| BackgroundThe health planning is an important means for the government to adjust allocation of health resources and achieve development of health fairly and efficiently.The core of health resource planning is the optimal allocation of health resources.China's central and local governments have carried out many health plannings and those plannings have played an important role in optimizing the allocation of health resources.However,because the existing health planning lacks a reasonable theoretical framework and model,which brings that the allocation of health resources do not match with the actual needs of residents.During the development of deepening the reform of the health system and promoting the construction of healthy china,how to use the health resources planning still needs to make efforts in the making plannings scientificly and implementing planning effectively.The existing planning is mainly based on the analysis of the current situation and trend of health resources.The researches on the health needs of residents are insufficient,which affects the connection between the allocation of health resources and the health needs of residents.Those are not conductive to the reform of the supply of health services and the construction of healthy china and deepening the reform of health system.Health needs of residents are the basis for optimizing the allocation of health resources.And it is of great significance to conduct more scientific analysis and predict the total and structure of health demands.The existing related researches show that the demand forecasting methods include trend extrapolation method,regression model method and micro-simulation model.Most domestic researches on the forecasting health demands use the method of trend extrapolation and regression model,which use the data of the supplier.The international researches use the method of micro-simulation model,and the data of this model come from the demand side.The method of trend extrapolation has less consideration for the influencing factors of future health demands,and also affects the effect due to the need of long-term data.The method of regression model cannot consider the relationship between parameters.Therefore,how to allocate health resources on the basis of the demand is a need to further deepen the research field in the future.The health-relate macroeconomic policies in the future have changed a lot,such as urbanization,population aging,the policies of fertility,healthy lifestyles,economic policies and medical insurance policies,policies of setting up a hierarchical medical system.Individual factors are also changing,such as age,education,marriage,et al.These changes of polices and individual factors will affect the total demand and the structure of future residents.How to quantify the impact of these macro policies on the total demand and structure of residents?In order to meet the needs and structure of residents and ensure the effective connection of health needs with health resources,how should different types of health resources be configured?What are the strategies for adjusting different types of health resources based on the current status of resources?Answering the above questions could provide relevant core information for optimizing the allocation of health resources in the future.ObjectivesThe general objectives of this study are to predict the total and structure of health demands from 2020 to 2030,and provide core information for the optimizing allocation of health resources in order to meet the health needs of residents and improve the health of residents.The specific aims include:(1)to analyze the current status of residents' health needs and health service systems;(2)to build a micro-simulation model of health demands;(3)to predict the total and structure of health demands from 2020 to 2030 by using micro-simulation model;(4)to estimate the total and structure of health resources from 2020 to 2030;(5)to quantify strategies for adjusting health resources and provide the policy recommendationsMethodsIn this study,the source of data is mainly divided into three survey data.Those data consist of the 2014 health service survey in Liaoning Province,the survey data on the doctors' efficiency and the patients' reasonable flow survey data.Multistage stratified random sampling method was employed in the health service survey.Fourteen cities were selected in Liaoning Province,and one district or one county was randomly selected in each city.Next,4-6 communities/townships were randomly designated from each district/county,and two committees/villages were randomly identified form each community/township.Then,60-70 households were randomly selected in each committee/village.Finally,a total of 27477 household members of 9434 households in 158 committees/villages of 84 communities/townships were investigated.Survey of efficiency of doctors.Typical sampling was used in the survey data for doctors'work efficiency.Five cities were selected.One provincial/municipal hospital and one county hospital were selected in each city.Finally,a total of three provincial hospitals,three municipal hospitals and five county hospitals were selected.According to the doctors' sampling principle,2984 doctors were investigated in those hospitals.Survey of patients' reasonable flow.Based on the hospitals participating in the survey for doctors' efficiency,two provincial hospitals,two municipal hospitals and two county hospitals were randomly selected.Finally,a total of 6 hospitals were selected for the survey of patients' reasonable flow.According to the principle of case extraction,a total of 4619 cases were selected.The research mainly adopts the micro-simulation model,that is,the model is divided into four steps.The first is to clarify the main parameters and outcome indicators of the micro-simulation model,then the model used the correlation function between the main parameters and output indicators and the change probability function of the main parameters,also predicted the output indicator values in the future by R language programming.Fourth,the model verification is performed using the existing statistical yearbook data from 2015 to 2017 and the chronic disease survey data of Liaoning Province.Secondly,the time series model,the mixed Logit regression model,and the contingency table function were used to clarify the change probability function of the main parameters in the micro-simulation model in the future.Then,logistic regression,passion regression and negative binomial regression were used to determine the correlation function between the main parameters and the output indicators in the micro-simulation model.Finally,the study uses the health service demand to estimate the total amount and structure of health resources by combining with the efficiency of different health resources.Results1.Building a micro-simulation model.According the purpose of the micro-simulation model and the conceptual framework of the study,the main parameters and output indicators in the model were identified.Based on the data of the survey,data from various statistical yearbooks,and the target of macroeconomic policies,the contingency table function and the polynomial Logit regression equation were used to calculate the change of the independent variable probability in this model.Based on the individual data of the health service survey in Liaoning Province in 2014,this model used the correlation between the independent variable and the outcome index by different multi-factor analysis equations,as well as used the probability variation functions and multi-factor analysis equations.When constructing the correlation function,90%of the sample data was used,and the remaining 10%of the sample was used for the check of the function.The prediction threshold and accuracy of the function were clarified.Then,the main parameter probability functions were used to update the main parameters,and the correlation functions were used to bring the updated value of parameters in 2015-2030 to the result indicators.According to the critical value of the correlation function,this study could determine whether the individual result event occurred,and calculate the outcome indicators for 2015-2030 finally.The model was verified using the results of the chronic disease survey from 2015-2017 and the statistical yearbook data.The verification results show that the model was valid.In summary,the micro-simulation model was effective in predicting health demand.2.The micro-simulation constructed by this study was used to predict the total health demand from 2020 to 2030.According to the different methods of parameter in the model,the model predicted the two-week visit rate and annual hospitalization rate of the residents in the future.As well as the two-week visit rate and annual hospitalization rate by urban-rural and age groups were predicted in four different scenarios.In 2020,the two-week visit rate and annual hospitalization rate of residents in Liaoning Province were 15.2%-17.6%and 12.9%-15.1%,respectively.In 2025,the two-week visit rate and annual hospitalization rate of residents were 15.6%-19.0%and 14.6%-18.8%,respectively.In 2030,the two-week visit rate and annual hospitalization rate of residents were 17.1%-21.1%and 18.7%-20.6%respectively.3.The demand structure was mainly divided into six different scenarios,which included three types of trends,such as the status structure of demand maintained in the future,rational structure of demand(the patients were diverted from the large hospital to PHCIs(Primary Health Care Institutions)),deteriorated structure of demand(the patients were concentrated in the large hospitals).According to six structures of demand and the efficiency of different types of beds and practicing(assistant)physicians,the number of beds and practicing(assistant)physicians per 1000 population were calculated for different types of institutions.4.The number of beds and practicing(assistant)physicians per 1000 showed an increasing trend between 2020 and 2030.Compared with the current situation,the number of beds and practicing(assistant)physicians showed opposite trends.Take the situation of diverting patient flow to PHCIs as an example.In 2020,the number of beds and practicing(assistant)physicians per 1000 are 4.04-5.49 and 2.34-4.05,respectively.In 2025,the number of beds and practicing(assistant)physicians per 1000 are 4.38-6.48 and 2.27-4.42,respectively.In 2030,the number of beds and practicing(assistant)physicians per 1000 are 5.11-6.92 and 2.68-4.95,respectively.The structure of the beds and practicing(assistant)physicians are shown in the case of patient flow scenario 3.In 2030,the number of beds and practicing(assistant)physicians per 1000 in PHCIs are 0.98-1.16 and 1.58-1.82,respectively.The number of beds and practicing(assistant)physicians in county hospitals are 1.24-1.37 and 0.45-0.52,respectively.The number of beds and practicing(assistant)physicians inmunicipal hospitals are 1.30-1.54 and 0.61-0.68,respectively.The number of beds and practicing(assistant)physicians in provincial hospitals are 0.56-0.66 and 0.33-0.38,respectively.5.Compared with the current situation,the number of beds per 1000 population in 2030 is reduced by 0.39-1.75,the percentage of the reduction is about 5%-20%.The number of practicing(assistant)physicians per 1000 is increasing by 0.22-1.17,the proportion of the increase is about 10%-40%.In the same outpatient/inpatient context,the number of beds for the reasonable flow is minimized and the number of practicing(assistant)physicians is the highest in this scenario.6.For the number of beds and practicing(assistant)physicians in different types of institutions,the situation of patient flow has a greater impact on them.In the scenario 3 of 2030,it is necessary to increase the number of beds and practicing(assistant)physicians in the PHCIs and increase the number of practicing(assistant)physicians in county hospitals and municipal hospitals.There is a need to reduce the number of beds in municipal hospitals and provincial hospitals and the number of practicing(assistant)physicians in provincial hospitals.In 2030,it is necessary to increase the number of beds and practicing(assistant)physicians per 1,000 population in PHCIs,with an increase of 0.11-0.36 and 0.87-1.09,respectively.In 2030,the number of practicing(assistant)physicians in county hospitals should be increased by 0.05-0.13.For the municipal hospitals,in 2030,the number of beds per 1,000 population needs to be reduced,the reduction is 0.33-0.57,and the number of practicing(assistant)physicians per thousand population needs to be increased,increasing by 0.03-0.13.It is necessary to reduce the number of hospital beds and the number of practicing(assistant)physicians,the reductions are 0.10-0.20 and 0.01-0.05 respectively.Conclusions and Policy implicationsThe main conclusions of this study are as follows.(1)To predict the health needs in the future by the micro-simulation model,the two-week visit rate and annual hospitalization rate of residents in Liaoning province show an upward trend.(2)The main task was to control the total number of beds and adjust the structure of beds in the future.For practicing(assistant)physicians,the main task was to increase the total number and adjust the structure of physicians.(3)For the number of beds and practicing(assistant)physicians in different types of institutions,the situation of patients' flow had a greater impact on them and trends in different scenarios were different.The adjustment of resources brought by rational patient flow is conductive to the effective implementation of the grading diagnosis,and is also the direction of future development of health policy.According to the above conclusions,the following recommendations are proposed.(1)Ensure the implementation of hospitals function by quantifying the assessment indicates of different levels of hospitals positioning.(2)To play the role of financial input in guiding the hospital to implement functional positioning.(3)To pay attention to the medical insurance in promoting the rational flow of patients.(4)The government should take various measures to enhance the capacity of PHCIs in order to promote the implementation of grading medical treatment policies.(5)The number of beds in the PHCIs is transformed into family beds and pension beds with the role of medical function.(6)Strengthening cooperation among hospitals at different levels and combinations between hospitals and primary health care institutions.Innovations(I)At present,most studies on health resource allocation are based on supplier data.This paper is to allocate health resources based on the health needs of the demand side in China.First,this study constructs a micro-simulation model to predict the future health needs of residents,and then uses the health service demand method to turn health needs into health resources.(2)This study applies micro-simulation model in the field of health resource allocation in China.The application of micro-simulation model in foreign counties shows that the model can solve the core technology in the process of health resource allocation.This model can use the individual data to predict the total demand and structure of residents in the future.In the process of forecasting demand,the model can integrate multi-level,multi-factor and non-linear dynamics of demand factors,and can effectively integrate individual-level demand-influencing factors and macro-policy-level demand-related factors.(3)This study measures the total allocation of health resources and the allocation of health resources in different types of institutions under the scenario of reasonable patient flow,and promotes the effective implementation of the grading diagnosis and treatment policy.At the same time,it provides detailed indicators for the implementation of health resource planning and operational indicators for planning assessment.Limitations(1)Due to data limitation,the data used in the individual parameter of the micro-simulation model is derived from national or provincial related data.These parameter probability functions do not better reflect the differences between each city,urban and rural areas,and different age groups.In the future,relevant investigations of these parameters will be carried out in Liaoning Province to obtain the corresponding probability function of this parameter.(2)Based on the demand based on the micro-simulation model,this study needs to use the efficiency indicators of various types of health resources when using the health service demand method to convert demand into resources.However,the efficiency index is mainly obtained by combining the changes of health resource efficiency,status quo and expert consultation,etc.so the health resource efficiency indicators are relatively subjective. |
PDF Full Text Request