Study on Relationship between Characteristics of Stroke Referral Network and Medical Costs and Length of Stay:an Retrospective Analysis based on Medical Insurance Data in Guangzhou
Objective Explore the relationship between hospital characteristics and medical costs and length of stay,and thus provide policy recommendations for promoting the system of stroke classification and treatment.Methods Social Network Analysis methods were employed to explore the characteristics of medical institutions in stroke referral network,with multiple linear regression analysis used to explore the impact of network structure metrics on patients'length of stay,total hospitalization costs,and out-of-pocket expenses.Results A total of 11 918 stroke inpatients and 158 hospitals were included in the initial network.As the network threshold increases,the average length of stay per admission decreased from 15.76 days to 15.60 days,the average total hospitalization cost per admission decreased from 12 570.57 yuan to 1 1 897.40 yuan;and the average out-of-pocket cost per admission decreased from 1 745.66 yuan to 1 558.91 yuan.In the network with a threshold of 16,patients were referred a total of 89 740 times,with tertiary hospitals having the highest number of referrals(23 485 times for referral out and 21 371 times for referral in).Tertiary hospitals had the highest proportion of hospitals and centrality metrics in the referral network,except for degree centrality in the densest network.Higher centrality was associated with lower total hospitalization costs and out-of-pocket expenses,while higher closeness centrality was associated with shorter length of stay.Conclusion Tertiary hospitals occupied a dominant position in the stroke referral network.Secondary hospitals did not play an intermediary role in the system of stroke classification and treatment.Increasing the closeness of the referral network and forming closer and faster cooperation between medical institutions is beneficial to lower medical costs and length of stay.
strokereferralmedical costslength of staySocial Network Analysis
NETL
NSTL
万方数据
