Patients in the intensive care unit (ICU) often suffer from delirium, a mental condition that involves disorganized thinking, general confusion, and sometimes, hallucinations. Physicians screens for delirium regularly since it affects 2 out of 3 hospitalized patients and is correlated to morbidity in the ICU. The screening process requires verbal or physical methods of communication (e.g. eye blinks or hand squeezing); however, endotracheal tube and mechanical ventilators as well as traumatic brain injuries may prevent patients from communicating effectively. Brain computer interfaces (BCI) could potentially help those patients who lack the necessary motor control. In this work, we have developed a multi-sensory (tactile, visual, and auditory) binary communication BCI using electroencephalography (EEG) signals. The system consists of the necessary hardware for brain stimulation, as well as the software needed to perform intent detection. Communication between modules was implemented using OpenDDS, a real-time publish-subscribe communication module in order to facilitate integration with other BCI applications. In addition, a graphical user interface was built to allow caregivers to perform standard delirium assessments with this tool. We successfully demonstrated the capabilities of our prototype in both visual and tactile BCI paradigms. The system can potentially be used with other vibro-tactile paradigms.