Effects of the essential medicine policy in China on outpatient service use and medicine cost: a secondary analysis of 5 year panel data

Abstract

BackgroundWith the introduction of the essential medicine policy in 2009, the Chinese government mandated the distribution of medicines to health-care facilities at a low cost and without profit to increase use of medical service by improving equity in health-care access and reducing patients' medical costs. The aim of this study was to examine the effect of the essential medicine policy on outpatient service use and medicine cost MethodsThis is a secondary analysis of annual national surveillance system data of the average medicine cost per outpatient visit (COST) and the number of outpatient service usage (USAGE) for 2008–12, collected among 32 514 grassroots-level health-care facilities in 2645 counties and 31 provinces in China. The policy was implemented in 2009 to cover about 27% of facilities, and then up to 53% in 2010, up to 75% in 2011, and finally scaled up to all facilities in 2012. We use four-level hierarchical random effects models to address possible dose-response effects of the policy and possible variation of such effects at levels of province, county, and facility. All analyses were adjusted for potential effects of time period, resource availability, and inflation, as appropriate on the outcomes, plus necessary data transformations to account for skewed distributions. As a secondary analysis with the data holder and collector being coauthors, the study did not require further approval nor register. FindingsHaving adjusted for health labour force and assets of facilities, both outcomes differed significantly by regions in China. East China, with the highest urban population, showed significant COST reduction as length of policy exposure increased (from ¥26·52 to ¥19·72 by an annual reduction factor e−0·074t). Similarly, USAGE increased significantly from 15 214 visits to 17 997 (annual increase e0·042t). In contrast, west Chinese facilities had significantly lower COST before policy implementation (¥16·62) with no reduction over 4 years of policy exposure. They also had significantly lower USAGE than facilities in east China before the policy implementation (estimated 6884 visits) with a significant reduction over the policy period (down to 5813 by an annual reduction rate at e−0·042t, p=0·023). Facilities in central China had significantly lower COST and USAGE than that in east China before policy implementation and showed no significant dose-responsive effects (baseline COST at ¥19·07 with an annual change factor e−0·031t, p=0·126; and baseline USAGE at 8263 visits with an annual change rate at e−0·038t, p=0·061). Furthermore, rural facilities had significantly lower estimates in COST and USAGE than urban facilities before policy implementation (an average of ¥2·99 less COST and 2419 fewer visits with p<0·0001 at township facilities, and of ¥1·75 less COST and 2490 fewer visists with p<0·0001 at town-centre facilities). The reduction rate in COST was significantly less in rural facilities than in urban facilities. There were significant variations of the dose-response effects among provinces and counties. InterpretationAlthough the essential medicine policy tended to have the expected effects on COST and USAGE, large variation of such effects exist between west and east parts of China, between rural and urban facilities, and between the period before and after implementation of the policy. The effects of the policy were particularly limited in western and rural areas. Further research of the contributions of each facility's contextual and micro-level factors to the present results are needed. The data were collected from the health administrative system, which might containt reporting errors and missing information at some time points. Lack of facility-level and county-level contextual factors in the data made it impossible to separate the effect due to policy intervention from that due to area contextual and micro-level factors. The study was strengthened by a large national data with advanced analytic method to address cluster effects in a pseudo stepped-wedge design for assessing the policy effects. FundingCentre for Health Information and Statistics, the National Health and Family Planning Commission of China and the China Medical Board.