Data Mining Procedural Reports of Percutaneous Interventions in Dialysis Access for Quality Assurance Programs
Keywords:diagnostic imaging, patient safety, health information technology, quality improvement, hemodialysis
Purpose: Ongoing quality assessment of dialysis access interventions is critical in the care of hemodialysis-dependent patients. The objective of this study was to mine data from interventional radiology (IR) reports of dialysis access interventions to determine if these reports contain the data necessary to retrospectively calculate quality outcome metrics required to support quality assurance (QA) programs.
Methods: A retrospective analysis was conducted of IR reports created at the institution between May 28, 2019 and October 16, 2020. Radiology reports pertaining to percutaneous image-guided management of thrombosed or dysfunctional hemodialysis circuits were included. Reports that only described diagnostic intervention(s) were excluded. Relevant reports were manually annotated according to a checklist of items derived from published reporting standards to determine whether reports contained the data necessary to retrospectively calculate quality outcome metrics, such as postintervention primary patency (PiPP) and postintervention lesion patency (PiLP).
Results: A total of 130 reports describing 78 patients were included in the analysis. Documentation of items derived from published reporting standards for dialysis fistulogram and interventions ranged from 28.5% to 100%. Only 18.5% and 15.4% of radiology reports were independently sufficient to calculate PiPP and PiLP, respectively.
Conclusion: Few reports were independently sufficient to retrospectively calculate quality outcome metrics, PiPP and PiLP. Results of this study suggest the need for greater standardization in reporting practices. Standardized reporting has the potential to improve communication, promote adherence to guidelines, and provide data for quality improvement projects to optimize patient care.