AMS AHMED1, PD LAWTON1, G GORHAM1, M CHATFIELD1, A CASS1
1Renal Unit, Wellbeing & Preventable Chronic Diseases Division, Menzies School of Health Research, Darwin, Northern Territory, Australia
Aim: To compare patient numbers, key date, modality and comorbidity recording between the ANZDATA Registry and NT public hospitals Admitted Patient Care (APC) datasets for NT RRT patients.
Background: With calls from patient and advocacy groups in the NT for treatment closer to home, there is a need to explore various treatment models’ effectiveness and sustainability in remote Australia. The Dialysis Models of Care (DxMOC) Partnership is a cross-organisational, mixed methods study, examining the patient experiences, outcomes and whole-of-government cost of different modalities and models of care for patients receiving RRT in the NT.
Methods: Previously validated ICD-9 and ICD-10AM codes (from 1 July 1998) were used to identify individuals receiving RRT from NT APC data, and linked to ANZDATA Registry data for NT patients to create a cohort of all prevalent patients who received maintenance RRT in the NT between 01/01/2000 to 31/12/2014 inclusive.
Results: 2,375 patients were identified in either dataset having received RRT in the NT. Of 993 patients in APC data but not in ANZDATA, 133 had no identifiable reason for not appearing in ANZDATA. 22 were in ANZDATA but not in APC data. 1,360 patients were matched in both datasets; To examine variation between datasets of dates and diagnosis codes, 1,142 matched participants (starting RRT from 1 July 1998) were selected. There was RRT start date variation (>±90 days) in 11.3% (129/1142) participants. Diabetes had substantial agreement (84.9%, kappa=0.65), coronary artery disease fair agreement (72.6%, kappa=0.29) and other comorbidities slight agreements.
Conclusions: Although variation exists between NT APC and ANZDATA in patient numbers, RRT start date and diagnosis codes, the linked datasets allow more detailed analysis than either alone.