In a clinical trial, a researcher compares the mean blood pressure reduction $\mu_T$ of a treatment group and $\mu_C$ of a control group. The estimated difference is $\hat{\Delta} = 12.5$ with a standard error $SE = 3.2$. Under the null hypothesis $H_0: \Delta = 0$, the researcher assumes a normal distribution for the sampling error. What is the approximate lower bound of the 95% confidence interval for the treatment effect?