A critique of the evidence base for non-pharmacological sleep interventions for persons with dementia