Abstract TP234: "Last Known Normal" Accuracy Among Initial Responders: An Update

Abstract

Background: Because acute treatment in stroke is time-based for inclusion, efficacy and safety, obtaining an accurate Last Known Normal (LKN) is of critical importance in stroke codes. We sought to assess with a larger sample if the assessment of 1st documented LKN times has improved since our prior 2013 data. Methods: Data was obtained from an IRB approved stroke registry in a single center from July 2013 to December 2018, for LKN time documented by a neurologist (“LKN2”). Chart review was done to document 1st reported LKN time as documented by EMS (or ED if no runsheets available) (“LKN1”). Inpatient stroke codes and hospital transfers were excluded. Differences in LKN1 and LKN2 were computed and stratified into Groups A (LKN1 is earlier in time than LKN2), B (LKN1 is the same as LKN2), and C (LKN1 is later in time than LKN2). Baseline characteristics, thrombolysis rates, stroke code time interval metrics, 90-day disability and death, discharge disposition, and symptomatic ICH rates, were compared between groups. Results: Of 990 stroke codes, 397 or 40.1% had agreeable LKN1 and LKN2 times (Group B) (increased from a historic 26.4%;p=<.001), while 593 or 59.9% had a discrepancy in LKN1 and LKN2 times. Of 593, 177 (29.8%) had an LKN1 earlier than LKN2 (Group A), 416 (70.2%) had LKN1 later than LKN2 (Group C). The mean age in Groups A, B, and C were 63.5, 63.4, and 66.1, respectively (p=0.04). Discharge disposition to home/self-care was seen more in Group C (n=284, 69.4%;A n=117, 63.2%;B n=255, 66.2%;p=0.03). There were no other differences in baseline characteristics, r-tPA rates, 90-day disability and death, or sICH rates. Among Group C patients who were excluded from IV-tPA based on time, 55.6% would have been treated outside of stroke guidelines had LKN1 been used (55.6% vs. prior report of 69.7%;p=0.2). Conclusion: Though initial LKN times obtained by EMS and ED responders have improved over time, there remains a significant discrepancy with 60% incorrect initial reports. Caution should be used when considering rt-PA treatments based on these LKN1 reports as 56% of cases could have been treated outside of current guidelines and evidence. This study highlights the need for continuous training in obtaining accurate LKN times and caution about using initial estimates of time.