In this paper, we present a robust model for the community healthcare service network design problem. The community healthcare service network can determine the location of central hospitals and the allocation of the community medical service stations in order to minimize that cost of the whole network. Beside, we employ the Box+ellipsoidal set to deal with the objective which obtain the uncertain parameter. In addition, we reformulate the proposed model into a computationally tractable robust counterpart form under Box+ellipsoidal set. Finally, we verify the validity of the model by a case study.