Table 3

Spin in abstract conclusions

Type of spinNo. (%) (n=57)
Claim benefit due to a statistically significant primary endpoint and ignored one or more statistically nonsignificant primary endpoints18 (32)
Claim benefit based on a statistically significant secondary endpoint15 (26)
Claim equivalence/non-inferiority versus comparator for a statistically nonsignificant endpoint11 (19)
Focus on another objective (eg, trial is statistically nonsignificant but they say they accomplish some goal that they did not prespecify)7 (12)
Claim benefit based on statistically significant subgroup analysis3 (5)
Claim benefit based on statistically significant modified treatment population (eg, per-protocol instead of intention to treat)1 (2)
Claim benefit based on statistically significant within group analysis1 (2)
Emphasise on the magnitude of difference in the presence of a statistically nonsignificant p value1 (2)