Di Shen, Yuwei Wang, Xin Li, Xiduo Yang, Mengfan Yu
In this single-center study conducted at a tertiary infectious disease hospital in China, HIV/AIDS inpatient costs were concentrated in a small proportion of high-cost hospitalizations. High resource-use intensity, prolonged stay, and greater coded disease burden characterized high-cost cases, especially in the upper tail. These findings highlight the value of distributional cost analysis for identifying high-resource-use hospitalizations and informing complex case management and hospital resource planning in HIV/AIDS care.
BACKGROUND: With the expansion of antiretroviral therapy, human immunodeficiency virus/acquired immunodeficiency syndrome (HIV/AIDS) has increasingly become a manageable chronic condition. However, late presentation, opportunistic infections, and critical care can still generate substantial inpatient costs. Mean-based analyses may obscure upper-tail cost concentration. This study quantified HIV/AIDS inpatient cost concentration and resource-use profiles of high-cost hospitalizations.
METHODS: We analyzed HIV/AIDS hospitalizations at a tertiary infectious disease hospital between 1 January 2019 and 31 December 2024. High-cost hospitalizations were defined as total inpatient costs at or above the 90th percentile. Cost concentration was assessed using the Lorenz curve, Gini coefficient, and expenditure shares of the top 5 and 10%. Multivariable logistic regression identified factors associated with high-cost hospitalization. Conditional quantile regression examined associations between medical resource-use intensity and costs across quantiles; recentered influence function unconditional 90th-percentile regression was used as a supplementary analysis.
RESULTS: Among 3,240 hospitalizations, the 90th-percentile threshold was RMB 62,683.41. The top 5 and 10% accounted for 23.49 and 35.97% of total inpatient costs; the Gini coefficient was 0.481. Compared with low resource-use intensity, moderate and high resource-use intensity were associated with high-cost hospitalization, with adjusted odds ratios of 18.27 (95% CI, 4.41-75.70) and 90.62 (95% CI, 21.94-374.34), respectively. Prolonged length of stay, a greater number of coded opportunistic or AIDS-defining conditions, and non-local care seeking were also associated with high-cost hospitalization. The estimated cost difference associated with high resource-use intensity was numerically larger at higher cost quantiles, ranging from RMB 18,261.89 at the 25th percentile to RMB 60,178.15 at the 90th percentile.
CONCLUSION: In this single-center study conducted at a tertiary infectious disease hospital in China, HIV/AIDS inpatient costs were concentrated in a small proportion of high-cost hospitalizations. High resource-use intensity, prolonged stay, and greater coded disease burden characterized high-cost cases, especially in the upper tail. These findings highlight the value of distributional cost analysis for identifying high-resource-use hospitalizations and informing complex case management and hospital resource planning in HIV/AIDS care.