Inject randomness into ventilator waveform settings to augment alveolar function while decreasing bacterial colonization of ventilator-dependent patients. (Inspired by ideas from Antifragile by N. Taleb).
Onion-like sheaths within inner-liner of endotracheal tube that can be peeled off as microorganism biofilms form, decreasing the need for potentially risky tube-changes and decreasing bacterial load in the airway. Combine with intermittently purging endotracheal tube with CO2.
Calculate the number of minutes (theoretically) saved by ambulance use of lights and sirens, then use epidemiological data of disease prevalence in the prehospital setting to ascertain if the number of outcomes changed for the better (ie what disease processes truly benefit from arriving in 10 versus 20 minutes) offsets the increased risk of crashes that injure the patient, EMS providers and the public. Adjust for level of provider training, time of day, weather, traffic conditions, population density. Primum non nocere would dictate that the (likely) increased risk to the general public alone cannot be justified.
Outfit a Roomba with UV lamps to rove hospital hallways and reduce overall bacterial load. Place UV lamps along under-cabinet lighting and use a infrared body heat detector to ascertain when no humans are present.