Back

Is mRankin scale correlated with mTICI? A systematic review and meta-regression on RCTs and registries.

De Rubeis, G.; Pampana, E.; Prosperini, L.; Fabiano, S.; Bertaccini, L.; Anticoli, S.; Saba, L.; Gasperini, C.; Cotroneo, E.

2021-05-17 radiology and imaging
10.1101/2021.05.12.21257077 medRxiv
Show abstract

Background and PurposemTICI [≥]2b/3 is one of the strongest positive predictors of mRS [≤]2. Quantitative analysis is poorly investigated. Reconcile results from RCT and registries is still a challenge. The purpose was to evaluate the numeric correlation between mTICI[≥]2b/3 and mRS[≤]2 in RCT and registries. MethodsLiterature research was performed on Pubmed for studies in 2015-2020. mTICI, mRS and sample size were recorded. Exclusion criteria were monocentric study, not-human and not-English. Studies quality were assessed with MINORS and RoB2. Meta-logistic and meta-linear regressions were used to correlate mTICI and mRS in both RCTs and registries. Z-test was used for comparing coefficients between RCTs and registries. ResultsTwenty-six studies were evaluated (13 registries; 14 RCTs) for 24423 patients (21914 from registries [average per registry 1685{+/-}1277]; 2509 from RCTs [average per RCT 179{+/-}160]). RCTs involved anterior circulation only, 7/13 (53.8%) registries considered also posterior one. The OR of obtaining a mRS[≤]2 for a singular increased of mTICI [≥]2b rate was 1.65 (CI95% 1.22-2.01) for all studies, 1.65 (CI95% 1.10-2.46) for RCTs and 1.50 (CI95% 1.00-2.23) for registries. mTICI[≥]2b and mRS had a positive correlation with a coefficient of 0.49 (CI95% 0.19-0.80, p=0.001) for all studies, 0.54 (CI95% 0.09-1.00) for RCTs and 0.42 (CI 95% 0.04-0.81) for registries. No differences were found in the coefficients between RCTs and registries (p=0.63; p=0.65; respectively). ConclusionsUnitary increased of mTICI[≥]2b rate correspond to an augment of mRS[≤]2 by 0.50 (CI95% 0.19-0.89) with OR of obtaining mRS[≤]2 of 1.65 (CI95% 1.22-2.01), without significantly differences in coefficients.

Matching journals

The top 6 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.