Introduction:
Symptomatic intracerebral hemorrhage (ICH) is a feared complication of reperfusion therapy in acute ischemic stroke. The Joint Commission (JC) national quality measure Comprehensive Stroke (CSTK-5b) reflects the rate of symptomatic ICH after mechanical thrombectomy (MT) and is used as a marker of comprehensive stroke center (CSC) performance. We sought to determine the accuracy of the CSKT-5b as determined by vascular neurologists.
Methods:
We reviewed the CSTK-5b failures in our JC certified CSC program for the years 2018 and 2019. Our CSC abstractors follow the instructions from the Joint Commission National Quality Measures manual without deviation and do not allow for clinician over-ride or adjudication (consistent with manual instructions). All CSKT-5b failures were reviewed by study neurologists to determine the true CSTK-5b rate. Data was collected on National Institute of Health Stroke Scale (NIHSS), symptomatic ICH, and neuroimaging results. European Cooperative Acute Stroke Study (ECASS) definitions were used and included hemorrhagic infarction (HI) (grade 1 or 2) and parenchymal hematoma (PH) (grade 1 or 2).
Results:
Among 361 MT patients for the 2 year period, the CSTK-5b failure rate reported to JC was 34/250 (13.6%), whereas the true rate was 21/250 (8.4%). Among the 13 cases that were miscoded, the ECASS grading included 1 HI-1, 7 HI-2, 1 PH-1, and 4 subarachnoid hemorrhage, and all had a 4 point change in NIHSS that was determined to be clinically unrelated to ICH. Comparing annual data, the 2018 CSTK-5b failure rate was 14/111 (12.6%) whereas the true CSTK-5b failure rate was 7/111 (6.3%). In 2019, the CSTK 5b failure rate was 20/139 (14.4%), whereas the true CSTK-5b failure rate was 14/139 (10.1%).
Conclusions:
Non-physician abstractors relying on instructions from the JC manual over-estimate the rate of CSKT-5b measure failure. The determination of symptomatic ICH after MT is complex and requires clinical knowledge for accuracy.