Identifying care fragmentation in the management of non-metastatic muscle-invasive bladder cancer.

Abstract

e13527 Background: Measuring care fragmentation facilitates exploration of breakdowns in cancer care delivery that may impact quality of care and patient outcomes. Bladder cancer often requires a multimodal approach of surgery, chemotherapy and radiation, increasing the risk of fragmented care. This study assesses the extent of care fragmentation across different inpatient facilities and examines patient characteristics associated with inpatient care fragmentation. Methods: We adapt the modified Herfindahl-Hirschman market concentration index to construct a Fragmentation Index (FI) capturing the distribution of a patient’s care across inpatient facilities. We calculate FI for the 15 months following diagnosis for non-metastatic muscle-invasive bladder cancer (MIBC) patients who underwent radical cystectomy from 2010-2016 and had at least 2 admissions during the study period. We use Pennsylvania Cancer Registry data linked to statewide facility discharge records to estimate patient-level multivariable regression models of the relationship between FI and patient demographics, disease features, rurality, area deprivation index (ADI), payer, and Commission on Cancer hospital (CoC-H) accreditation status. Results: Of 965 patients, 87.3% (n = 842) lived in urban areas, 8.9% in large towns and 3.8% rural. Approximately 8% had Medicaid and 5.7% had uninsured or unknown payer status. A majority (64%; n = 614) were age 60-79; 76% were men and 92% non-Hispanic White; 19.5% (n = 188) had T4 disease and 25% (n = 239) received neoadjuvant chemotherapy. Almost half of patients (45.3%; n = 437) had an FI of 0, indicating all admissions were at the same hospital. The mean FI was 0.254 for the total cohort and 0.464 for patients with FI > 0. Residency in a large town or rural area was associated with 0.59 (CI 0.003—0.114) and 0.14 (CI 0.060—0.220) higher FI relative to urban areas. T4 disease was associated with 0.063 higher FI (CI 0.019—0.107) relative to <T2 disease. Factors associated with lower FI included Non-White race (-0.092, CI -0.151 – -0.033) and no admissions to CoC-H (-0.138, CI -0.182 – -0.095) relative to patients admitted to at least 1 CoC-H. Patients never admitted to a CoC-H (n = 134) mostly lived in urban areas (88%) and 49% had Medicare insurance: they were disproportionately from the most disadvantaged ADI quartile in the cohort (37.3%). Conclusions: Our findings highlight that many MIBC patients receive inpatient care across at least two different hospitals. Advanced disease and patient residence in non-urban areas are associated with greater fragmentation of inpatient care. Patients who never received care at a CoC hospital experienced less fragmentation but were more likely to be relatively economically disadvantaged. Future research will investigate relationships between fragmentation and clinical outcomes to inform policy and clinical strategies aimed at optimizing patterns of care.