Patient　home　waiting　during　a　pandemic　may　trigger　patients　to　baulk　or　renege.　Meanwhile,　hospitals　need　to　ascertain　whether　patients　have　contracted　infectious　diseases.　Given　the　patient　baulking　and　reneging,　hospitals　should　devise　optimal　operational　management　plans　supported　by　rational　admission　decisions,　buffer　wards,　and　dynamic　bed　and　nurse　allocation.　Using　a　non-linear　programming　model,　this　study　explores　the　role　of　buffer　wards　to　derive　the　optimal　decision　plan　for　the　dynamic　allocation　of　hospital　beds　and　nurses,　and　the　possible　number　of　patients　to　be　hospitalised　in　the　presence　of　patient　baulking　and　reneging.　The　optimal　plan　for　patient　admission　and　dynamic　bed　and　nurse　allocation　is　developed　based　on　the　Gurobi　solver.　Furthermore,　a　data-driven　decision　approach　is　proposed　to　update　the　model　parameters　over　time　based　on　the　rolling　horizon　procedure.　As　a　result,　the　aforementioned　model　proposes　an　optimal　decision　plan　for　the　rational　operational　management　of　the　hospital　including　the　number　of　reneging　and　baulking　patients,　the　movement　of　beds　between　different　wards,　and　the　number　of　rejected　patients.