Public health decision-making during the 2009 H1N1 flu epidemic

Swine Flu

You remember the great swine flu epidemic of 2009, right? Really, you don't remember the school's being closed across the country after the first wave of fatalities? And how people stopped eating pork to such an extent that farmers simply slaughtered most of their pigs and then burned the meat? You don't remember that? Well, of course not. No-one does, because it didn't happen. It also true that no one knows why it didn't happen.

The interesting question at this point, as one looks back to the way in which decisions were made to stop an epidemic before it got started. In the aftermath of the flu season, when there had been no outbreak, many people criticized public health officials for having over-reacted. Those officials in turn argued two points. First, it was better to over-react than to under-react because the consequences of the former were far more dire than the consequences of the latter.

Furthermore, public health officials argued, there was no way to determine in the aftermath of the flu season to what extent the actions that had been taken had prevented an epidemic or pandemic of flu. Sometimes the fact that there is not a catastrophe that proves that the degree of response was exactly right. This paper examines the decision process that was followed in preparing for the 2009 flu season.

Rather than making any judgement about whether the right decision was made -- because it is of course impossible to determine afterward whether an epidemic was suppressed or simply never arose -- this paper walks through the decision-making process according to different models of choice that emphasize different underlying assumptions.

Bazerman and Moore's Decision-making Model

Bazerman & Moore (2008) and Bazerman (2008) have developed a six-stage decision-making model that can be applied to any field. This section of the paper uses this model to analyze the decisions made in preparing for the 2009 flu season. As will be discussed at greater length following the breakdown of the individual steps of this model and the ways in which they could have been used in setting public health policy, while this and similar models have many supporters, such models are often criticized for being overly mechanized and reductionistic.

The first step in Bazerman & Moore's (2008) model is to determine whether or not there is a problem. This sounds so elementary as to not needing to be said, but it is remarkable how often the "if it's not broke don't fix it" axiom gets ignored. Also included in this step are the following considerations: If there is a problem, does it have to be solved now? What happens if the decision if delayed? Who is responsible for making the decision? These steps are intuitive and indeed would be the first steps for most decision-making models that are being applied directly to problem solving.

The answer to whether the H1N1 flu virus had the potential to cause significant harm was certainly yes, something that was well-established in public health research. In the case of a potential flu pandemic, there is always the possibility that there may be truly catastrophic consequence. Indeed, thousands and usually tens of thousands of Americans die every year from the flu, and at least half a million people die from flu annually, many from secondary bacterial infections (Influenza, 2003).

In fact, the worldwide death rate of individuals from flu may be much higher than this because countries where there is likely to be a very high rate of infection are those same countries (primarily undeveloped nations in the tropics) that are both burdened with highly limited public health systems and limited means (and perhaps limited incentive) to report the actual rates of morbidity and mortality rates from flu. Even developing nations such as China have been slow to report the levels of infection.

The [government's] threats [to report the number of flu cases accurately] came after a prominent Chinese doctor, Zhong Nanshan, questioned why China's official H1N1 death toll remains in the double digits while other countries with much smaller populations post much higher numbers. He's convinced that [national] the government is covering up the spread of the virus.

That's a serious accusation in China, which suffered global embarrassment after attempting to conceal the deadly SARS [Severe Acute Respiratory Syndrome] virus in 2003. This time, China's sheer size may contribute to the problem of under-reporting. (Hatton, 2009)

The fact that initial reports of the seriousness of flu transmission rates may be artificially (even if not intentionally) low must also be considered in assessing how long any decision on acting on a flu epidemic can wait because initial reports may suggest that the threat is far less than it is.

The Spanish flu pandemic of 1918-19, which killed more people than had died in World War I, swept across the globe along with the soldiers going home from the war. The flu killed even healthy adults, who could feel "fine in the morning" and yet be dead by the time that the sun had fallen (Pandemic flu history, 2011).

Some of the statistics that help make the extent of the Spanish flu pandemic are the following: 1) Approximately 20% to 40% of the worldwide population became ill; 2) An estimated 50 million people died 3) Nearly 675,000 people died in the United States (Pandemic flu history, 2011). One of the reasons that this flu pandemic proved to be so lethal was that the virus responsible for it was new and so people did not have an immunity to it.

Flu viruses, like most viruses, can change quickly. It is for this reason that having the flu one year does not protect an individual the next year. Although the virus that comes around each year is similar enough to the previous version to be considered to be a form of the same disease, it can be significantly different. H1N1 was a virus that was significantly different from previous flu viruses, and so it had a greater potential than most versions of the flu to cause a high rate of infection and a high rate of death.

"It's a new virus whose biological properties we're still not sure of," said Andrew Pekosz. "And with influenza, it's been documented that different strains have a differential ability to cause disease in animals and in humans. And right now, we're getting mixed signals about this virus' ability to cause disease." (Researchers race to develop swine flu vaccine, 2009)

The 1918 flu pandemic also spread so quickly because it was carried home across the globe by soldiers returning home from war. H1N1 would not be transmitted in this way, but with international travel of hundreds of thousands of people during a single flu season, the spread of the disease could be incredibly rapid.

There had not been a flu pandemic as severe since that time, but the devastating consequences of a flu virus to which the human population had not been exposed, which had been demonstrated by the 1918-19 pandemic, that the emergence of a new strain of flu was enough to set off alarm bells throughout the public health community. Thus, step one of Baverman & Moore's model had been satisfied: There was a potentially serious problem, and it was one that had to be acted on with alacrity.

One of the key axioms of public health is that a vaccination program to be effective it must be started early on. Once a certain percentage of the population becomes infected (what the percentage is depends on a number of factors, including the type of pathogen) then a vaccination program will have very little effect. This precept is known as "herd immunity" or "community immunity" and is a well-established medical fact:

When a critical portion of a community is immunized against a contagious disease, most members of the community are protected against that disease because there is little opportunity for an outbreak. Even those who are not eligible for certain vaccines -- such as infants, pregnant women, or immunocompromised individuals -- get some protection because the spread of contagious disease is contained. This is known as "community immunity." (Community immunity, 2010)

This concept of "herd immunity" answers the question of why it would have been highly problematic to wait any significant length of time before making a decision to act.

Among the other key questions posed in step one of this model is, after determining that there is a problem is to determine who is responsible for offering a solution to the problem. The answer to this part of the model is that the most important actors in the solution are public health officials working on national and international levels. Viruses can fly across an international border instantaneously and so any intervention that has a chance of being effective must have international cooperation.

But other actors are also responsible as a necessary part of any intervention: Primary care physicians for diagnosing flu patients, individuals for being vaccinated and for acting in ways that limited the potential of infecting someone else, local officials for making decisions such as whether to close schools, virologists and other researchers for creating a new vaccination against the H1N1 virus.

Finally, in addition to the cost in health and human suffering, flu pandemics are costly in simple monetary terms. According to the World Health Organization:

Influenza rapidly spreads around the world in seasonal epidemics and imposes a considerable economic burden in the form of hospital and other health care costs and lost productivity. In the United States of America, for example, recent estimates put the cost of influenza epidemics to the economy at U.S.$71-167 billion per year. (Influenza, 2003)

One final point must be considered in assessing the need for a speedy decision in terms of public health actions taken when considering how to respond to a flu pandemic: The length of time required to make a vaccine that would prove to be effective against the H1N1 virus. While, of course, a flu vaccine already existed, the fact that the flu season actually consists of a different viral mix each year, each year requires a reformulation of the flu vaccination if it is to be effective.

To have vaccines ready, you really have to be preparing six to eight months in advance," Brennan says.

Once a virus is identified, one of two FDA-approved methods is used.

In the first, called the attenuated method, drug companies grow the new strain of virus in chicken eggs. Even though vaccine is more efficiently produced than in the past, say WHO officials, some infectious-disease experts say eggs can be unreliable. "They can go bad, the strain might not grow fast enough, or at all," says Brian Currie, medical director for research at Montefiore Medical Center in New York.

Brennan says it can take months to harvest the virus. (Marcus, 2009)

Step Three

Step three of this model focuses on weighting the different possible steps in the process. The CDC argues that there are three key steps in reducing the spread and effects. The first of these is to vaccinate as many people as possible, so this must be weighted the most heavily. (What the CDC does not note is that a previous step is to create the appropriate virus, so in fact this is the step that should be most heavily weighted.)

The next step is 1) to take all possible steps to prevent transmission. This includes a number of simple but imperative steps such as handwashing, not going to work or school when sick, and covering mouth when sneezing or coughing (CDC Says "Take 3" Actions To Fight The Flu, 2011). To get individuals to incorporate these actions into their daily lives takes repeated reminders and even warnings about the consequences of not doing so, and even after such a barrage of information many individuals may not follow any of the advisories that they hear.

The final step, which would receive the lowest weighting, is for those people for whom it is medically advisable to receive anti-viral drugs. These drugs are highly effective: It is not that they are ineffective that this step is given the lowest weighting but rather that it affects the fewest number of people so that it is ranked the lowest in terms of urgency.

Steps Four through Six

The next step of Bazerman & Moore's (2008) model is the least relevant to this particular problem since it asks the problem solvers to reexamine solutions that had already been dismissed and to come up with reasonable alternatives. The history of flu pandemics has demonstrated that there are no reasonable alternatives to widespread vaccination, all reasonable methods to prevent transmission, and providing primary medical care to those who are infected. Given the success of these steps over the past century, there is no reason to question their validity.

Having noted this, the only possible alternative that has not already been assessed that has had some public health success in limiting transmission of infectious diseases is the policy of quarantining those who have become sick and are in the contagious phase of the illness. While there has been some success in terms of quarantining small populations (such as foreign visitors to China) or small island nations (Wilson etal, 2008), large-scale quarantines of populations is not feasible on a political level, may violate U.S. civil rights law (as well as civil rights laws in other nations), and are unlikely to be effective in terms of virology (Wilson etal. 2008).

It would be possible to isolate some animal populations such as the swine populations in which the virus was first isolated. However, while this would circumvent legal and civil rights issues, this too would prove to be ineffective in terms of virology and transmission. Because there are no reasonable alternatives, this step can effectively be dismissed in this particular application of this model.

Step Six, which focuses on weighting the steps of the alternative step (s to determine which of these steps would be the most important to implement first or the most important one to which to assign resources. This step too has very little relevance for this particular problem, as noted above, because there are no acknowledged effective alternatives for public health responses to flu pandemics. Given the long history of research into this issue, it seems unlikely at this point that significant alternatives will appear.

Limitations of Baverman & Moore's Model

As noted above, while Bazerman & Moore's (2008) model has many valuable elements to it -- perhaps the most valuable of all being that the model focuses on re-evaluating alternative processes even after one has selected a working model, a step that is much too frequently ignored -- it has also received significant criticism for being overly mechanistic and simplistic and insufficiently adaptive, as the above application of this model shows. Applying this model to the H1N1 planning processes, in fact, demonstrates both its strengths and its shortcomings.

Some decisions, at least, seem inappropriately based on doing the numbers. But is the emotional dismissal of Bazerman's 6-step calculation method justified? We can certainly see some notable advantages of the calculation method over the intuition method. First, it is set up to avoid neglecting relevant alternatives and goals. Second, it makes explicit the consideration of how the various alternatives contribute to the various goals. Third, it puts the decision making process out in the open, enabling it to be carefully reviewed by a particular decision maker and also by others involved in a group decision process. (Thaguard, 2001)

Thaguard contrasts this by-the-numbers decision-making model with one that emphasizes intuition. A first glance at his description of the intuitive decision-making model suggests that it would not be particularly useful in making decisions about public health issues.

Thaguard argues that there are two highly important advantages of intuitive decision-making: 1) "One obvious advantage is speed: an emotional reaction can be immediate and lead directly to a decision." And 2) "Another advantage is that basing your decisions on emotions helps to ensure that the decisions take into account what you really care about. If you are pleased and excited about a possible action, that is a good sign that the action promises to accomplish the goals that are genuinely important to you." 3) "Finally, decisions based on emotional intuitions lead directly to action: the positive feeling toward an option will motivate you to carry it out." (Thaguard, 2001)

Making a speedy decision, we have already established, is important in public health decisions and being committed to a course of action when the situation has such high stakes is also beneficial within the realm of public health. However, personal opinion based on feeling is not something that should be relied on when dealing with scientific and medical decisions. (A caveat here. Experts in any given field may base their decisions on what they describe as intuition but may in fact be an opinion well-grounded in experience and knowledge.

Thaguard (2001) goes on to describe the qualities of intuitive decision-making strategies:

1) [E]motion-based intuitive decision making can also have some serious disadvantages. An option may seem emotionally appealing because of failure to consider other available options. & #8230; Intuition is also subject to the intense craving that drug addicts call "jonesing." If you are jonesing & #8230; your intuition will tell you to choose what you crave, but only because the craving has emotionally swamped other desires that you will be more aware of when the craving is less intense.

2) Another problem with intuition is that it may be based on inaccurate or irrelevant information. ... It is difficult to determine introspectively whether your intuitions derive from reliable and relevant information.

3) [I]ntuitive reasoning is problematic in group situations where decisions need to be made collectively. If other people disagree with your choices, you cannot simply contend that your intuitions are stronger or better than the intuitions of others. (Thagard, 2001)

This last condition alone disqualifies intuitive-based decision-making models from being considered to be the best option for public health decisions, which must necessarily be cooperative decisions. Public health decisions at a minimum must involve the expertise provided by researchers and physicians as well as government officials who have the power to set policies that will implement the best practices that researchers and doctors recommend.

In addition to these two groups, other groups must be recruited to participate in public health decisions, from local government officials, to individual families. Such a set of decisions, ranging from cutting-edge research to family decisions about whether an individual can afford lost wages for taking time off from a job when sick, cannot be made on the basis of intuition or emotion. There is simply no way in which such decisions could be resolved if undertaken with such a process.

However, Thaguard (2001) does suggest an additional model or underlying philosophy of decision-making that has not yet been discussed in this paper but that shows important promise for making the kind of rapid, complex, and very important kind of public health decision described here.

Thaguard (2001) refers to this as the "informed intuition model":

This procedure combines the strengths and avoids the weaknesses of the intuition and calculation models of decision making. Like the intuition model, it recognizes that decision making is an unconscious process that involves emotions. Like the calculation model, it aims to avoid decision errors caused by unsystematic and unexamined intuitions. One drawback of the Informed Intuition procedure is that it is not so intersubjective as the calculation model, in which the numerical weights and calculations can be laid out on the table for all to see. & #8230;

The members of the group may be poor at specifying the emotional weights they put on different goals, and they may be unaware of their assumptions about the extent to which different actions facilitate different goals. Achieving consensus among a group of decision makers may require extensive discussion that reveals the goals and beliefs of decision makers to themselves as well as to others. & #8230;

The discussion, including the exercise of working through a calculation together, may help the members of the group converge on evaluations of goal importance and belief plausibility that produce a shared reaction of emotional coherence. Scientific consensus concerning competing scientific theories can emerge from a process of individual coherence and interpersonal communication. (Thaguard, 2001).

So did the public health officials make the right decision to notify people that H1N1 might be terrible? Did their preparations spare many lives? And did sparing those lives in 2009 mean that people will not take the next set of warnings seriously, leading to the loss of far greater life than were saved three years ago?

As unsatisfying as this answer is, there is simply no way to know.