Cost and influencing factors of treatment for neoplasms in underdeveloped regions of China based on SHA2011 and Random Forest Model

Xiaobin Hu,Xinyao Che,Lei Qi,Feng Dou,Yini Zhao,Jianjian Li,Hongmei Du,Zhibin Ma,Chao Yang,Yaya Yang,Lei Wang

DOI: https://doi.org/10.21203/rs.3.rs-1943188/v1

2022-01-01

Abstract:Abstract Background: Cancer poses a severe health hazard and a heavy economic burden to mankind. This study, applying the System of Health Accounts 2011 (SHA 2011), aimed to calculate the economic burden of neoplasm in Gansu of China, and to find out problems in the entire financial flow of neoplasm treatment service delivery (including financing scheme, services provider, disease and patients distribution), and to explore influencing factors of hospitalization expenditure with neoplasms.Methods: Macro data was obtained from official yearbook and sampling data were collected with multistage stratified sampling method. Top-down approach was performed to decompose the Current Curative Expenditure (CCE) of neoplasm and the formula to calculate CCE was introduced step-by-step. Multiple linear regression and random forest regression model were used to analyze influencing factors of hospitalization expenditure with neoplasms.Results: The CCE of neoplasm amounted to 4442.269 million CNY in Gansu 2017, more than half of which were directly burdened by patients and their family, and patients were over-reliant on inpatient service provided by hospitals while the service delivered by primary health care establishment were not fully utilized. The middle-aged adults (40-64) and elderly (over 65 years old) consumed majority of expenditure, and the age of female (45-54 years old) with peak of CCE of neoplasm was almost 15 years younger than male (60-64 years old). Additionally, for male, the top three neoplasms in terms of consuming CCE of neoplasm were stomach cancer, colorectal cancer and lung cancer, whereas that were uterine leiomyoma, breast cancer as well as cervix uteri cancer for female. The random forest regression model showed that the top 5 influential factors in the ranking of importance were the length of stay, operation, proportion of drug, neoplasms and institution level.Conclusions: SHA 2011 is a powerful tool to describe financial flow and to link expenditure to treatment of disease, and in aspect of treatment of neoplasm in underdeveloped region of China, the unreasonable financing scheme, unbalanced allocation of medical resource to heath provider, as well as the distinct distribution characteristics of beneficiaries’ expenditure should attract wide attention to policy maker.