INTERHOSPITAL TRANSFER OR DIRECT TRANSPORT TO SPECIALIZED CENTRES FOR THROMBECTOMY: INTERPRETATION OF QUÉBEC RESULTS IN LIGHT OF LATE BREAKING REAL-WORLD EVIDENCE 2019 CADTH Symposium Michèle de Guise, MD, FRCPc Director of health technology assessment, INESSS on behalf of the Cardio-neurovascular Evaluation Unit
DISCLOSURE INESSS is publicly funded. I have no actual or potential conflict of interest in relation to this topic or presentation. 2
PUR URPO POSE SE OF TH THIS S PRESE SENT NTATION TION To reflect on how RWData and RWE are complementary to RCTs in order to provide more adapted and contextualized recommendations. Using as case study the mandate we received from the Ministry of health to evaluate how to optimize access to endovascular treatment (EVT, i.e. thrombectomy) in Québec, focusing on the optimal pathway. 3
STRUCTURE OF THE SYSTEM OF STROKE CARE IN QUÉBEC (2017-18) 80 hospitals providing different levels of care, 4 thrombectomy programs concentrated in urban areas 4
PATIENT TRAJECTORY FOR EVT TREATMENT Triage NT Non tertiary Needle Scan (Door) (NT) centre t-PA Departure from NT centre First medical Symptoms contact First Triage EVT Needle arterial EVT centre Scan (Door) t-PA puncture PATIENT PREHOSPITAL REFERRAL INTERHOSPITAL EVT FAMILY (EMS) (NT) CENTRES (EMS) CENTRES POST-PROCEDURE CARE AND REHABILITATION 6
SYNT NTHE HESIS SIS OF ALL LL T TYPES S OF EVIDE DENCE NCE TO DEVELOP OP PROPOS OPOSED ED RECOM OMMEND ENDATI TION ONS 7
SC SCIEN ENTIFIC TIFIC EVIDE DENCE NCE • Meta-analyses of five randomized trials: • In comparison to thrombolysis, thrombectomy is associated with improved patient outcomes if treatment is received within 7 hours of the start of symptoms • The shorter the treatment delays, the better the outcomes. • Among 1000 patients achieving endovascular reperfusion; • For every 15 minutes faster emergency department door-to- reperfusion time • 39 patients would have less-disabled outcome at 3 months • 25 among those would achieve functionnal independence • Assuming we can replicate the condition of the RCTs HERMES collaboration: Saver et al. and Goyal et al. (2016) 8
RCT T SE SELE LECTI TION ON CRITE TERIA RIA • Selection criteria for patients: the vast majority were functionally independent pre-stroke mRankin score ≤2 2/5 trials had age restrictions (e.g., ≤80 years old) • Selection criteria for participating centres: high volume e.g. ≥40 -60 EVT/year; >500 stroke patients/year high expertise • Study protocols emphasized fast treatment Thus, patients and hospitals not necessarily representative of those in the real-world context of stroke care 9
FIELD EVALUATION OF EVT BY INESSS April 1, 2017 March 31, 2018 Observation period Hospitals All hospitals with an EVT program (n=4) All patients who presented to an emergency room Patients (by ambulance or other means) and were treated with t-PA or EVT in an EVT centre Documentation by EVT clinical teams Data collection by INESSS Data collection Centralized secure web site (REDCap) Read only access to REDCap Data validation Validation of preliminary results by EVT teams Analysis by INESSS in collaboration with clinical Data analysis expert committee (reps from each EVT program) 10
VOLUME EVT ± t-PA (2017-18) 200 183 180 Total volume EVT ± t-PA in Québec 160 = 375 140 120 100 80 77 80 60 35 40 20 0 A B C D The four thrombectomy centers have a wide range in volume of cases 11
PATIENT CHARACTERISTICS: RCT VS QUÉBEC (1) RCT Québec meta-analysis 2017-18 N=634 N=375 Median age in years 68 71 (25 th -75 th percentile) (57-77) (60-80) Age ≥ 80 years 17% 29% Women 48% 51% 12
PATIENT CHARACTERISTICS: RCT VS QUÉBEC (2) RCT Québec meta-analysis 2017-18 N=634 N=375 Pre-stroke mRankin ≤ 2 98% 94% Median (25 th -75 th percentile) 9 9 Initial ASPECTS score (7-10) (7-10) 17 16 Initial NIHSS score (14-20) (11-20) Start of symptoms to 99 min 151 min triage at EVT centre (door) (52-191) (60-222) 13
LATE-BREAKING EVIDENCE • American registry publication (online Jan 31, 2019): 37,260 EVT patients treated at 639 centres, 2012-17 14
PROCESSES: RCT VS QUÉBEC VS USA REGISTRY RCTs Québec USA N=634 N=375 N=37,260 Mode of arrival: Direct admission 70% 41% 57% Interhospital transfer 30% 59% 42% Use of t-PA 83% 61% 58% The proportion of transferred patients in Québec were much more similar to the US registry than to the RCTs, but still higher. The proportion of use of t-PA was much more similar to the US registry 15
RCT VS QUÉBEC VS USA REGISTRY TREATMENT DELAYS FOR TRANSFERRED PATIENTS RCT Québec USA N=184 N=209 N=15,975 Median delay: Start of 295 247 289 symptoms to EVT (255-342) (202-310) (NR) (25 th -75 th percentile) Median delay: Door of 81 25 68 EVT centre to EVT (58-105) (14-43) (NR) (25 th -75 th percentile) For transferred patients, delays from symptoms to hospital arrival and from triage to thrombectomy results were favourable in comparison to both the trials and the American registry 16
RCT VS QUÉBEC VS USA REGISTRY TREATMENT DELAYS FOR DIRECTLY ADMITTED PATIENTS RCT Québec USA N=421 N=166 N=21,285 Median delay: Start of 210 135 213 symptoms to EVT (158-270) (107-185) (NR) (25 th -75 th percentile) Median delay: Door of 116 70 128 EVT centre to EVT (82-160 (50-100) (NR) (25 th -75 th percentile) Shorter delays from triage to thrombectomy showed the good performance of our current thrombectomy programs 17
USA REGISTRY EVIDENCE ON OUTCOMES Directly Adjusted OR Transferred admitted OR for treatment Outcome patients patients (95% CI) delay N=15,975 N=21,285 (95% CI) 1.01 In-hospital 1.17 (0.92, 1.11) 14.7% 13.4% mortality (1.10, 1.24) P=ns Independent 0.87 0.80 ambulation at 33.1% 37.2% (0.80, 0.95) (0.76, 0.85) discharge P=0.002 Shah et al., 2019 18
USA REGISTRY EVIDENCE ON OUTCOMES In comparison to patients treated after inter-hospital transfer, Direct admission EVT was associated with : • Significant decrease in in-hospital mortality • Significant increase in independent ambulation at hospital discharge • The decrease in mortality was no longer significant after adjustment for differences in treatment delays • Time is brain ! Shah et al., 2019 19
CONCLUSION New evidence from our own context in Québec and from a large registry : • Was comforting in terms of implementation of the EVT programs in Québec; • Added weight to the conclusions of the RCT meta- analyses that “ Time is brain ”; • Increased the strength of our recommendations concerning prioritization of direct transport for EVT; • Showed the importance of documentation of real- world outcomes to aid decision-making. 20
Québec Montréal 2535, boulevard Laurier, 5e étage 2021, avenue Union, bureau 10.083 Québec (Québec) G1V 4M3 Montréal (Québec) H3A 2S9 Téléphone : 418 643-1339 Téléphone : 514 873-2563 Télécopieur : 418 646-8349 Télécopieur : 514 873-1369 inesss.qc.ca inesss@inesss.qc.ca 21
Recommend
More recommend
