Ventilator Sharing during Shortages. A Siren's Song?

La seule façon de lutter contre la peste, c’est l’honnêteté.

—Albert Camus, La Peste, 1947

Coronavirus disease (COVID-19) is an extremely dynamic condition, and as of this writing, a total of more than 8 million cases and more than 450,000 global deaths have been confirmed (1). About half of these cases have occurred in Europe and North America. On March 26, an article in The New York Times entitled “‘The Other Option Is Death’: New York Starts Sharing of Ventilators,” echoed that New York Presbyterian Hospital began ventilator sharing during the pandemic. It also mentioned that Governor Andrew M. Cuomo of New York said, “We need 30,000 ventilators. We have 11,000” (2). This critical shortage of life-saving devices and the colossal pressure and uncertainty created by social media and COVID-19 drive an urgent search for solutions. With few companies having the expertise to build ventilators, boosting supplies is no easy task in the midst of a pandemic.

In this issue of the Journal, and in the context of COVID-19–associated acute respiratory distress syndrome in New York Presbyterian Hospital, Beitler and colleagues (pp. 600–604) (3) discuss the feasibility of ventilator sharing, using a single ICU ventilator to support two patients. This approach has been addressed in recent bench studies (4, 5). The original and novel aspect of this research letter, however, is that the authors tested the short-term feasibility of ventilator sharing in patients with COVID-19 with acute lung injury. They provide an accurate description of its technical implementation as well as the potential risks and the way to limit these. Beitler’s comprehensive strategy requires not only careful independent monitoring of each patient but also cautious selection of pairs of patients so as to minimize the risks of major mismatching when patients share a machine with the same settings. How to assess patient compatibility is not straightforward. As the authors underlined, numerous criteria must be met. For example, ventilator settings, respiratory system mechanics, and hemodynamic status must be similar. There can be no contraindication for neuromuscular blockade, and respiratory pathogens should be the same. The authors used the pressure control mode of standard ICU ventilators—a key safety feature. They also used a freestanding respiratory monitor so as to have a continuous display of patient-specific airflow, Vt, airway pressure, and capnography during ventilator sharing. In addition, the indications for this approach were cautiously limited to a 48-hour time span. The authors precluded the use of anesthesia machines for ventilator splitting owing to technical reasons.

The practical limitations of ventilator sharing, however, cannot be overlooked. First, the need for an unoccupied rescue ventilator within each cluster of ventilator-sharing patients, for example, may perhaps defeat the purpose of putting two patients on the same machine. Furthermore, this strategy impedes weaning from mechanical ventilation, a fundamental step to get rid of a ventilator. Ventilator sharing also demands additional patient care, adding even further strain to health systems with already limited resources. And besides the shortage of ventilators, there is also a shortage of personnel with sufficient physiological background and skills to manage severely hypoxemic individuals under mechanical ventilation (6, 7). The clinical approach itself is daunting. Beitler and his colleagues explained that whenever patients were on ventilator-sharing mode, patient care was assured by the usual clinical team and a consult of either of two intensivists familiar with the system who alternated around the clock. They also pointed out that “In acute ventilator shortages, after exhausting alternatives, ventilator sharing is a reasonable stopgap….” What they meant by “exhausting alternatives” is somewhat vague. One may wonder if, as a temporizing measure, manually bagging patients—this time with the only innovation being the addition of a positive end-expiratory pressure valve—would also be a reasonable alternative. After all, manually bagging patients by volunteers was successful during the polio epidemics in August 1952 in Copenhagen (8): mortality dropped from near 90% to <30%. It was a dramatic moment, and its consequences constituted a turning point in the history of medicine (9). Delivering “gentle” and noninjurious ventilation, however, may be more challenging in conditions of acute lung injury than in pure ventilatory failure. An additional strategy might be to have a repository of old ventilators that could be used to provide life-saving treatment in an acute shortage (10). Appropriate upkeep of these machines is possibly more practical than trying to manufacture new ventilators from scratch in a moment of crisis. One may remember that old pneumatic systems, in spite of their limitations, had much longer half-lives than their more modern “electronic” counterparts.

To summarize, what the study by Beitler and coworkers shows is that if patients can be matched in pairs and paralyzed, they can be kept alive under mechanical ventilation using a single ventilator for a couple of days. Feasibility and safety beyond the authors’ highly specific context are unknown, and paradoxically, ventilator sharing might be harmful for both patients if not conducted by personnel with appropriate expertise.

So where do we go from here? Major respiratory disorders are a threat that we cannot continue to ignore as a society. How can we be better prepared for the next pandemic? As mechanical ventilation is clearly the cornerstone for supportive treatment in this setting, more mechanical ventilators will again be needed, as will professionals trained in basic physiology and aware of the complexity, the risks, and the benefits of mechanical ventilation. There is a pressing need to develop a vision, in coordination with global pandemic management. This challenge must be addressed in an honest and nimble way, and better sooner than later.