Introduction: The Risk of Aneurysm Rupture (ROAR) study is a UK multicentre natural history study to determine the risk of rupture from a known unruptured intracranial aneurysm (UIA). It will use national healthcare databases for patient follow up which will need to identify events of UIA treatment for censoring patients. This study was to establish the sensitivity of different codes to identify these treatments.
Method: Patients were identified at a single neurosurgery unit from records dated 2006-2020 and linked to the Hospital Episode Statistics - Admitted Patient Care (HES-APC) database. All cases underwent case note review to identify UIA treatments during that time. All HES episodes containing an OPCS4 code for aneurysm treatment underwent further individual case note review, recalling records from external providers as required.
Results: 318 instances of elective UIA treatment were identified, of which 310 were found in the HES-APC database. The sensitivity of HES-APC for identifying elective UIA treatment is 95.6%, and the estimated sensitivity of combining HES-APC with neurosurgery unit electronic patient records, as will be done in the ROAR Study, is 99.88%. The L33 or O01-4 OPCS4 codes were used for 93.8% of elective aneurysm treatments in HES-APC.
Discussion: The HES-APC database is an effective source for identifying elective UIA treatments, and its high sensitivity makes it ideal for long-term follow-up in the ROAR Study.
Keywords: Hospital Episode Statistics (HES); ICD-10; Unruptured intracranial aneurysms; aneurysm treatment; healthcare data linkage.