From January 1, 2020, the price of SAC/VAL has been reduced from ¥ 17.33 per 100 mg to¥ 9.95 per 100 mg, greatly improving the drug use rate and bringing good news to HF patients in China.
As far as we know, this study is the first time to evaluate the economy of the treatment of HFrEF patients with SAC/VAL after national negotiation price in China. China HF study, CN HF study and HERO study make the epidemiological data of HF patients in China more comprehensive and authoritative[21,22]. With the exception of the parameters related to SAC/VAL and the hospitalization rate of enalapril, all data are almost from the epidemiologic data of HF patients in China. Our analysis based on Markov model shows that: at the national negotiation price (¥9.95 per 100mg), SAC/VAL was associated with clinical benefit and may be cost-effective compared with an ACEI (the current standard of care) in patients with HFrEF.
With the exception of the relative risk of dying for SAC/VAL, one-way sensitivity analysis reveals that ICER is all lower than ¥ 212676 per QALY. The relative risk of dying for SAV/VAL has the most important driver on the cost-effectiveness, which means that avoiding deaths in the SAC/VAL treatment group lead to more QALYs compared with enalapril. The second driver is the health utility value of the stable heart failure and the relative risk of hospital for SAC/VAL because of the additional gained QALYs by avoiding hospital and living longer, the price of SAC/VAL has little impact on the ICER.
The scenario analysis shows that the pre negotiation price of SAC/VAL has a decisive effect on the cost-effectiveness and the negotiation price of SAC/VAL has little impact on the cost-effectiveness. The long-term impact of SAC/VAL including reducing death, reducing the number of hospitalizations and prolonging life expectancy is the fundamental factors to promote the use of drugs.
In our model, the cost of drugs and hospitalization, HR of hospitalization, rehospitalization and all-cause death and the utility value have a distribution, but those data including hospitalization, rehospitalization and all-cause death in enalapril do not have a distribution in the model, because these are not available, but all-cause death, in-hospital death, and rehospitalization are based on large numbers and inherent uncertainty may be small. The PSA of our model shows that probability of SAC/VAL being cost-effective was about 53% at a WTP threshold of ¥212676 per QALY. It may be that most of our data are from the epidemiological data of HF patients in China. Compared with PARADIGM-HF trial, all-cause mortality, in-hospital mortality and readmission rate of HF patients in China are relatively high, which is why our model predicts lower survival rate in 27 month and mean survival time. The cost of drugs in our model may be higher than actual cost, because according to PARADIGM-HF, the average dosage of SAC/VAL and enalapril is 375 ± 71 mg, 18.9 ± 3.4 mg respectively, which is almost two times of actual dosage. The reason that we take the dosage from PARADIGM-HF is to use the HR of hospitalization, rehospitalization and all-cause death from PARADIGM-HF in our model and to not reduce the effect of SAC/VAL.
The results are robust and reliable in general in the sensitivity analyses for several reasons; First, HF is a chronic disease with high morbidity and mortality whose median survival time is less than 5 years[31], within a reasonable price range, treatment that reduces death and hospitalization can be cost-effective; Second, relieving the progress of HF may contribute to the relative improvement of the overall quality of life, because the decline of the quality of life is related to the severity of the disease; Third, the high cost of hospitalization is saved, including general ward care and intensive care unit, and the implantation of high consumable heart assist device is reduced.
In our model, the use of SAC/VAL could lead to more than one year of life for each patient and significant cost savings by avoiding hospitalization. Because of the increase in life expectancy of 8-10 million HF patients in China[3], the benefits of the use of SAC/VAL could be significant. Considering that up to half of the HF patients were accompanied by the reduced ejection fraction, a considerable number of patients were likely to receive SAC/VAL treatment. According to our model, it was assumed that 10 million HFrEF patients in China are treated with SAC/VAL. This strategy could reduce patients in hospital by 300 thousand, deaths by 280 thousand, and the medical cost of the hospitalization by ¥9 billion in two years, which could greatly reduce the burden of Chinese medical security system.
There have been many studies abroad to analyze the cost-effectiveness of the treatment of heart failure with SAC/VAL. Solomon et al. considered it to be cost-effective by calculating the ICER of $45017 per QALY at the cost of $375 per month for SAC/VAL and $0.96 per month for enalapril [32]; The team in Dutch also considered it cost effective to spend £ 5.25 per day for SAC/VAL and the ICER of £ 19133 per QALY[15]; Germany also announced that it was cost-effective to spend £ 5.33 a day for SAC/VAL and the ICER of £ 26278 per QALY[33] . It can be seen that the treatment of heart failure with SAC/VAL in many countries was cost-effective, which is consistent with our results. Compared with our model, the cost including drugs and hospitalization and the utility value is much higher than that in our model. In their sensitivity analysis, the price of SAC/VAL has a driving effect on whether it is cost-effective, while our model shows that: SAC/VAL on death has a driving effect on whether it is cost-effective.
There are the following limitations in our study: First, we did not consider the cost of the hospitalization for non-HF in our model, but in the PARADIGM-HF trial, the HR of the hospitalization for non-HF was 0.92 (0.85-0.99), and our analysis results could be conservative; Second, because we can't achieve the data of parameters related to SAC/VAL and health utility values of each transfer state of HF patients in China, we use data form foreign trials to simulate, which may lead to some racial bias in the simulation results;Third, we assumed that HF patients in China could tolerate the recommended dose of each drug, regardless of the adverse events, in the PARADIGM-HF trial, the most common adverse events included hypotension, heart failure, hyperkalemia, renal damage and cough, but the proportion of discontinuation of the trial (10.7%) due to the occurrence of the adverse events in the SAC/VAL was lower than that of enalapril (12.3%), which further emphasized our analysis could be conservative;Fourth, the cost of other treatments such as B blocker, aldosterone receptor antagonist, diuretic, digitalis, ICD and cardiac resynchronization were not considered in the calculation of drug cost, but the PARADIGM-HF trial showed no significant difference between the two groups;Fifth, the NYHA classification was not incorporated in the model because these data were not available;Sixth, Productivity losses were not taken into account for this model. because this study was from the health care payer’s perspective and not from the societal perspective. Nevertheless, most patients werelikely to be retired when HF is diagnosed , the effect of this is likely to be low.