Background: Last known well (LKW) time is critical information when determining eligibility for acute stroke treatments, but is often unknown or not documented. Our aim was to use the Michigan Stroke Coverdell Registry to identify independent predictors of missing LKW and the association between missingness and outcomes.
Methods: Forty hospitals contributed data to the registry between 2008-2017. Patients were excluded from the analysis if they had an in-hospital stroke or arrived via transfer from another healthcare facility. Independent factors associated with missing LKW time documentation were identified using multivariable logistic regression.
Results: Out of total of 51,066, LKW was missing in 24,675 (48.3%) of the cases. Characteristics of cases and their association with LKW missingness are shown in the Table. Missing LKW documentation was higher among non-Caucasians, those with an onset in a health care facility (versus home), those who arrived via private vehicle (vs EMS), those with hemorrhagic strokes and those who presented with aphasia or weakness. Documentation of LKW was lower in those who presented with TIA or had a history of atrial fibrillation or prior TIA. LKW missingness varied widely between hospitals (Range: 10.3 - 72.5%) (Figure). Patients without documentation of LKW had longer lengths of stay, higher in-hospital mortality, and were more likely to have missing NIHSS data (63% versus 37% for those with documented LKW).
Conclusions: Documentation of LKW time is common, varies substantially between hospitals and is associated with poor outcomes. These data provide an opportunity for education and quality improvement efforts to improve documentation.