MANUAL resuscitation bags may be assumed mistakenly to be simple devices, but they consist of multiple valves to create positive pressure, permit exhalation, and segregate inspiratory and expiratory gases. Erroneous assembly or misuse can profoundly impact patient outcome. We wish to draw attention to such a case, which had fatal consequences. Although the resuscitation bag was not used as intended by the manufacturer, we believe that a well-designed device should not permit a misassembly with such serious consequences. We wish to draw attention to the potential hazards of failing to monitor airway pressure during bag–mask ventilation.
A 27-yr-old woman with stage 2B large-cell lymphoma was admitted to the hospital for treatment of spinal cord compression. The results of magnetic resonance imaging were consistent with thoracolumbar nodules, and the patient was treated using intrathecal cytosine arabinoside and methotrexate, in addition to local irradiation. Her disease was complicated further by intermittent fevers, which yielded negative cultures. Three weeks before her death, bilateral perihilar infiltrates developed, along with increasing oxygen requirements, indicative of radiation pneumonitis or opportunistic infection. Empirical antibiotic treatment was initiated. Results of bronchoalveolar lavage were consistent with cytomegalovirus and Pneumocycystis carinii pneumonia infection, and ganciclovir and γ-globulin were added.
Despite maximal oxygen therapy with use of a face mask, arterial oxygen saturation (Spo2) decreased to 85%. The patient and her family had previously requested full resuscitation. After sedation with intravenous morphine and diazepam, tracheal intubation was attempted but was unsuccessful. Glottic visualization and tracheal intubation were finally achieved and confirmed by auscultation. Attempts to transfer the patient to the critical care unit were complicated by a decrease in Spo2to 70%, prompting a return to her room. Manual ventilation using a disposable self-inflating bag (FeatherFlex; Brathwaites-Olivier, Winnipeg, Canada) did not correct the hypoxemia. The reservoir bag was not filling despite high flow, and a decision was made to increase the oxygen supply by connecting a second flowmeter to a proximal nipple (fig. 1). The reservoir bag immediately became distended, and subcutaneous emphysema, abdominal distention, and pulseless electrical activity were noted. Nasogastric tube insertion and bilateral needle thoracostomies were performed with 14-gauge catheters. An air leak was noted, and chest tubes were inserted bilaterally. Incorrect endotracheal tube placement was suspected; the tube was withdrawn and replaced without difficulty. Chest compressions were commenced, and intravenous epinephrine was administered, resulting in restoration of a pulse. Profound hypoxemia persisted, and further resuscitative efforts were deemed futile. Consent for an autopsy could not be obtained.
The resuscitation bag used was not available for subsequent evaluation. However, an identical device was connected to a test lung (Lung Simulator, Model LS122; Medishield, Harlow-Essex, UK) to recreate the problem. The initial failure to inflate the reservoir bag could not be reproduced. We do not have an adequate explanation for this event. Both flowmeters were functional. It is possible that there was a kink in the oxygen tubing.
We demonstrated that the addition of oxygen flow via the proximal nipple, intended for airway pressure monitoring, distorted the inspiratory duckbill valve and prevented exhalation. This is the probable cause of the fatal barotrauma experienced by this patient.
The device was not used as the manufacturer had intended. It is our opinion, however, that such a design is potentially dangerous. Airway-pressure monitoring is desirable to avoid exposing the lungs to excessive pressure. High airway pressures can inadvertently be caused with use of numerous manual resuscitator bags (manuscript in preparation). It is our belief that a manometry connection should be designed so that erroneous assembly will not result in valve malfunction. Ideally, the manometry nipple should also be incompatible with oxygen tubing.
It is not clear whether the effect on exhalation was a design fault or a production fault. The device used is manufactured by Brathwaites but widely sold by other vendors, with various private labels. In Canada, Brathwaites-Olivier has voluntarily recalled and replaced all manual resuscitators with manometry ports (D. Olivier, president, Brathwaites-Olivier, written communication, April 1999). They have also made modifications to an expiratory valve to prevent excessive pressure production within the bag (D. Murdoch, representative, Brathwaites-Olivier, oral communication, December 1999). Such initiatives may not have taken place in other jurisdictions. The described modifications may reduce the likelihood of such an occurrence, but product familiarity and proper assembly are vital. The purpose of this report is to draw attention to the potential consequences of faulty use and to encourage proper use of manometry or the purchase of bag valve devices that do not have manometry nipples.