Multi-site Retrospective Cohort Study of Central Venous Access Device (CVAD) Removal in Patients with Haematology Malignancies at four Victorian Comprehensive Cancer Centre Clinical Partner Organisations.

Introduction Central Venous Access Devices (CVAD) are critical for the administration of systemic anticancer therapy for patients with haematology cancers. CVADs in this patient cohort have some of the highest CVAD related complications and premature rates in the literature. However, in Victoria these rates and their impact on patient outcomes is unknown. Objectives/Aims This study primarily aimed to describe all reasons for CVAD removal, complications and expenditure. The secondary aim was to identify opportunities to mitigate preventable, premature CVAD removal, to be tested in future healthcare service initiatives. Methods This multi-site, retrospective cohort PhD study collected quantitative data from hospital administrative and health records of adult haematological cancer patients who had a CVAD inserted between 1 September 2020 to 31 August 2021. Data was obtained from numerous sources to maximise data completeness and quality. Analysis included descriptive statistics and regression modelling to identify modifiable predictors for premature removal. Results/Outcomes Of the 1078 CVADs inserted in 673 patients, 503 CVADs (57%) were removed due to expected reasons for removal, 366 (42%) were removed prematurely and 12 (1%) were removed due to unknown reasons. Most CVADs experienced complications (n=919, 85%). This results in increased system and hospital level expenditure. Variances in multidisciplinary management to current evidence were identified. Conclusion This study highlights unacceptably high rates of complications and premature removal of CVADs in patients with haematology cancers. This negatively impacts the patient, hospital and system levels. Opportunities for multidisciplinary initiatives to align clinical practice with current evidence are critical to improve CVAD outcomes in this high risk patient cohort. [ABSTRACT FROM AUTHOR]