Objective: To determine the oxygen outflow delivered by seven different models of manually operated self-inflating resuscitation bags (with and without an oxygen reservoir connected), which were tested using different oxygen supply rates without manipulating the bag, by simulating their use in patients breathing spontaneously. Methods: The oxygen outflow was measured using a wall oxygen flow meter and a flow meter/respirometer attached to the bag, together with another flow meter/respirometer attached to the patient connection port. The resuscitation bags that allow the connection of an oxygen reservoir were tested with and without this device. All resuscitation bags were tested using oxygen supply rates of 1, 5, 10, and 15 L/min. Statistical analyses were performed using analysis of variance and t-tests. Results: The resuscitation bags that allow the connection of an oxygen reservoir presented a greater oxygen outflow when this device was connected. All resuscitation bags delivered a greater oxygen outflow when receiving oxygen at a rate of 15 L/min. However, not all models delivered a sufficient oxygen outflow even when the two previous conditions were satisfied. Conclusions: Of the resuscitation bags studied, those that allow the connection of an oxygen reservoir must have this reservoir connected to the bag when used as a source of oxygen in nonintubated spontaneously breathing patients. All of the models studied should receive oxygen at a rate ≥ 15 L/min. It is not safe to use manually operated self-inflating resuscitation bags for this purpose without knowing their characteristics.

Keywords: Oxygen/administration & dosage; Resuscitation; Equipment and supplies; Respiration; Intensive care.