Matthew Wynia, director, Center for Bioethics and Humanities at the University of Colorado, and James G. Hodge, Jr., professor of law, Sandra Day O’Connor College of Law, Arizona State University, moderated a discussion of ethical and legal considerations within crisis standards of care (CSC). This chapter highlights the ethical difficulties of implementing CSC in real time, the questions and challenges of research to inform these real-time decisions, the differing perspectives of individual ethics versus population ethics, and how those varying points of view can impact CSC planning.
While most CSC conversations often occur in a theoretical, planning environment, there have been occasions where these concepts become operational. Speakers discussed their first-hand experiences with these scenarios, and the ethical conflicts that emerge when difficult decisions have to be made in real time. Wynia began with his experience working on the USNS Mercy following the 2004 tsunami in Indonesia. The ship has a 900-bed capacity, but he said they never had more than about 50 patients on board, and some of his colleagues on board were frustrated at the relatively low volume of service being provided on the ship in view of its potential capacity and the obviously tremendous need. Eventually, he found out that there was an unwritten triage rule being applied by the triage officer on shore that “no one was to die on the ship,” so there was a hesitancy to bring high-acuity patients on board. Sheri Fink, The New York Times journalist
and author of Five Days at Memorial, voiced her frustration in this same scenario, saying she was an aide worker on the ground in Indonesia trying to get patients onto the ship, with little success.
Fink continued with her own first memory of CSC-type operations having a personal impact. Following the 2010 earthquake in Haiti, she was sent to one of the Disaster Medical Assistance Team hospitals to observe and see how care was delivered in a resource-constrained environment. She described a case of a woman who survived the earthquake and came in seeking care because her chronic condition required oxygen. Unfortunately, the hospital had a shortage and was having supply-chain issues, so they were forced to ration oxygen, limiting its use to surgery only. Because this woman did not meet this “alternate standard” of care that had been set, she was triaged to be slowly weaned off of her oxygen. But no one communicated this plan to her, or weaned her from the oxygen, and suddenly, Fink said, she was just taken to another location without oxygen at all. Fink accompanied her and watched as she slipped into respiratory crisis. Fink took a moment to remind the workshop participants how different the experience is when you are in a room developing a hypothetical plan to follow versus what the implications are on the ground when your plan means that patients will die. Fink added that, fortunately, once the woman arrived at the next hospital, they realized her critical state and were able to optimize her and give her the care she needed, and she survived for several more months.
Hodge said that for all the partners working in this arena, you will experience situations like these if you have not already. This is why we need CSC, he argued, because the absence of definitive standards in the void would be heinous. He described his experience with one state legislator, discussing different courses of action in a pandemic. He realized this legislator wanted to mandate that limited resources go first to state government representatives and their families, irrespective of need. These types of political or other actions are possible if we do not implement CSC wisely and ethically, with medical guidance and community input.
Given the critical ethical implications of CSC decisions in a disaster or crisis state, the knowledge base and efficacy of the standards being set are important questions, Fink noted. “What disturbed me after looking at this 10 years ago,” she said, “was that there were just a few small studies looking at outcomes. What would have happened if the agreed-upon standard was applied? Would we have increased survival?” she asked. In H1N1, she pointed out that studies eventually found that you actually needed more
time on a ventilator to improve than the standard time that was given—so the ethical question becomes, is it okay to worsen outcomes based on these standards? She asked if these protocols had been used and tested and stated that it is a duty of professionals to study potential outcomes for what is being proposed. Lee Biddison, chief wellness officer, Johns Hopkins Medicine, admitted it is an uncomfortable but important question. There is a tension between the moral obligation to make it better and having to work with the best information we have available at the time. After creating the CSC framework in Maryland, she said, we thought we should test it, but data to test it with is very difficult to obtain. Hodge added that proving efficacy is difficult in routine medicine, much less with CSC. Yet, planning in advance and executing CSC decisions with the best information available is legally defensible. If findings emerge later that initial choices were not ideal, adaptations in real time are made. Fink added the example of the 2014–2015 West Africa Ebola outbreak, and the delay in doing research because no one knew the best, most ethical way to conduct studies during an ongoing emergency. What are the ethics of coming out of a catastrophe like that without new information to better guide the next one, she asked? Luckily, there was enough research conducted during that outbreak to show the safety and efficacy of a vaccine, which is now able for subsequent Ebola outbreaks (WHO, 2019).
John Hick, professor of emergency medicine and medical director of emergency preparedness, Hennepin Healthcare, University of Minnesota, offered a comment related to sequential organ failure assessment (SOFA) scores often used in resource allocation planning. The SOFA scores are designed as a research tool to categorize groups of patients and their risk of death (ASPR, 2017), where the higher the score, the higher the likelihood of mortality. But Hick explained that SOFA scores were not designed to be predictive at the individual patient level, and instead, should not be used as a comparator between someone that has a resource and someone that needs a resource. However, it should also not be the only tool used in this assessment, he noted. He also added that with the data gathered from the 2009 H1N1 outbreak and other evidence from the United Kingdom, they were able to revise the guidelines in 2014 to be more accurate. Hick argued that the clinical community needs to be careful to ensure that the data are being used for the right purposes and are used in a comparative and not exclusionary way. While we have evolved as professionals in response to ongoing literature, he said, we are trying to work carefully with states that have outdated information written into their plans. He added, “I’ve come around to saying it’s more about the processes in place and less about the practices because they will be incident and disease dependent. But without a process to integrate clinical expertise into these triage decisions, you’ll be out of luck from the start.”
Suggestions from Other Fields
While many comments acknowledged the extreme difficulty in conducting research during disaster scenarios in order to effectively test these protocols and plans, a few participants offered suggestions from other fields as guides. Hanfling suggested having a model similar to the National Transportation Safety Board for a disaster event, which could arrive on a scene quickly after a disaster to assess the landscape of what went wrong. He called for a federal, dedicated capability with the right expertise that can be deployed in real time during a disaster to begin the process of data acquisition analysis and real-time reporting. “That’s what we do in the hospital when cases go awry, and I think within the CSC field we’ve reached a level of maturity that is deserving of this next step,” he said. Looking forward, we need to find a way to ensure we are doing the best possible job we can to protect the public, he added. He also emphasized again the importance of engaging the public, saying that in his experiences, community members get these issues every bit as clearly as the doctors and lawyers in the room. Part of our biggest mistake as responders, he admitted, is that we forget the public is there first, taking care of themselves and each other before responders arrive; our connection to that is critical. Hick added that this type of retrospective analysis is important, but if we find ourselves doing proactive triage, we need a method for ongoing quality assurance to ask whether this is the best thing we should be doing at each stage along the way. Planners need to be willing to go through their process to figure out how they can accommodate additional patients or what paperwork is needed to exceed legislated nursing ratios.
A participant from the American Burn Association offered another idea that could be applied to the difficult scenario of research in disasters, from the battlefields. If you look at trauma in war zones, he said, the joint trauma system has been one of the biggest successes in the military. The military did not have randomized controlled trials, but people were dying and they needed to improve their system quickly, so they practiced “focused empiricism” and developed a learning system that reduced combat casualties to the lowest levels they have seen. Even under such austere conditions, the percentage of wounded service members who died of their injuries in Afghanistan decreased by nearly 50 percent between 2005 and 2013. In 2016, the National Academies published the report A National Trauma Care System: Integrating Military and Civilian Trauma Systems to Achieve Zero Preventable Deaths After Injury, which provided recommendations on how to best incorporate these trauma advances into the civilian sector (NASEM, 2016). A similar “learning health system” effort could be applied to disaster research.
As various stories emerged from the conversation, it became clear that the ethics could be seen from two very different lenses: the individual and the population levels. Fink described a case at Bellevue Hospital in New York City during Hurricane Sandy in 2012, where the intensive care unit thought they were going to lose power, and only had 6 working outlets for 50 patients. Drawing from CSC principles, they began making a list of how to prioritize those six outlets, and came up with a plan that was ethically defensible. However, the head nurse told Fink that they really were lacking guidance on how to communicate that plan to the families of patients that were not prioritized. How do you tell people that they will not have a lifesaving resource, she asked? Biddison also noted that this example from Bellevue highlights the “real-time learning” need that is also embedded in the focused empiricism model. Every epidemic and disaster are different, she said, and part of our responsibility is to challenge people to optimize triage so it can be as effective as possible.
Hodge added that this is an example within the continuum of shifting from individual to broader public health ethics, which requires a different way of looking at the situation. Because we are trying to maximize the number of lives saved, it is not possible to try and save everyone. There may be people who we cannot help, but we are seeking to employ standards to save the most people possible under a public health ethic. A participant from the Office of the Assistant Secretary for Preparedness and Response shared a modeling study that was done in 2011, which focused on triaging surgical procedures 48 hours after a nuclear detonation. He said that the authors found that in poorly resourced settings, “prioritizing victims with moderate life-threatening injuries over victims with severe life-threatening injuries saves more lives and reduces demand for intensive care, which is likely to outstrip local and national capacity” (Casagrande et al., 2011).
One participant noted that these protocols and guidelines suggest we can save many more lives if we prioritize the moderately injured over the severely injured—but we do not know which of the moderately injured lives we have saved. Is it better to save a person you know, or save the lives of many who you do not know, he asked? Wynia agreed, saying the ethics language for this question is the “rule of rescue,” which says that an identified person under threat will tend to receive an inordinate amount of resources and assistance. It is much more difficult to withhold resources from a person when you know their name, face, and identity, compared to withholding those same resources from nameless people who might be in the same situation. He explained that the “rule of rescue” is thought to hold
value for a community, where people can trust that if they were suffering in that scenario, they would receive the same resources. Fink added that when shifting into this public health and utilitarian mindset for ethics on a broader population level, the potential for magnifying health care inequities increases. She highlighted a 2011 PLOS Currents article from Wolf and Hensel that found that some of the protocols have proposed exclusions that “run afoul of federal antidiscrimination laws, including the Rehabilitation Act and the Americans with Disabilities Act” (Wolf and Hensel, 2011). She expressed concern about some of the categorical exclusions based on severe cognitive impairment and having quality-of-life assessments, saying there are several ways that some of these protocols can result in dangerous ethical implications.
Incorporating Equity into Practices
This discussion of differing ethical lenses continued into important points on equity. Biddison described an experience in Maryland in creating CSC frameworks, saying one of their concerns was not wanting to exacerbate existing disparities. Someone who has lesser access to care routinely will likely have more complications of their disease, she explained, so if they are penalized from a resource allocation perspective based on that, it is the moral equivalent of double jeopardy. Hodge added that without attention to CSC and building these frameworks, there are other inequities that would result, such as basing access to care or resources on factors unrelated to medical prognosis, like gender, sexual orientation, ability to pay, or domicile. McKinney also added comments from experience in CSC work in Illinois. All stakeholders come to the planning table from different perspectives; it is very difficult work. The community has to want to be engaged, she said, and people who are historically underrepresented and have adverse health outcomes do not want to talk and think about it. Even starting the conversation of prospective research with these populations is really problematic because of past transgressions, she explained. The important thing at the state and local levels is that even with the obstacles, we keep going, and keep showing up. Hick added another consideration when thinking about incorporating equity into the health and medical landscape. When we talk about social equity, he said, you cannot ask from medicine what medicine cannot deliver. As providers, the only factors we can really take into account for triage practices are medical. Unless we can get a framework for equity, with traction and support from the community, then we really have to stick to medical factors that we know, Hick explained.
Existing Ethical Consensus and Potential Next Steps
In conclusion, speakers offered issues within CSC that they believe to have “ethical consensus” to show the progress made in the field over the past several years. Hodge suggested going beyond simply talking about equity, and actually operationalizing it. CSC is legally defensible. The next step is to build equity into a CSC plan. Biddison sought to keep systems out of the crisis and contingency phases of care, and said there is agreement on the importance of doing everything possible to stay out of situations requiring altered standards. Fink highlighted the duty to plan that falls to government and health leaders in communities, but she noted a need for wider dissemination. The effort over the past 10 years has been so worthwhile, she said, but after years of gaining input from the community and other providers and jurisdictions, how can we translate that to practice? Wynia said he was glad to see the increased focus on process and guidance rather than adherence to substantive rules and tools that are strictly applied. He also offered the consensus that no one should be left to die—though Fink cautioned the group about the moral distress that health care workers experienced during the Ebola outbreak, and how resource intensive the provision of palliative care can be. Finally, Wynia said, we need stories. Wynia elaborated that we have talked about engagement of politicians and health care leaders, but the type of engagement we really need is deliberate, bidirectional, and longitudinal. We will not get people involved through statistics and reports, but we really need to collect and tell stories in a way to capture the attention this topic deserves.
This page intentionally left blank.