highly significantP < 0.001
After each control session, safety variables were compared with baseline values in the whole sample by paired tests, to detect possible highly significant ( P < 0.001) and clinically relevant differences.
After each control session, safety variables were compared with baseline values in the whole sample by paired tests, to detect possible highly significant ( P < 0.001) and clinically relevant differences.
In the first case (best possible result, Supplementary Table 3 ), borderline significance could be obtained between groups and for interaction term, but without any significant progression of proBNPage over time.