1: \begin{abstract}

2: We use an effective field theory for short-range forces (SREFT) to analyze systems of three identical bosons interacting via a two-body potential that  generates a scattering length, $a$, which is large compared to the range of the interaction, $\ell$. The amplitude for the scattering of one boson off a bound state of the other two is computed to next-to-next-to-leading order (N$^2$LO) in the $\ell/a$ expansion. At this order, two pieces of three-body data are required as input in order to renormalize the amplitude (for fixed $a$). We apply our results to a model system of three Helium-4 atoms, which are assumed to interact via the TTY potential. We generate N$^2$LO predictions for atom-dimer scattering below the dimer breakup threshold using the bound-state energy of the shallow Helium-4 trimer and the atom-dimer scattering length as our two pieces of three-body input. Based on the convergence pattern of the SREFT expansion, as well as differences in the predictions of two renormalization schemes, we 

3: conclude that our N$^2$LO 

4: phase-

5: shift predictions will receive higher-order corrections of $< 0.2$\%. In contrast, the prediction of SREFT for the binding energy of the ``deep" trimer of Helium-4 atoms displays poor convergence.

6: \end{abstract}

7: