To sleep, perchance to leak.

Intermittent positive pressure mechanical ventilation (IPPV) can be applied invasively (through a tracheotomy or an endotracheal tube) or noninvasively, using nasal, oral or facial interfaces. Three different types of physician have been involved in treating patients with this kind of therapy: intensive care specialists, pneumologists (also called pulmonologists or respiratory physicians) and sleep physicians. Each group has its own background, points of view and prejudices, which, at least in part, determine the way in which noninvasive mechanical ventilation is implemented. A stereotypical (though not far from real) intensive care unit patients is a naked human being, lying flat on a relatively sophisticated bed, unaware of the environment. Tubes are inserted into most natural orifices, sometimes more than one tube into a single orifice. Tubes are also inserted through the skin. Some of these tubes are connected to machines and some to bottles, plastic bags or a series of syringes and pump injectors. Sensors are also taped over the chest, or one finger, and around one arm. The bed itself is generally surrounded by monitoring screens displaying numbers and physiological curves, as well as at least one breathing machine and sometimes also an artificial kidney, or a second respiratory machine pumping blood from, and back into, the body of the patient. All this equipment is provided with alarms and beeps; thus, the room is far from silent. To keep patients alive during the organ(s) failure episode that brought them to intensive care in the first place, all of this paraphernalia is required for hours, days or weeks. To make life bearable, patients are heavily sedated and, sometimes also paralysed. Hours go by, the sun sets, and then rises again as electric lighting makes night the same as day. For the attending intensive care specialist who controls a patient's breathing by means of an external mechanical repirator and endotracheal tube, tidal volume is tidal volume, irrespective of the hour. Outside intensive care units, human beings may sometimes be ill enough to justify the replacement of a failing organ with an external mechanical device. Examples of such devices are artificial kidneys, or a mechanical heart. Dialysis allows for extended life expectancy, yet it is carried out periodically; the patient is connected to the artificial kidney machine approximately three times per week for several hours. The rest of the time, the patients can live a quite normal life on their own. To increase free time, some patients are dialysed while they sleep, in order to maximize the time available for the pursuit of personal goals while awake. Respiratory failure is often treated in the following way: patients are connected to an external mechanical device during sleep, thus leaving them free of respiratory assistance during wakefulness. However, contrary to dialysis, free time while awake, is not the only reason for using this adapted method. It is justified by the specific effects of sleep on respiration. Physicians experts in this type of knowledge are generally far removed from intensive care medicine, and pertain, for the most part, to the domain of respiratory and/or sleep medicine. For this type of specialists, tidal volume at 0:00 may be quite different from tidal volume at 12:00 h. A curious (but trivial) observation is that experts in different aspects of medicine rarely speak to each other. More troublesome is the fact that, when they speak to one another, they rarely understand what the other person means as the discourse goes through the various filters imbedded in the brain. Thus, it comes as no surprise that intensive care and sleep specialists do not communicate. This lack of dialogue is facilitated by the exotic nature of each other's knowledge. For the intensive care specialist, sleep lies outside his field of personal or professional experience. Patients sleep very badly in the intensive care unit, with its noisy environment and permanent artificial lighting, in which nurses and physical therapists frequently perform various controls, irrespective of the time of day or night [1, 2]. For the traditional sleep specialist, organ failure and support, invasive monitoring, sedation and paralysis, and life and death are remote notions, rarely encountered in their everyday professional experience. In the treatment of patients with respiratory failure, this absence of dialogue may have deleterious health consequences. Until very recently, the essential difference between mechanical ventilation applied inside or outside the intensive care unit was that patients in intensive care units were intubated and tracheotomised, whereas patients outside intensive care were connected to ventilators through nasal or facial masks (noninvasive (nasal) IPPV (nIPPV)). The latter has now, definitely, found its way into intensive care [3]. The problem is that when patients sleep during nIPPV (whether inside or outside intensive care), they behave in strange ways, that may surprise intensive care (but not sleep) specialists. For instance, they may show large amounts of leaks, whereas no leak is evident when the same patient is ventilated during wakefulness with the same ventilator, the same mask and the same settings. This has serious ventilatory consequences: when leaks are large, *Cliniques Universitaires Saint-Luc, Universite catholique de Louvain, Brussels, Belgium. **Dept of Respiratory Medicine and Sleep Laboratory, PRETA Laboratory TIMC UMR CNRS 5525, University Hospital, Grenoble, France.