I'm trying to understand the proof of theorem 8.3 in Paulsen's book "Completely bounded maps and operator algebras". Here is (part of) the proof:
Why can we conclude that $$\Phi\begin{pmatrix} p & 0 \\ 0 & 0\end{pmatrix}= \begin{pmatrix} * & 0 \\ 0 & 0\end{pmatrix}?$$
I guess it might be sufficient to show that $$\begin{pmatrix} 0 & 0 \\ 0 & 0\end{pmatrix} \le \begin{pmatrix} a & b \\ b^* & c\end{pmatrix} \le \begin{pmatrix} 1 & 0 \\ 0 & 0\end{pmatrix}\implies b = c = 0$$ where the middle matrix is a positive matrix. However, I can't show this and it should be straightforward.

I'll call the three matrices in question $0$, $A$ and $E_{11}$. For $\xi\in \mathcal H$ the inequality $$ 0\leq \left\langle A\binom{0}{\xi},\binom{0}{\xi}\right\rangle\leq\left\langle E_{11}\binom{0}{\xi},\binom{0}{\xi}\right\rangle $$ implies $0\leq\langle c\xi,\xi\rangle\leq0$, hence $c=0$.
Moreover, $$ 0\leq\left\langle A\binom{\xi}{\eta},\binom{\xi}{\eta}\right\rangle=\langle a\xi,\xi\rangle+2\mathrm{Re}\langle \xi,b\eta\rangle. $$ If $b\neq 0$, then there exist $\xi,\eta\in H$ such that $\langle\xi,b\eta\rangle\neq 0$. Multiplying $\eta$ by a suitable scalar makes this inner product negative and arbitrarily small, while $\langle a\xi,\xi\rangle$ stays unchanged. Thus we can make $\langle a\xi,\xi\rangle+2\mathrm{Re}\langle \xi,b\eta\rangle$ negative, a contradiction.