Factors associated with unplanned hospital readmissions and emergency presentation following percutaneous coronary intervention: insights from the Victorian Cardiac Outcomes Registry.

Abstract

Background: Resource constraints faced by clinical quality registries when performing longer-term follow-up can potentially be overcome by data linkage to administrative datasets. This study aimed to determine whether linkage of the Victorian Cardiac Outcomes Registry (VCOR) to a comprehensive administrative dataset available from the Department of Health, Victoria, could assist in the follow-up process of percutaneous coronary intervention (PCI) patients, and identify factors associated with postPCI unplanned emergency department (ED) presentations and hospital readmissions (UHR).

Methods: The study prospectively collected data on 8,435 VCOR public hospital index PCI episodes between January 2013 and May 2015, which were then linked to data from the Victorian Admitted Episodes Dataset (VAED) and the Victorian Emergency Minimum Dataset (VEMD). International classification of disease (ICD-10) diagnosis codes in VAED were used to identify unplanned cardiac readmissions (UCR).

Results: A total of 1,379 patients (16.3%) re-presented to the ED in the first 30 days after PCI. Within 30 days of PCI, 996 patients (11.8%) had a UHR, of which 363 (4.3%) had a UCR. A total of 4,182 patients were able to have linkage performed up to 12 months after PCI. Of these, 1,375 patients (32.9%) had a UHR, with 506 (12.1%) having a UCR. Independent predictors of UHR were aged >80 years, being female, having ≥1 hospital admission in the 12 months before PCI, acute coronary syndrome presentation, and moderate left ventricular dysfunction (p < 0.01).

Conclusion: A significant proportion of patients who underwent PCI had unexpected ED presentations and readmissions to hospital. Targeted strategies for high-risk patients may be useful. Administrative dataset linkage provided an effective method for clinical registries follow-up.