You have enough data to make a complete cross-tabulation:
$$\begin{array}{r|c|c|c} &\text{pos.}&\text{neg.}&\text{total}\\\ \hline \text{pregnant}&35&9&44\\\ \hline \text{not pregnant}&8&48&56\\\ \hline \text{total}&43&57&100 \end{array}$$
There are $56$ women who are not pregnant, and $48$ of them test negative, so the probability that a randomly chosen non-pregnant woman tests negative is
$$\frac{48}{56}=\frac67=0.\overline{857142}\;.$$
This is slightly larger than your $\frac{56}{66}=\frac{28}{33}=0.\overline{84}$.