Effect of Fragmentation of Cancer Care on Treatment Use and Survival in Hepatocellular Carcinoma
BACKGROUND: Fragmentation of care (FC) refers to treatment received at multiple facilities. Implications of FC include increased health care costs and amplification of existing healthcare disparities. This study aimed to identify patient and hospital-level factors associated with FC and analyze the effect of FC on patient outcome measures (overall survival and time to treatment). OBJECTIVE: Fragmented care in HCC patients is associated with worse overall survival and increased time to treatment compared to patients receiving non-fragmented care. METHODS: The Texas Cancer Registry (TCR) was queried from 2004-2015 for a 12-year study span. Patient- and hospital-level factors were characterized within 2 groups: patients receiving fragmented care (FC) and those receiving non-fragmented care (NFC). Cox proportional hazards regression models were used to identify those factors that were independently and significantly associated with overall survival and time to treatment. Kaplan-Meier curves were generated to evaluate differences in overall survival between the FC and NFC groups, as well as between every type of transition within the FC group (stratified on the basis of safety net hospital status or volume status). All statistical analyses were performed with SPSS. RESULTS: Of our cohort (n = 4329), 72.6% received NFC, and 27.4% received FC. In comparison to patients receiving NFC, patients receiving FC had larger median tumor sizes at diagnosis (≥4 cm, 52.6% vs 35.2%; p < .001). NFC patients also tended to present with regional or metastatic disease (35.9% vs 26.7%; P < .001). A subset analysis of patients with localized stage HCC who received curative therapy showed that FC was associated with decreased odds of curative therapy (odds ratio, 0.83; 95% confidence interval [CI], 0.7-0.9). In this subgroup analysis, FC was associated with worse OS (median survival, 67 vs 43 months; HR, 1.2; 95% CI, 1.0-1.4) and increased TTT (HR, 0.74; 95% CI, 0.7-0.8). Ultimately, in our global cohort, FC was associated with worse OS (hazard ratio [HR], 1.14; 95% CI, 1.05-1.24) and increased TTT (HR, 0.76; 95% CI, 0.7-0.8). CONCLUSION: Patients receiving FC had worse OS and increased TTT compared to patients receiving NFC. Several patient and hospital-level factors were found to be associated with FC, including age, insurance, non-safety net hospital status, accreditation, and disease stage. This work has implications for encouraging initiatives geared toward increasing care coordination, especially when managing cancer. Future work may aim to elucidate the reasons for the associations described and delineate steps by which to mitigate FC in the context of these factors.