study.(2)According to t response evaluation criteria in solid tumors(RECIST)version 1.1 of WHO,four state advanced pancreatic cancer development were divided:mitigation,stability,progress and death.Then establish a multistate Markov decision tree model by Treeage Pro2011.The Markov model was used to simulate the distribution of different states of the patients with APC after 5 years.The results were corrected by half cycle correction,and take cost effectiveness analysis based on simulation results.(3)Take the single factor sensitivity analysis and probabilistic sensitivity analysis of the cost effectiveness results.Single factor analysis with Tornado chart can express the influence of various parameters on the degree of comparison results.Probabilistic sensitivity analysis evaluates the results in scatter diagram by Monte Carlo simulation.Result:(1)The result of Roll Back analysis shows that the G scheme is the optimum scheme.The G project cost effectiveness ratio(C/E)is(?)145228.52/12.26QALMs.The cost of obtaining a QALM is(?)11845.72;the C/E ratio of GM is(?)154783.88/11.39QALMs,The cost of obtaining a QALM is(?)13589.45;the C/E ratio of GS is(?)315485.28/23.26QALMs,The cost of obtaining a QALM is(?)13563.44.The threshold standard WTP set for this study is 12563 yuan/month,so the G scheme is the preferred scheme.(2)Through the cost effectiveness analysis,the incremental cost effectiveness ratio(ICER)of the GM and G schemes is negative(-10999.89),which belongs to the absolute inferiority scheme;and the ICER value of the GS and G schemes is positive(15479.64).Thus,within the threshold range,the G scheme is the preferred scheme.(3)The results of sensitivity analysis.A comparison between the G and GM,the results of sensitivity analysis and probabilistic sensitivity analysis of the single factor influence is not very stable,easy to receive some parameters.So it’s unable to determine which one is the optimal scheme;Between the G and GS,the single factor sensitivity analysis and probabilistic sensitivity analysis result is relatively stable,the parameters had little effect on the results come,and the G scheme is an optimal scheme;Between the GS and GM,the single factor sensitivity analysis and probabilistic sensitivity analysis is stable between GM optimization scheme,but the advantage is not obvious.In conclusion,although the sensitivity analysis results can not determine what kind of program is the best,but the cost effectiveness accectablility curve(CEAC)and net benefit curve(NBC)showe that G is the best project in the range of WTP by Monte Carlo simulation.