Evaluating Reporting Bias in Acute Ischemic Stroke Thrombectomy Trials

Introduction/Purpose Acute ischemic stroke thrombectomy trials provide the backbone of evidence‐based practice, yet their design and reporting remain heterogeneous. Selective reporting of efficacy and safety endpoints, inconsistent definitions, and variable use of analysis populations risk introducing bias and complicating comparisons of devices, techniques, and outcomes across studies. We aimed to systematically evaluate reporting practices in recent thrombectomy trials to identify patterns of design‐ and reporting‐related bias. Materials/Methods We performed a systematic review of thrombectomy trials published between 2014‐2024. Recorded variables related to study design: randomization, screening logs, and analyzed populations (intention to treat (ITT), modified intention to treat (mITT), per protocol (PP)). Outcomes included reported first pass and final recanalization (TICI 2b‐3), complications, early neurological status (NIHSS at 24h and at 5 days or discharge), functional status (mRS at 90 days), and mortality. Results 46 trials were reviewed. Median patient age was 69 years (IQR 65‐72), 52% were male. Median recruitment duration was 30 months (IQR 24‐38), with a median of 20 participating sites per study (IQR 12‐45) and a mean recruitment rate of 0.6 patients/month/site. Design features included randomization in 83.9% of studies and reported screening logs in 53.2%. The primary outcome was reported for the ITT in 77.4% of trials, for the mITT in 21.0%, and for the PP in 1.6%. On average, 6.7% of patients were excluded from ITT to conform mITT, and 12.3% from ITT to conform PP. While 97% trials reported 90‐day mRS and mortality, only 26% reported NIHSS at 24h and 17% at 5 days or discharge. Complications reporting was inconsistent: symptomatic ICH was universally reported, but other adverse events (e.g., embolization to new territory, vessel perforation, vasospasm) were heterogenously documented. Reported recanalization outcomes were heterogeneous: 72% reported final angiographic results, but only 38% reported first‐pass recanalization outcomes, and definitions varied across studies (TICI 2b‐3 vs. TICI 2c‐3). Weighted mean first‐pass recanalization (TICI 2b‐3) was lower in ITT population (32%), compared to mITT (39%) and PP (42%). Rates of first pass recanalization were higher in non‐randomized studies (63.1%) than in randomized studies (48.6%). Median % of mRS 0‐2 at 90 days was lower in mITT populations (52.4%) compared to ITT (45.6%). There were no differences in the reported mortality rates between mITT (16.8%) and ITT populations (17.6%). Conclusion Recent thrombectomy trials demonstrate heterogeneity and selective reporting, limiting comparability between new devices and techniques. Outcomes appear systematically more favorable when reported in mITT or PP populations, when screening logs are absent, or when studies are non‐randomized. Complications are incompletely captured, limiting transparency. Standardization of outcome definitions, consistent ITT reporting, and comprehensive complication capture are urgently needed to strengthen the reliability and comparability of future trials.