Hi Constance,
Percentage of variance is expressed by R2, which is exactly what it sounds like. R is the Pearson correlation coefficient, R is its square, so percentage of variance is:
0.152 = 0.0225 = 2.25%
Technically, we cannot gauge statistical significance without more information--specifically an ANOVA table-- but an R2 that low is unlikely to be significant. Only 2.25% of variability in anxiety scores is explained by the drug. Cohen classifies this percentage as "weak" or "small" depending on regression type, so unlikely to be significant.
I hope this helps.