Retrospective database analysis of 1907 hospitalized patients in Japan, using PSM to compare 74 remdesivir patients and 195 control patients, not showing significant differences in mortality and ventilation/ECMO, and showing signifcantly longer hospital stay with treatment. PSM did not fully match the patients, for example 65% of matched remdesivir patients were over 65, while only 55% of matched control patients were.
Tsuzuki et al., 3/10/2021, retrospective, propensity score matching, Japan, Asia, preprint, 20 authors.
risk of death, 8.8% lower, RR 0.91, p = 1.00, treatment 9 of 74 (12.2%), control 26 of 195 (13.3%), PSM.
risk of mechanical ventilation or ECMO, 17.1% higher, RR 1.17, p = 0.76, treatment 4 of 74 (5.4%), control 9 of 195 (4.6%), PSM.
hospitalization time, 27.3% higher, relative time 1.27, p < 0.001, treatment 74, control 195, PSM.
Effect extraction follows
pre-specified rules
prioritizing more serious outcomes. For an individual study the most serious
outcome may have a smaller number of events and lower statistical signficance,
however this provides the strongest evidence for the most serious outcomes
when combining the results of many trials.