The National Medication Price-Negotiated Policy
In order to further regulate the price of anticancer medication and alleviate the financial burden of cancer patients, the Chinese government launched the NMPNP for anticancer medication in 2017. The Ministry of Human Resources and Social Security of the People’s Republic of China organized experts involved many related medical majors, such clinical medicine, pharmacy, pharmaceutical economics, and medical insurance, to formulate expected prices though evaluating the therapeutic effects, consumption of treatments and market price of anticancer medications. Subsequently, based on the expected prices, the experts negotiated the prices, types and payment standards of anticancer medications with pharmaceutical companies. If the negotiation was successful, the anticancer medications would be permitted into the NRDL, and the final prices of anticancer medications were regarded as the uniform payment standard throughout the country. In 2017, 18 anticancer medications covering lung cancer, stomach cancer, breast cancer, colorectal cancer, lymphoma, and myeloma were permitted in NRDL through negotiations, and the average procurement price of anticancer medication was reduced by 44% [11]. Since the National Healthcare Security Administration was established, the Chinese government successively permitted a total of 39 anticancer medications to go into NRDL by NMPNP from 2018 to 2019, and the average procurement price of anticancer medication continued to drop by 65% [12]. Until 2022, there were 76 anticancer medications in the NRDL [13].
Moreover, as a national compulsory policy, the National Healthcare Security Administration required the price of anticancer medication in all the Provincial Reimbursement Drug Lists (PRDL) must be updated according to the annual negotiated results within confined time, and public hospitals throughout the country were required to purchase those medications at the negotiated price [10]. In this case, in September 2017, the Shandong’s PRDL was updated the price of anticancer medications based on the negotiated results of NMPNP 2017, which was intended to alleviate the financial burden of cancer patients and increase the access of anticancer medication as soon as possible.
Study design
As one of the provinces that has a large population, Shandong province has 7.2% of the Chinese population and has 7.33% of China’s gross domestic product (GDP) [14]. Additionally, the morbidities of gastric cancer, thyroid cancer, and breast cancer are separately higher than other provinces in China [15]. So, we took the implementation of NMPNP in 2017 in Shandong province as a quasi-experiment and took September 2017 as the intervention point in this study when Shandong’s PRDL was updated based on the negotiated results of NMPNP 2017. We also used an ITS design, which covers the complete claim data records of cancer patients from 2016 to 2021.
Data sources
As the basic medical insurance that covers the largest insured population in China, Urban and Rural Resident Basic Medical Insurance (URRBMI) covers 74.03% of the Chinese basic medical insureds [16] including elders, children, and low-income rural and urban area residents [17]. According to the population and economic levels rank [18], we purposively sampled four cities in Shandong province (The sampling principle and data sources are shown in Additional file 1). We collected and aggregated the monthly medical claim data of 45,895 cancer patients who enrolled in URRBMI from the municipal healthcare security administrations of the four cities. The medical claim data of cancer patients are the related outpatient and inpatient medical costs of cancer treatments from January 2016 to December 2021. It contains healthcare service utilizations of outpatient and inpatient care, medical costs of outpatient and inpatient care, OOP expenditure of outpatient and inpatient care, and medication costs of outpatient and inpatient care.
Outcome measures
In order to accurately quantify the access and affordability of anticancer medication, we assessed several outcome variables during study observation including outpatient and inpatient care visits per capita (Eq. 1), proportion of OOP expenditure in outpatient and inpatient medical costs (Eq. 2), and proportion of medication costs in outpatient and inpatient medical costs (Eq. 3). Regarded the China consumer price index (CPI) in 2016 as the base year, all data in the study was discounted by the CPI [14]. The skewed distribution of data in this study was log-transformed.
$$The\ outpatient/ inpatient\ care\ visits\ per\ capita=\frac{number\ of\ outpatient/ inpatient\ care\ visits\kern0.75em }{number\ of\ outpatient/ inpatient\ patients}$$
(1)
$$The\ proportion\ of\ OOP\ expenditure\ in\ medical\ costs=\frac{total\ OOP\ expenditure\kern0.75em }{ total\ medical\ costs}\times 100\%$$
(2)
$$The\ proportion\ of\ medication\ costs\ in\ medical\ costs=\frac{total\ medication\ costs\kern0.75em }{total\ medical\ costs}\times 100\%$$
(3)
Statistical model and analysis
We used the ITS analysis to assess the change of access and affordability of anticancer medication in this study. The new itsa command contains the two ordinary least squares (OLS) regression in the Stata packages prais and newey, which can perform the ITS analysis for multiple inventions with single or multiple groups. In our ITS analysis, a segmented OLS regression model with a Newey-West test was separately used to assess whether NMPNP can reduce the financial burden of outpatient and inpatient care [19]. The segmented regression model we adopted is shown below:
$$Y_t=\beta_0+\beta_1T_t+\beta_2X_t+\beta_3T_tX_t+\varepsilon_t$$
Y
t is the dependent variable we measured at every monthly point t, Tt is the time series variable representing the time in months since the start of observation until to the time t, Xt is a dummy variable representing the intervention point (preintervention period is 0, post-intervention period is 1), Tt Xt is an interaction term of the time and intervention, and εt is the residual term representing the unknown variation component of the regression model. β0 represents the baseline level; β1 represents the baseline trend prior to intervention; β2 represents the immediate level change after the intervention compared to the preintervention; β3 represents the trend change after the intervention compared to the preintervention; and β1+ β3 represents the trend after the intervention [20].
The actest command was used to test the autocorrelation of time series [21], and the autocorrelation results were both present at lag 1. Due to the time trend of proportion of OOP expenditure in outpatient and inpatient medical costs in each month displayed conspicuous seasonal effect, and thus we used OLS regression model to perform the seasonality adjustment [22]. STATA 16.0 was used to perform all statistical analysis. Two-sided P < 0.05 was considered statistically significant.
Source link
