|Questions for Dr. Lewis Rubinson from the Flu Wiki Community
Q. Dr. Rubinson, in the current Chest supplement, the task force writes
"This Supplement on the management of mass critical care for ill patients represents the consensus opinion of a multidisciplinary panel convened under the umbrella of the Critical Care Collaborative Initiative." How and why was the task force convened? Do you feel the publication of the reports contained in the supplement ends the work of the task force? If not, what's next? If so, what's next from other professional specialty groups, in your opinion?
A. In 2005 the Working Group on Mass Casualty Critical Care, also a large group of senior-level experts, published a conceptual framework for increasing the numbers of people who can receive life-sustaining care in the wake of an epidemic. This initial guidance was a good start, but it lacked sufficient details for hospitals and regions to simply translate into preparedness.
In 2006, there had been increasing grass roots uneasiness among critical care health professionals that current levels of clinical guidance remained inadequate to provide optimal care to their communities were a severe influenza pandemic to occur. Subject matter experts representing several society members of the Critical Care collaborative elected to provide much more detailed guidance so hospitals could prepare for and achieve large critical care surge capacity.
Also, Minnesota, Ontario and New York State had proposed means to optimize use of scarce resource such as mechanical ventilators during extreme catastrophes. These activities were well done, but there were some operational issues which remained undeveloped. In addition, there were compelling reasons to create a national framework. The collaborative therefore elected to further develop an allocation framework which could be used across the nation.
The steering committee recognized that the evidence-base was limited for mass critical care. An event would have to occur for data to become available, and we felt it unacceptable to wait for such a situation to undertake preparedness efforts. Given the paucity of evidence, expert opinion would be needed for most of the guidance. World-class experts from a number of fields were assembled to give the best possible expert guidance. It is worth noting that the focus of the Task Force was a severe influenza pandemic, but all of the recommendations were developed to be relevant and appropriate for responses to other hazards which could cause mass critical care (e.g. large-scale inhalational anthrax or chlorine inhalation.
These manuscripts represent another step in critical care preparedness, but they are not the end of the work. Much work now must be done to implement Emergency Mass Critical Care and fair and just allocation of scarce life-sustaining resources at the hospital, regional, state and federal levels. Establishment of effective decision-making bodies at all of these levels, establishing detailed response processes and assuring excellent real-time bidirectional communication will be essential. Much work remains in most communities.
Utlimately the project was formatted as follows:
Using a broad-based panel of experts; 1) Undertake a rigorous gap analysis to identify key limitations to current critical care disaster response and use this information to 2) develop guidance to drastically increase the number of people who can receive life-sustaining care in the wake of a disaster. Lastly, for the most catastrophic events where the "surge capacity" was still not enough to assure all seriously ill had access to life sustaining interventions such as mechanical ventilation, we set out to develop a fair and just means to allocate these scarce resources.
Our group did not deal specifically with pediatric issues. The Centers for Disease Control and Prevention will be providing assistance to assemble a future Task Force for pediatric critical care surge capacity and scarce resource issues.
I believe the most important next efforts are to engage community members-at-large and front-line clinicians to evaluate, revise and improve our guidance. Also community members must determine the value-sensitive issues. A major question which remains is what is the right outcome to optimize if we switch from individual to population-focused care. Should minimizing mortality alone or maximizing expected years of survival (this would weigh care towards pediatrics) or expected "quality" years of survival be used to prioritize allocation. Only the community can ultimately decide.
Lastly,a number of Task Force's clinical suggestions require validation and will require research.
Q. In the report, the task force covers the shortages likely in a disaster such as a pandemic ("staff, stuff and space"). As part of that, ventilator needs are outlined. Non invasive ventilatory support is mentioned in passing.
"The Task Force concurs with these cautions and does not propose NIPPV as a principal strategy for managing mass casualty respiratory failure, for the following reasons: the need for experienced users28; requirement of significant initial staff time29; limited benefit and infrequent use in practice for ARDS; and the uncertainty as to whether NIPPV may generate significant respiratory aerosols that would be difficult to scavenge during an epidemic of a respiratory-transmitted pathogen" Can you elaborate on that, particularly in light of experience with NIPPV and SARS in Hong Kong?
A. In the US, the Department of Homeland Security National Planning Guidelines are intended to coordinate and prioritize emergency preparedness efforts at all response levels. Contained within the Guidelines are 15 National Planning Scenarios, and at least two-thirds of these may cause catastrophic levels of acute respiratory failure (2). Successful surge PPV strategies must be grounded in accurate predictions of patients' needs as well as healthcare systems' and communities' capabilities for these events. Predicted distributions of types and severity of respiratory failure as well as the characteristics of the affected population (e.g. underlying COPD or previously healthy) vary widely. This is where the direct evidence-base is thinnest. The predicted distribution of respiratory failure even for a specific disease such as an influenza pandemic cannot be definitively known currently. Despite these cautions,there are a number of reasons to assume that most people with respiratory failure due to novel influenza will have hypoxemic respiratory failure due to unilateral pneumonia or ARDS/ALI.
The context for the Task Force's recommendations for mechanical ventilation included: 1) most patients without underlying COPD will have acute respiratory distress syndrome (ARDS) or hypoxemic respiratory failure due to pneumonia, 2)need far surpasses resources (including staff and equipment), and 3)the cause of mass respiratory failure could be transmitted nosocomially.
The most recent reviews of NIV support its everyday use for acute respiratory failure in hemodynamically stable patients without ongoing cardiac ischemia due to acute exacerbations of chronic obstructive pulmonary disease, cardiogenic pulmonary edema, or ALI/ARDS in immunocompromised patients. Its use in hemodynamically unstable patients and most patients with hypoxemic respiratory failure due to ARDS or pneumonia is much more uncertain. If the simple consideration is respiratory failure and resources are not limited and the potential for secondary disease transmission are not factors, than a trial of NIV in some patients may be warranted. For everyday patients, though the number who can be successfully ventilated is still a small number, and for those who fail the mortality may be higher than if they were intubated right away (as with post-extubation respiratory failure, and suggested by Schettino G, Altobelli N, Kacmarek RM. Noninvasive positive-pressure ventilation in acute respiratory failure outside clinical trials: experience at the Massachusetts General Hospital. Crit Care Med. 2008 Feb;36(2):441-7).
Assuming surge PPV must be able to manage ARDS is not unrealistic. Patients who have persistent PaO2/FiO2 > 300 and without other indications for intubation or mechanical ventilation, will not require PPV---they should not to be considered for surge PPV. For those requiring PPV due to hypoxemia, the spectrum of disease is unlikely to be well defined by the level of respiratory failure (e.g. presence or absence of ARDS). Patients with and without ARDS/ALI with low PaO2/FiO2 have similar mortality. Only very low PaO2/FiO2 correlated with mortality and discriminating between patients with and without ALI/ARDS is not very reliable. Also most patients with ALI meet criteria for ARDS very early in disease. Even in a group of ARDS patients (mean PaO2/FiO2 128.6 + 33.3 at day 0), outcomes have been shown to be highly variable. The temporal changes in PaO2/FiO2 suggested that initially patients are quite ill and then some start to resolve their lung injury. This disease course suggests an upfront strategy of a less supportive PPV device may be imprudent. Most patients with ARDS die of multi-organ dysfunction and not refractory hypoxemia. Hence, spectrum of severity of disease during mass respiratory failure events will probably vary by number of non-pulmonary organ dysfunction and presence of hemodynamic instability. ARDS alone is not the worst case scenario and does not describe a futile response situation. We therefore believe that surge PPV equipment must be able to successfully manage patients with ARDS.
In 1999 Confalonieri demonstrated a reduction in intubation rates (n=56, 21% vs 50%, p=.03) for patients with acute respiratory failure due to community-acquired pneumonia during a randomized controlled trial of NIV versus standard treatment. More than 40% of patients had COPD; these patients had a reduction in intubation, ICU and hospital lengths of stay, and 2-month mortality. For the patients without COPD, there was no statistical difference in intubation rates and a non-statistically significant point estimate increase in hospital mortality and 2-month mortality. Ferrer et al studied NIV in patients with acute hypoxemic respiratory failure without both underlying severe chronic respiratory co-morbidities and hypercapnea. Intubation rates, incidence of septic shock and ICU mortality were all lower in the group receiving NIV. These results are compelling, but it is worth noting that they may not be generalizable to all comers with respiratory failure since 64 of 172 eligible patients (hypoxemia on 50% delivered through a venturi facemask) were excluded. Some were unable to cooperate with NIV due to agitation (n=45), had a severely depressed level of consciousness (n=5), hemodynamic instability (n=4) or required immediate intubation (n=10). Also worth noting is intubation and mortality rates were not better with NIV versus standard treatment in the approximate 15% of patients with ARDS. In contrast to the Ferrer et al study demonstrating the benefit of NIV in select patients with pneumonia, Honrubia et al showed no apparent benefit to NIV in patients with pneumonia. As with the other studies, significant proportions of patients with respiratory failure were unable to be randomized.
The above studies may be optimistically interpreted as NIV having a role for a select subset of patients with hypoxemic respiratory failure; however, the utility in all-comers with respiratory failure due to pneumonia and ARDS remains suspect. Of course use of NIV in hemodynamically unstable patients is even more uncertain. For a 20 day period in 2002 among 70 ICUs in France, 1076 of 1943 (55%) patients required PPV. The majority of patients (74.9 %) were invasively intubated prior to or at ICU admission. 55.8% of those who received NIV (12.9% of all requiring PPV) were able to forestall intubation; patients with de novo respiratory failure who had RR> 35 and PaO2/FiO2 <200 were more likely to fail NIV. Again NIV may have had benefit in a small subset of patients, this time demonstrated in everyday practice, but overall, NIV was not an appropriate option for the overwhelming majority of patients. This is likely in part due to patients with conditions less likely to respond to NIV (e.g. de novo respiratory failure (42%) being more common than patients with acute on chronic respiratory failure (16%).
Another explanation could be that the limited overall benefit of NIV was due to missed opportunities through underutilization of NIV in France. This explanation is not supported by recent reports from ICUs which are known to have extensive experience with NIV. At these hospitals, the overall percentage of patients with ARDS and acute hypoxemic respiratory failure who can remain unintubated is relatively small. Antonelli et al, demonstrate a 50% success rate of NIV after patients with hypotension, excess secretions, more than one organ failure, bleeding, and neurologic disturbances were excluded. Close inspection of the data reveal that of the 479 patients meeting the ARDS criteria, 322 (69%) were already intubated due to altered mental status, inability to manage secretions, hemodynamic or electrocardiographic instability, severe trauma, and/or more than two organ failures. The remaining 147 patients (30.6%) were then studied. In fact then, only half of these select patients with ARDS (15%) were successfully ventilated with NIV. The experience that a high proportion of patients with hypoxemic respiratory failure cannot be successfully managed with NIV, was also reported from a large US academic center. These authors as well as other recognized ventilator experts have therefore cautioned against liberal application of NIV to patients with ALI and ARDS.
In summary, NIV may have a role for respiratory failure due to novel influenza for some patients with hypoxemic respiratory failure, but caution should be applied when expecting to manage all patients with such modalities.
Importantly our Task Force considered mass respiratory failure when resources are far outnumbered by need. For these situations, we recommend even further caution about use of NIV due to the risk of worsening respiratory failure in patients who cannot be adequately monitored by skilled health professionals due to the resource limitations. Several studies have shown that for patients with hypoxemic respiratory failure, a low PaO2/FiO2 after 1 or 2 hours and a high SAPS II score portend need for intubation. SAPS requires a number of laboratory studies and ideally should be scored as the worst variables over 24 hours. Time and laboratory capacity may not always be abundant resources during mass respiratory failure. Also, experienced staff are likely to be in short supply and patients who fail NIV may not be identified at the 1 hour and 2 hour mark, which data support as the period as being very important to get patients to tolerate NIV and identify failure. Even after the first day, intense observation will still be necessary. Schettino et aL reported that 38% of NIV failures in hypoxemic respiratory failure occurred after 24 hours and Antonelli reported that 30% of failures occurred after 48 hours. A high proportion of these patients required intubation and their mortality rate was high.
The third consideration is the potential for disease transmission to staff and other patients. the uncertainty regarding NIV and secondary transmission of respiratory infections. Some facilities used NIV successfully during the SARS epidemic albeit with many modifications (expiratory filters in rooms with negative pressure and > 8 air exchanges as well as use of PAPRs), others restricted use, and some reported a possible mode of transmission. If the pandemic strain of influenza is able to bind to receptors on the proximal respiratory epithelium than perhaps transmission can occur by NIV. Also, if we apply the small series of NIV reporting on SARS transmission from Asia methodology to all personal protective equipment, we could argue that the lack of secondary transmission of SARS in the US despite inadequate PPE use, argues that even basic PPE may not needed (Park BJ, Peck AJ, Kuehnert MJ, Newbern C, Smelser C, Comer JA, Jernigan D, McDonald LC. Lack of SARS transmission among healthcare workers, United States. Emerg Infect Dis. 2004 Feb;10(2):244-8.) I would caution against this interpretation. Paramedics are not obligated to rescue a person in an unsafe building who would clearly benefit from their assistance (e.g. fully involved with fire or structurally unsound). Similarly, staff providing respiratory care for patients with contagious diseases for which effective prophylaxis is not available and which can cause severe disease should be protected. Data are limited which supports NIV and SARS transmission, but that does not mean that the data are conclusive nor do we understand if NIV will be implicated with transmission of other pathogens such as influenza. We therefore stand by our recommendation not to plan for widespread use of NIV even with existing equipment, when a contagious pathogen is suspected.
While NIV devices should not be stockpiled, existing devices at hospitals and other clinical facilities which are capable of ventilating a patient through an endotracheal or cuffed tracheostomy tube can be re-purposed during a disaster.
Q. The scope of the task force's evaluation covers the difficulties encountered when patient needs outnumber critical care beds and staff. Does this suggest to task force members the need to consider home care and pre-pandemic preparations by individuals as a needed adjunct? If this is not the scope of the task force, then whose question is it to address?
A. The Task Force does not recommend home care for patients with acute respiratory failure requiring mechanical ventilation or hemodynamic instability. We do however believe in expanding the use of long-term care facilities, alternate care sites/facilities and home health for non-critically ill people who would traditionally be hospital inpatients. Models for such planning have been expanded over the past decade (SBCCOM, AHRQ and other organizations have published concept documents). A number of community efforts and national working groups have tackled various pieces of this planning.
Q. The Agency for Healthcare Research and Quality has developed an emergency program called Project Xtreme to help extenders supplement respiratory needs in a disaster or emergency . Did the task force look at this? Are there other things that emerged as promising answers to the 'staff' question?
A. We direct the audience to our Chest manuscript on medical resources for critical care surge capacity. In the section on staffing we specifically address Project Xtreme. I was a member of the External Advisory Committee for Project Xtreme. I had extensive familiarity with the project. Rich Branson was the AARC rep to the Task Force. While Rich was not directly involved in development or evaluation of Project Xtreme, he was well versed in the effort through his role in AARC. Capt Ann Knebel's office co-sponsored Project Xtreme with AHRQ. She was also a member of the Task Force and she was very familiar with Project Xtreme.
To address the second part of the question, I recommend the section on staffing from the Working Group on Mass Casualty Critical Care as well as our manuscript which covers medical resources.
Q. The issue of palliative care was raised in the recommendations. The task force recognized its importance, but in the face of potential medication shortages ("stuff"), are there recommendations as to how to implement palliative care?
A. This is a great question as the actual palliative care resources which would be required are unknown. We have an ongoing research project related to this. That being said, drugs to reduce anxiety and suffering may be available in sufficient quantities, can be substituted with a variety of drugs in the same class, can have their use modified to extend their availability and are not prohibitively expensive if hospitals want to increase the PAR value for the amount they keep on hand.
More specific recommendations for palliative care would be very useful and we encourage additional work on this important topic.
Q. Do you feel the task force report, given the fairly wide media coverage, has reached the people you wanted it to? Have you gotten feedback from hospitals and critical care units?
A. The message which was initially put out through Lindsay Tanner's AP article was sensationalized and not very comprehensive. We hope to see many communities engage in this topic but believe sensationalized journalism is not the way to take on this difficult topic. We believe many community groups must get involved with these discussions, planning recommendations and implementation activities. I do not believe we have done an effective job at getting this discussion going. We are looking at future activities to more deliberately engage communities.
We have been presenting drafts of this guidance for awhile at many medical and public health conferences. The feedback has usually been very positive. Our recent experience since the manuscripts have been published has been similar. There is much room for improvement and I hope as the clinical community evaluates our suggestions, their insights, recommendations and additional efforts will improve the current planning suggestions.
Q. It's not recommended that tamiflu be acquired before a pandemic, but early use, if administered at the onset of illness, might ameliorate the need for critical care, at least for some. Given the task force recommendations and conclusions, do you feel this changes the equation in regard to tamiflu prescriptions pre-pandemic?
A. Tamiflu resistance does not necessarily correlate with tamiflu use (Japan has low level of seasonal flu resistance but uses a lot of tamiflu, and some Scandanavian countries have high resistance and limited tamiflu use). Given the logistical challenges of distributing tamiflu during a pandemic to all those in need, I believe there is not consensus about minimizing tamiflu access pre-pandemic. I am not sure our work specifically affects these decisions, since the logistical challenges to pandemic tamiflu distribution have been known for some time.
Q. Do you have any suggestions as to how these recommendations might be practiced or drilled?
A. Exercising and validating the suggestions are very important.
There are processes to be evaluated: For instance, one can test communication (initiation, termination, protocol modification) and uniformity of implementation across hospitals.
For validation of the actual clinical protocols, one can look at appropriate elements of everyday critical care patients for process validation.