When the spread of disease becomes a global event: The classification of pandemics

Abstract

The classification of novel disease events is central to public health action surrounding them. Drawing upon the sociology of scientific classification, this article examines the role and contestation of the World Health Organization’s Pandemic Alert Phases, as applied to the spread of 2009/10 H1N1 Influenza. The analysis of World Health Organization texts, including policy documents, public statements and epidemiological documents, has been utilized to examine the Organization’s actions and public narratives around the event of H1N1. Analytically, the functional role of such classificatory schemes and the social construction of scientific classifications are examined. It is argued that in understanding the World Health Organization’s 2009/10 application of the Pandemic Alert Phases, the critical limitation of the functions served by the classificatory scheme led to the breakdown of its construction. This case study highlights the importance of classification for the successful production of scientific ‘facts’, the constructed nature of classificatory systems and the potential for contestation that arises when such classifications do not adequately fulfil their functional roles.

Keywords

classification functionality pandemic sociology World Health Organization

Introduction

Classifications are central to the epistemological, institutional and political ordering of reality. Within the sciences, classificatory schemes are crucial to the investigation and understanding of the natural world, from the basics (e.g. the periodic table in chemistry) to complex systems of understanding (e.g. diagnostic manuals). Under the conditions of normal science (Kuhn, 1970), scientific classifications are taken for granted as grounded in an ‘objective’ and measurable reality, black-boxing the assemblages of actor(–network)s that underpin them (Latour, 1987, 2005; Law, 1992). Historically, the act of classification has been pivotal to the development of scientific disciplines (Stichweh, 1992), as well as being fundamental to everyday scientific practice (Roth, 2005). Bloor (1982a), in an essay on Durkheim and Mauss’ (1963) work, notes the foundational nature of classifications, which, as Hesse (1974) suggests, do not merely mirror natural realities but serve to constitute the phenomenon they classify. Understanding disciplinary-bound systems of knowledge and their development means understanding the classificatory systems surrounding them, as evidenced (for example) by the work of Bloor (1982b) on the development of German mathematics in the 19th century, Weindling (1981) on cell theory in Imperial Germany and by Rudwick (1982) on the 19th century development of geology. Such disciplinary-bound classification remains significant to contemporary acts of knowledge-making and provides fertile grounds for sociological investigation.

Given that scientific controversy serves to reveal the act of knowledge production, disputes over classificatory events provide transparency to the nature of classification (Messeri, 2010; Steingart, 2012). This article examines the application and subsequent contestation of a classificatory scheme, which has recently received widespread political and public interest – the World Health Organization’s (WHO) Pandemic Alert Phases, as defined by the International Health Regulations (IHRs). After its application to the 2009/10 H1N1 Pandemic, the WHO’s 2009 revision of the Pandemic Alert Phases has been challenged as inadequate to the task of understanding and defining ‘pandemic’ disease threats (Flynn, 2010a, 2010b; Jefferson, 2010). These criticisms, which went to the heart of the WHO’s account of H1N1 and use of the Pandemic Alert Phases in their definitions of risk and preparatory action (Council of Europe, 2010; Flynn, 2010a, 2010b), demonstrate that for key public health actors (particularly in national governments), the application of this classificatory scheme was both unclear and inaccurate. In this way, from the viewpoint of the key stakeholders themselves, the WHO’s Pandemic Alert Phases were critically flawed. Furthermore, following the pandemic, the WHO itself has acknowledged some inadequacies with the Alert Phases and become concerned with constructing a more effective process (Chan, 2011; WHO, 2011a).

Using a range of documentary evidence, this article outlines the WHO’s construction of the Alert Phases and illustrates their crucial limitations when applied to the case of H1N1. This examination utilizes data gained from the textual analysis of epidemiological statements, policy documents, expert statements and public statements produced by the WHO during the period of March 2009 (the initial detection of the H1N1 virus) and August 2010 (the official declaration of movement into the Post-Pandemic Phase). These publicly available documents were chosen as part of a larger study, which sought to examine the WHO’s public construction and representation of H1N1 and its spread.

This article demonstrates that the WHO’s Phases reflected the tendency to construct classifications in a way that reaffirmed the WHO’s institutional memory (David, 1994; Douglas, 1989; Greenwald, 2008; Mahoney, 2000) and understandings of ‘exemplar’ pandemic events, particularly the Spanish Influenza Pandemic (Taubenberger and Morens, 2006). Following the tradition of Durkheim (1961) and Fleck (1979), it is argued that successful classification serves to stably reflect the ‘reality’ being classified in a way which is functional to the actors who use it. Thus, this case study demonstrates not only the social construction of the Pandemic Phases but also how such constructions fail when they cannot robustly serve their intended functions. When the requirement of basic functionality is not met, the classificatory scheme will necessarily be abandoned (Kuhn, 1970) and the actor–network’s endeavours to black-box the phenomenon will unravel (Callon, 1986; Latour, 1987, 1996).

This article illustrates the fragilities of the WHO Pandemic Phase classifications and the political difficulties faced by an institution that was responsible for issuing definitions surrounding a reality that could not, at the time, easily be classified. In particular, the WHO argued that the spread of H1N1 had crossed a boundary and become a (full, Phase 6) Pandemic, but this boundary was inherently unknowable and arbitrary. When the virus did not proceed to display attributes that the lay public and key public health actors understood as characteristic of a pandemic, the WHO classifications became a site of political and public contestation, and unravelled.

With its Pandemic Phases, the WHO attempted to create discrete categories through which to define socially and scientifically complex realities (the spread of infectious agents), creating boundaries that were by necessity artificial. In this case, the key question was ‘what constitutes a pandemic?’ For the Phases to be of any utility in circumscribing public health action, the WHO needed to provide a workable definition of the concept of ‘pandemic’ as well as to illustrate the measurable boundaries of such an event. However, the concept of ‘pandemic’ had not been previously well-defined. In application to the case of 2009 H1N1, which did not mirror the epidemiological and virological characteristics of the exemplar of Spanish Influenza, the inadequacies of WHO’s definition were highlighted. The liminal case of H1N1 disrupted the classificatory schema of the WHO’s Pandemic Alert Phases.

Constituting reality: the work of classificatory systems

Classification is intrinsic to thought and social structures (Foucault, 1970; Lewin, 1994; Moscovici, 1988). In theorizing or conceptualizing a novel phenomenon, there is a tendency to think in terms of analogy and comparison to phenomena that are already socially understood, recognized or defined (Douglas, 1969, 1973; Friese, 2010; Sontag, 1978). Within the sciences, this tendency to analogize manifests as formalized classificatory schema (Lewin, 1994). Such institutional classification is pivotal to medical work from diagnosis (Brown, 1995; Jutel, 2009) to the definition of disease categories, such as in the Diagnostic and Statistical Manual of Mental Disorders (DSM) and the International Classification of Diseases (ICD) manual (Bowker and Star, 1991, 1999; Gaines, 1992; Strand, 2011).

Defining health and illness, whether in an individual patient, or through a diagnostic category, is central to the discipline of medicine. Similarly, at the population health level, institutions such as the WHO that monitor and define public health problems, seek to construct robust epidemiological categories of investigation. These are most clearly mobilized when a ‘novel’ disease enters the diagnostic picture – such as occurred in the case of H1N1. Here, while the fundamentals of the disease were apparently discernible (novel Influenza A virus, rapidly spreading), the interpretation of what these characteristics meant was not. In the context of a contemporary interest in pandemic events, the questions of ‘if’ and ‘when’ H1N1 would constitute a pandemic were central to the public health action surrounding the virus.

Manifest functions of the Phases

One important characteristic of scientific classification is functionality – formal categories must provide some purpose in either sorting knowledge or defining action. The Pandemic Alert Phases serve as a signal of the pandemic potential of any circulating viral strain and are connected with the levels of preparation necessary to combat a potential influenza pandemic threat. The Phases are set to reflect the estimated probability of a pandemic, with Phase 6 indicating a pandemic in progress (WHO, 2009: 27). The WHO’s overall perspective on its definitions was that ‘[t]he phases are applicable to the entire world and provide a global framework to aid countries in pandemic preparedness and response planning’ (WHO, 2009: 24). Phase declarations therefore act as an important indicator of risk and a method through which to distribute key information and conceptualizations of pandemic threats from the WHO to its member states. States then react to this indicator by formulating management strategies.

The Phases undergo periodic revision by the WHO. They had been redefined immediately prior to the first recorded incidence of H1N1. This redefinition was outlined in the updated version of the WHO’s ‘core document’ regarding influenza management, the Pandemic Influenza Preparedness and Response (WHO, 2009) guidance document, produced by the Global Influenza Programme. Here, the WHO outlined the Phases, asserting that the 2009 definition ‘[r]etains the six-phase structure [from the earlier 2005 version] but regroups and re-defines the phases to more accurately reflect pandemic risk and the epidemiological situation based upon observable phenomenon’ (WHO, 2009: 3). The Phases thus serve to distinguish the degree of preparation needed in response to any influenza strain that threatens to develop into a pandemic. They classify what constitutes (and what does not constitute) a concern by outlining the factors that define a pandemic, or (in Phases 3 to 5) a potential pandemic. This includes identifying (potential) pandemic agents, describing the likelihood that these agents will cause a full (Phase 6) pandemic, defining a Phase 6 pandemic and demarcating appropriate public health actions towards these threats.

Sociological theorists who emphasize the functionality of classificatory schema (Douglas, 1989; Durkheim and Mauss, 1963 among others) provide a convincing argument for the functional role of categorization. Here, the importance of classification in circumscribing social roles and maintaining social boundaries and order is emphasized – classifications help social life function smoothly by defining roles and realities. In the context of scientific knowledge production, such as defining the concept of ‘pandemic’ and categorizing its ‘phases’, the act of classification is pivotal to understanding ‘natural’ phenomena. The functional role of the Pandemic Alert Phases also includes defining boundaries of authority and action among diverse global health actors, in addition to providing communication surrounding the nature of the virus (Brown et al., 2006; Fidler, 2001; Taylor, 2004).

Implicit functions of the Phases

In addition to these formally functional elements, classifications are also indicative of wider social realities. Constructionist approaches provide insight into these, suggesting that classifications produce the social ‘facts’ which they claim to elucidate. This emphasizes the often power-laden nature of classifications in the production of social reality (Haraway, 1991; MacKenzie and Wajcman, 1999; Treacher and Wright, 1982). For constructionist theorists, the success of a classificatory scheme lies in its correspondence to the thought collective which created it (Fleck, 1979; Foucault 1970), and the degree to which the construction can hold up against any challenge by external actors and events, or is applicable to locally applied contexts and needs. In the context of Western scientific theory, Kuhn (1970) and Fleck (1979), among others, demonstrate that classificatory knowledge is a product of the prevailing paradigm, discourse or disciplinary thought-style. Classificatory schemes which are thought to explain the world are actually a method of constituting it (Freeman and Frisina, 2010; Lewin, 1994; Vaihinger, 1949). In fact, shifts in scientific thought can be sociologically conceptualized as the outcome of the continuous revision of classificatory schema reflecting institutional or disciplinary perspectives (Fleck, 1979; Foucault, 1970; Martin and Lynch, 2009).

Once a categorization (here, the Phases) has been formed, any specific phenomenon in question (the H1N1 virus) is measured against the predefined criteria. The H1N1 virus was constituted as a pandemic agent at the point where it fit into the WHO’s category of ‘Phase 6 Pandemic’ – whereas, in the early Phases, it was constituted as merely a ‘potential’ pandemic – signifying an important category distinction. In the case of the WHO’s Pandemic Phases, the classificatory scheme is therefore not a purely theoretical constitution of reality, as in some scientific work, but also performative (Austin, 1980; Bowker and Star, 1999; Friese, 2010; Martin, 2004). At every Phase in which the virus was classified, national governments were expected to enact preparatory measures corresponding to the level of threat indicated. In this way, the WHO’s Phases provided a performative discourse and thereby constituted an important interface between meaning-making and action.

Once a classificatory scheme has been produced, it tends to acquire a taken-for-granted nature, where the categories are presumed to be ‘natural’ or ‘true’ (Douglas, 1969, 1973; Foucault, 1970). Although intended primarily as a description and signal, the Phases also implicitly predict and explain the nature of pandemics. In attempting to describe the boundary between ‘potential’ pandemic and ‘pandemic-in-progress’, the Phases necessarily serve to conceptualize and constitute what a ‘real’ pandemic is. This act of definition, therefore, foreshadows and fosters an understanding of what constitutes effective action against a pandemic.

Thus, a key function of classifications, particularly in the context of risk, is to erase uncertainty (Derksen, 2000; Martin and Lynch, 2009). Translating a phenomenon into a classificatory category constitutes it in accordance with the category’s defining characteristics. In this case, the uncertainty to be abrogated was the question surrounding how the virus should be managed. Classifying H1N1 as a Phase 6 Pandemic constituted it as a (risky) pandemic disease. Thus, when effectively employed, classifications eliminate ambiguity by definitively placing a phenomenon among similar things (Latour, 2005). This act of erasing uncertainty is central to scientific practice (Martin, 2004). To be constituted as a scientific ‘fact’, an idea/phenomenon needs to exhibit defined and constant characteristics; any uncertainty surrounding the classification must become covered/removed so as to maintain the solidity of the construction (Fleck, 1979).

As the discussion below demonstrates, neither the manifest purpose of the Phases (to communicate risk to national governments and order public health action) nor the social functions of the Phases (to constitute the pandemic event and to define the nature of the risk posed) was fulfilled, ultimately leading to the contestation of the Phases as a classificatory tool.

The dysfunction of the Pandemic Alert Phases

The WHO Pandemic Alert Phases, as illustrated in the ‘Pandemic Influenza Guidance and Response’ document of 2009, outline the six-phase progress of a pandemic event. The Phases had been formulated by the WHO as part of the procedures set out by the IHRs (Review Committee on the Functioning of the International Health Regulations (2005) in Relation to Pandemic (H1N1) 2009 (2011)) to allow for efficient identification and notification of problems, facilitation of global communication and coordination of the global response against infectious disease events. The determination of the Pandemic Phases is an integral part of the IHRs, allowing WHO member states to effectively identify risk and act against potential pandemic threats. Phases 1–3 are used to denote emerging problems and do not signify any need for action on the part of national governments. Phase 1 represents a base state, where there is no identified animal influenza virus which causes infection in humans. Phase 2 indicates a state of known zoonotic infection of humans in isolated events. Phase 3 refers to a state where although there is an increased incidence of infection, there is no sustained human-to-human transmission sufficient to demand community-level outbreaks. This is followed by Phases where the event is considered significant enough to warrant action. Phase 4 is adopted where human-to-human transmission is verified, resulting in the enactment of rapid containment measures and a state of readiness for a pandemic response. Phase 5 represents community-level outbreaks in at least two countries within one geographical region. At this stage, as a precautionary measure, national governments are advised to begin implementing pandemic response plans. The final stage represents a definitional shift from a potential pandemic to a pandemic-in-progress. This is Phase 6, which is designated when there is sustained spread of the virus over multiple geographical regions. This is then followed, at the conclusion of the Pandemic, by the Post-Pandemic Period.

The WHO Pandemic Phases were based primarily upon the key criteria of geographical spread and mode and mechanism of transmission (WHO, 2009). One of the commonly recognized characteristics of pandemic events is the ability of a novel virus to transmit swiftly between human hosts. Human-to-human transmission, signifying the fulfilment of the key mutation from animal-to-human transmission (since pandemic agents are invariably zoonotic transfers), represents an important feature in recognizing a specific viral agent as potentially pandemic-causing (Webster, 2002). In the case of H1N1, such human-to-human transmission was evident from early in Mexico’s reporting of the disease and the WHO’s first release of statements on the (then-labelled) swine flu on the 24th April 2009. From the first invocation of the Pandemic Phases by the WHO, H1N1 was already categorized as a Phase 3 threat, which specifies the presence of a sporadic outbreak but no known human-to-human transmission. As Figure 1 attests, this very quickly progressed to declaration of a Phase 5 threat.

Figure 1.

H1N1 geographic spread over countries and territories.

As Figure 1 demonstrates, the H1N1 virus spread rapidly over multiple geographic regions. In strictly applying the classificatory system of the Pandemic Alert Phases, the WHO would have been forced to declare a Phase 6 Pandemic very soon after its initial detection if it were following its own Phase definitions. The WHO, however, would not do this, since the Phase 6 declaration implicitly and socially signified a high level of risk and threat. In this way, the criteria of the official Phases were quickly set aside in favour of Phase declarations more informally based upon the WHO’s perception of risk. As discussed below, the disparity between risk and geographical spread (and thereby the formal Phases) then became highly evident. Consequently, the application of H1N1 in relation to the Pandemic Phases highlighted their definitional inadequacies, leading to the contestation of both the WHO’s characterization of H1N1 as a ‘pandemic’ and the Pandemic Alert Phases themselves.

The criticism of the WHO’s Pandemic Phases by other public health actors revolved around two questions, which arose at different temporal points in the event. During the later stages of H1N1’s spread, questions revolved around whether a pandemic declaration had been necessary and when a Post-Pandemic Period would be announced. However, during the early pre-declaration period, critics in fact argued that Phase increases were occurring too slowly. While these two criticisms might seem contradictory, they can be explained by the imprecise nature of the Phase categories themselves, which rendered those categories liable to diverse interpretations and diminished their capacity to provide discrete and functional categories. A third key point of criticism that the WHO’s definition of ‘pandemic’ did not effectively articulate the boundary between a potential threat and a pandemic in progress was also bound up in the inadequacies of the Phases and signified the ill-defined nature of the concept of ‘pandemic’ itself.

Knowing when a pandemic has started

During the early stages of the threat, most of the critical discussion of the WHO’s actions centred around its timing of the Phase declarations. From the earliest discussions, the WHO representatives offered suggestions that H1N1 was likely to result in a pandemic, but the timing of the Phase 6 declaration was a subject of contention. For example, as early as 26 April 2009, it was suggested by Dr Keiji Fukuda (2009a) (Special Advisor to the WHO Director-General on Pandemic Influenza and WHO Assistant-Director General) that

… the swine viruses appears to be affecting [a] significant [number] of people in at least a couple of different countries in different locations. This situation has raised questions about whether we are entering into a pandemic period.

A little over a week later, Dr Michael Ryan (2009) (WHO Director of Global Alert and Response) argued that

at the present time I would still propose that a pandemic is imminent because we are seeing the disease spread to other countries. We have not seen yet that sustained transmission outside one WHO region. At this point we have to expect that Phase 6 will be reached. We have to hope that it is not reached.

From the initial development of the threat, there was a clear expectation within the WHO that H1N1 was likely to constitute a Phase 6 Pandemic. Nevertheless, at the time, despite the evidence of geographical spread (upon which the Phases are defined) and a presumption that a pandemic was imminent, movement through the Phases was slower than the Phase guidelines themselves would dictate. The swift abandonment of the official guidelines provides evidence that the Phases were inadequate to their task. The WHO was criticized because the epidemiological situation (in relation to spread of disease) had fulfilled the definitions of each Phase well before the Phase change was implemented. This was a result of the conflict between the function of the Phases (to recommend action and measure risk) and their definition (which relied upon geographical spread). The wide spread of the virus did not necessarily signify that the disease would manifest as severe, although this link was apparently assumed within the Pandemic Phases since there was no other measure of threat built into the classification.

Press questions (and the WHO representatives’ answers) about the hesitancy to declare Phase increases illuminated the WHO’s attempts to reconcile this discrepancy, and its inability to effectively do so within the confines of the existing classificatory scheme. For example, one reporter questioned,

You said yesterday the Emergency Committee wanted to buy more time, but to many people here in the real world it looks very much like [this] does not fit the definition of Phase 3, and I am hearing from people in the infection control sector that this is really making life difficult for them. They don’t know if they are going to 4, 5, 6, sometime in the next few days … Should they be operating under the assumption that this is probably a pandemic? (Fukuda, 2009a)

Here, the inefficacy of the Phases was already becoming apparent. Fukuda (2009a) suggested that the Phases were not being declared, although they were technically applicable, because of the implications of such declarations:

… whenever we face any emergency situation there is of course a balance between the need to have a fair amount of information so that we feel that our decisions and our assessments are based on solid grounds. On the other hand, I think we are also mindful of the need for different groups to act and to make decisions. Now, if WHO goes ahead and makes a declaration that the phase has changed, then this is really a very serious signal to the world …

This exchange suggests that, in part, the (delay in) Phase changes were motivated by the WHO’s desire not to send ‘strong signals’ (which might instigate panic, and signify high risk) until the decision could be well justified through additional epidemiological evidence. Thus, early on in the event, the WHO was already defensive and was being heavily scrutinized on the matter of Phases because the progression of H1N1 did not conform to the expected progression of a pandemic (both in terms of temporal events and in regard to risk), as outlined by the Phase definitions.

The WHO suggested that it did not wish to declare a pandemic or increase Phases before member countries could be informed about what such statements entailed or before risk was more clearly defined. As an example, it was asserted by the WHO representatives that

Another question that has come up is that ‘Are we in Phase 6 [full pandemic] now, or why haven’t we declared Phased 6 now?’ … I simply want to say that we know that this virus is spreading and we are now seeing that activity is picking up in a number of countries and … we know that we’re getting closer to probably a pandemic situation. … I want you to know that WHO has been working extremely hard in terms of preparing countries, in terms of preparing populations for what a potential move to Phase 6 or pandemic would entail.

For example, there is work that is being undertaken right now so people understand really what does a pandemic mean, what does going to Phase 6 mean? Does this mean we are seeing something really severe change? Does this mean that there is a need for drastic actions to be taken? (Fukuda, 2009e)

Although, according to the classificatory criteria, a pandemic state already technically existed, the declaration was withheld until expectations of events could be better managed and until the WHO could more successfully define the case and position the global reaction. This represents a closely related failure in the classificatory scheme, a failure to serve its performative functions – WHO information was not translatable into public health action. Thus, while the Phases sought to inform member states of effective actions, the WHO thought that the technical application of the Phases would potentially produce confusion surrounding public health action.

The failure of the Phases as a classificatory scheme was reinforced in an exchange where a journalist questions why the pandemic has not been declared, since, by the WHO’s own definitions, the situation appears to warrant it. In this exchange, David Brown (journalist, Washington Post) asks,

… if you could please address the question of why there seems to be so much reluctance on going to Phase 6? It is a very clear definition. The point was made, you know, long ago, that it does not measure severity. What is to be lost by saying that it is community spreading, in the community and more than one place – which it obviously is – more than one region, we are going to go to Phase 6 and it is a mild Phase 6. Why not just bite the bullet? (Fukuda, 2009d)

In response, it was asserted that

The answer to that is really almost another question which is: ‘what is to be gained by going to another Phase?’ … Right now, when we look at the request: ‘Why cannot WHO look at going to Phase 6’ coming from the countries, there are a couple of concerns here. One of them is that in many of the countries they do not see H1 activity going on, and in these countries with the few cases, things are relatively mild. And so, behind that question is the sense that many countries are already doing things that are necessary right now to address the situation. But if you go and declare Phase 6 without very clear evidence that there is a sort of change in the global situation, it can lead to extra work for countries without much gain, it can lead to some level of panic, it can lead to some level of cynicism … (Fukuda, 2009d)

The WHO was attempting to minimize potential future criticisms of prematurely/unnecessarily calling a pandemic and was concerned with managing expectations. It was put in this position because the Phase classifications were at odds with the general social understanding of a ‘pandemic’ as a devastating and severe event (see below). As such, the WHO needed to actively manage the member states’ expectations in spite of the Phases, rather than through the Phases.

Distinguishing pandemic events

Another key site of criticism revolved around the definitional abilities of the Pandemic Phases. In order to be defined as a ‘pandemic’, an influenza virus strain must be distinguished from seasonal influenza and shown to be analogous to previous pandemic events. In general, drawing upon the collective memory of past contagion is fundamental to the construction of a new disease (Halbwachs, 1992; Herzlich and Pierret, 1987). This is because the characterization of any new threat necessarily reflects pre-existing conceptualizations of infectious disease (Fleck, 1979), and one of the ways in which novel disease events are understood is through reference to an existing comparative framework (Marková and Farr, 1995). Since such analogy construction constitutes an important device through which thoughts and ideas are represented and analysed (Arber, 1954; Sontag, 1978), the translations of the H1N1 pandemic actor–network necessarily involved the assembly of these links (Callon, 1986; Prout, 1996).

The WHO’s constitution of H1N1 reflected the WHO’s previous historical experience with infectious disease control (Campbell, 2010; David, 1994; Mahoney, 2000). From their opening press conference, the WHO introduced the H1N1 virus by reference to previous infectious disease threats. Reference to past pandemics also served to reinforce the unpredictability of pandemic influenza, suggesting that the world needed to be vigilant to the volatility of H1N1’s pandemic potential. Thus, as Dr Margaret Chan (2009) argued, ‘… experience of past pandemics warns us that the initial situation can change in many ways, with many, many surprises’. These types of analogies reinforced and justified concern over H1N1, despite what at the time was the mild epidemiological presentation of illness (when it was as yet uncertain how the pandemic would unfold). Here, past pandemics were specifically invoked to reinforce the potential impact of the virus, as illustrated in the following example from the WHO Director-General:

… the 1918 pandemic, the most deadly of them all, began in a mild wave and then returned in a far more deadly one. In fact, the first wave was so mild that its significance as a warning signal was missed.

… the pandemic of 1957 began with a mild phase followed, in several countries, by a second wave of greater fatality. The pandemic of 1968 remained, in most countries, comparatively mild in both its first and second waves.

At this point, we have no indication that we are facing a situation similar to that seen in 1918. As I must stress repeatedly, this situation can change, not because we are overestimating or underestimating the situation, but simply because influenza viruses are constantly changing in unpredictable ways. (Chan, 2009)

The WHO attempted to reinforce the argument that there was a lack of initial severity in the major past pandemics, suggesting that H1N1 could mimic these events and eventually manifest as a severe disease. Furthermore, such assertions served to construct the characteristic of unpredictability; H1N1 was represented as threatening and risky because its impact was unknown.

The Spanish Influenza serves as the prototypical example of severe and unpredictable pandemic. The example of the Spanish Influenza was an important framing device in translating the threat of H1N1. For example, it was again suggested that

… the worst pandemic at the last century started out mild in the springtime, it was fairly quiet during the summer, and then in the autumn when it really exploded, this is in 1918 and it was a much more severe form. (Fukuda, 2009b)

Such assertions emphasize the need for constant vigilance and the potential threat of the current situation. The analogy with Spanish Influenza was necessary for the translation of H1N1 into the category of ‘pandemic’ due to the prominence of Spanish Influenza within the collective understanding of pandemics. The Spanish Influenza pandemic of 1918–1919 is held up as a prototypical example of an influenza pandemic, despite its epidemiological uniqueness (Taubenberger and Morens, 2006; Tognotti, 2003), and has been reinforced as a fearful event in public memory (Barry, 2004; Crosby, 1976). In order to constitute and categorize H1N1 as a legitimate pandemic threat, the WHO constructed and reinforced links between Spanish Influenza and H1N1. Furthermore, the Pandemic Phases implicitly reflected the public experience and collective memory of the Spanish Influenza Pandemic (Barry, 2004; Crosby, 1976; Halbwachs, 1992; Olick and Robbins, 1998) that swift spread would correspond to severe disease risk. It was assumed that novel influenza agents would necessarily result in a severe event.

In addition to historical analogy, the WHO also referenced concurrent disease events. Linking H1N1 to past threats served to locate it in the context of a continuing history of disease. Drawing analogy to past pandemics also suggested that this experience ‘helps us to understand the situation, right now’ (Chan, 2009); it was argued that H1N1 can be recognized and understood through this accumulated knowledge. Furthermore, as a result of past pandemics, there is increased preparation for pandemic events so that ‘… the world today is much more alert to such warning signals [at the appearance of new strains] and much better prepared to respond’ (Chan, 2009).

In order to effectively problematize H1N1 as a current threat, the WHO needed to situate the emerging event of H1N1 within debates about contemporary infectious disease. References to more recent threats served a different purpose to historical analogy. Such allusions represent attempts to translate H1N1 as a significant phenomenon within contemporary global disease events. Thus, in one of the opening statements of the first conference, it was suggested,

Many of you know that the world has been talking about and preparing for pandemic influenza for at least the past five years and there are a number of reasons for this. We know that influenza pandemics have occurred at least a couple of times each century and in the last five years we have been working very hard … because of a specific pandemic threat known as avian influenza or H5N1 and because of that many countries have been very focused on strengthening their defences for such a situation. (Fukuda, 2009a)

In this statement, it was asserted that influenza pandemics are an ever-present risk for which we should prepare (and have prepared).

Ultimately, these attempts to distinguish the category of pandemic proved ineffectual with respect to H1N1. In addition to not fulfilling the function of effectively providing certainty to risk, the Pandemic Phases, as applied to H1N1, also failed to effectively distinguish the Phase 6 pandemic event from other (epidemic and seasonal) infectious disease events. As early as 5 May 2009, press questions focused upon the lack of distinction between H1N1 and seasonal influenza and the (in)applicability of the historical and contemporary pandemic analogies. The WHO responses to these questions were at times ambiguous and failed to provide a clear distinction. For example, in reply to one question which suggested that seasonal flu deaths were in fact potentially large (minimizing the distinction with H1N1), it was stated,

In fact the numbers we have for seasonal flu vary depending on the years. Some years we have a very mild seasonal flu, and other years we have a more severe seasonal flu. Global figures are really difficult to get because each country is monitoring the seasonal flu, and they provide their figures, but not necessarily on a regular basis. But to give you a kind of frame, in France for example, the number of deaths during seasonal flu varies from 5,000-15,000 deaths, in the United States you can have 40,000 deaths depending on the years, so these are numbers, but highly variable. (Briand, 2009)

It could be argued that the WHO needed to maintain the perception of the impact of seasonal influenza (even within discussions of pandemic strains), given that, on the global scale, seasonal strains do in fact represent a significant health burden and remain an important aspect of the WHO’s non-crisis health governance (WHO, 2011a, 2011b). However, in the context of the discussion of H1N1, the failure to downplay seasonal influenza, or to establish a strong distinction between seasonal and pandemic cases, constituted a point at which the WHO’s problematization of H1N1 became vulnerable. As the above-mentioned quote shows, in order to make distinctions, the representative appears to have implied that mortality from seasonal flu can vary, whereas mortality from pandemic influenza is always high. However, this does not represent a clear-cut marker of difference and fails to construct H1N1 as a distinct event. This was particularly apparent when, by the WHO’s official count, H1N1 only resulted in 18,500 laboratory-confirmed deaths by the end of the pandemic period (WHO, 11/08/10).

The WHO’s explanation of the difference between H1N1 and seasonal influenza was unconvincing. Direct comparisons between the two states of influenza (pandemic and seasonal) were made in a way that failed to establish a distinction. For example, in one conference, it was suggested that with respect to H1N1

In terms of the illness itself, in the people who are developing generally milder illness, this is similar to the kinds of influenza-like illnesses that we see, so this is typically people developing fever, cough, body aches, headaches, and this is generally in keeping with what the milder spectrum of illness is. (Fukuda, 2009c)

Thus, the WHO failed to effectively establish strong distinctions between H1N1 and seasonal influenza, or strong links between H1N1 and past pandemic agents, within definitions of pandemic. The distinction (from seasonal influenza) and the similarities (with pandemic forms) were assumed and implicit within the Phase construction but were not explicitly evident in the characteristics that the Phase categories measured, representing a key dysfunction of the Phase criteria.

Knowing when a pandemic has finished

The inadequacy of the Phase classifications was also clearly evident in the later stages of H1N1. Almost immediately following the full Phase 6 declaration on 11 June 2009, critics began to suggest that the WHO should declare the end of the pandemic (Flynn, 2010a). Questions centred particularly upon the suggestion that the decreasing spread, and therefore caseload, of infection indicated that the pandemic threat had passed. For the WHO, declaring the end of pandemic was a difficult process for a number of reasons. In the case of H1N1, because it had never been a severe pandemic, a clear ‘end’ was indeterminable. The Pandemic Alert Phases themselves were ill-defined in their categorization of exactly what a Post-Pandemic Period entailed. Eventually, following an extended period of ambiguity, the Post-Pandemic Phase was declared by Chan on 10th August 2010, when she announced that ‘[w]e are now moving into the Post-Pandemic Period. The new H1N1 virus has largely run its course’ (Chan, 2010).

Prior to the declaration of the Post-Pandemic Period, the WHO needed to justify its continued assertions of ‘pandemic’ (Phase 6) classification of H1N1. This was done through reference to the historical experience of pandemics, where it was suggested that history should provide a guide for declaring the end of the pandemic:

… in terms of how we move from a pandemic period to a non-pandemic period. Again if we look back at history for some guidance, we will see that we typically have a period in which pandemic infections are quite high. Then we go to a transition period in which those newly emerged viruses, pandemic viruses, often become seasonal influenza viruses. (Fukuda, 2009f)

Consequently, the question for the WHO became one of distinguishing the point of entry into a transition period when the pandemic period was finished and a non-pandemic period began. Here, it was pointed out that ‘… the ending of the pandemic is not an on and off phenomenon, we really expect it to be more of a trailing off phenomenon, it does not happen overnight’ (Fukuda, 2010b). Although it portrayed pandemic events as clearly defined, with a distinct endpoint, the classificatory scheme did not offer a clear distinction between pandemic and non-pandemic states. As a result, the WHO proposed that an ambiguous transition period was necessary including

… the post-peak period which is the transition period as well as the post-pandemic period which signifies when we have quite a good expectation that we are really getting close to the normal period out of the pandemic period. (Fukuda, 2010b)

This occurred where

The practical effect of indicating that we are in a post-peak period is really to give a broad signal to the world that even though we may continue to see pandemic activity that we expect that we are transitioning more towards a normal level. (Fukuda, 2010b)

However, although it was acknowledged that such a transition period was necessary, it was difficult for the WHO to effectively distinguish a point at which the end-of-pandemic could be declared for H1N1, due to both their ineffective classifications and the mild manifestation of the disease.

For example, in attempting to reinforce the distinction with seasonal influenza, Chan (2010) suggested that in the Post-Pandemic Period,

… we notice that in countries with H1N1 transmission, the level of intensity is now moving back to a pattern similar to the seasonal influenza pattern. The third thing we observed in all these countries that we have been getting good data, there is no longer a dominance of the H1N1 virus as we saw last year. We are seeing a mixed virus pattern. By that we mean we see H1N1 virus; [but] we also see H3N2 and we also see Influenza B virus …

But, last … is that we are seeing some level of community-wide immunity, either due to natural infection by the H1N1 or due to passive immunity by vaccination.

However, even in the Post-Pandemic Period, the disease was characterized by the WHO as exhibiting some of the same features as in the Pandemic Phase, due to the ill-defined nature of the categories and the lack of distinction between H1N1 and seasonal strains. The apparently ‘defining’ qualities of a pandemic were thereby rendered ambiguous in the accounts of a Post-Pandemic Period, given that they were apparently present in both Pandemic and Post-Pandemic contexts.

Furthermore, during Phase 6, WHO representatives noted that public and media calls for the declaration of the end of the pandemic did not take into account the global nature of its spread. In this way, the continued impact of H1N1 in specific regional areas served as a justification for the continued pandemic classification:

Based on the situation, our current assessment is that it remains too early to say that the pandemic is over. This is because we continue to see activity at elevated levels in a number of countries. (Fukuda, 2010a)

Thus, Fukuda (2010c) argued that ‘… pandemic activity is different at different places in the world’, it was ‘really too early to conclude that the pandemic was in a post-peak period in many countries’, and for these reasons, a post-peak period could not be declared at that time. Therefore, Fukuda (2010c) continued,

I think that, if we look at how the world deals with these large global events … some of the recommendations made at the global level certainly are blunt because they are really intended to be relevant and germane to the world.

The global attribute of an influenza pandemic is highlighted in these justifications. However, classificatory schemes often encounter problems when they attempt to consolidate localized and globalized problems into a uniform set of categorizations. There is often a disconnection between the locally experienced reality and the simplified global categorization (Bowker and Star, 1991; Mahajan, 2008). The WHO institutionally focused upon a global problem, whereas the member states experienced only national effects. To the extent that the Phases served as signals for action on the part of member states, this difficultly in consolidating the local and the global into one schema may have contributed the overall inadequacy of the Pandemic Phases.

Upon the WHO’s declaration of the end of the pandemic, Frank Jordans (journalist, Associated Press) commented that ‘[s]everal countries started scaling back their H1N1 efforts some months ago, yet WHO held back on downgrading the pandemic phase until now. Why did it take so long?’ (Chan, 2010). Dr Margaret Chan (2010) responded that

Yes, indeed, what you said is correct. Many countries in the northern hemisphere in fact scaled back on their public health response to the H1N1 virus. This is the right action to be taken because for countries, especially in the temperate zone in the northern hemisphere, the worst was over. But having said that, the World Health Organization has a duty to monitor the global situation and that is precisely what we are doing …

Now all in all, we are seeing clear signals and evidence pointing to the fact that the world is now – and I’m talking about it at a global level – the world is transitioning out of the pandemic into the post-pandemic period.

Similarly, the reporter Jules Caron asked, ‘[I] would like to know, between phase 6 and post-pandemic, what exactly does it mean? What is the WHO doing now that it does, didn’t do before?’ (Fukuda, 2010d). To this Fukuda (2010d) responded that

… this action simply notifies countries that we are transitioning out of a pandemic period in which we have seen unusual patterns related to influenza, back to a period in which we see influenza patterns more typical of seasonal influenza. However, during this period one of this things which we are strongly emphasising to countries is that it’s important to continue monitoring and (stay) alert …

So one action is to continue with surveillance. A second action that we are recommending is that it is important to continue with control efforts …

Here, it is evident that there was little distinction between actions occurring in the Pandemic Phase and actions occurring in the Post-Pandemic Phase – surveillance and control precautions were still emphasized and member states had enacted changes long before the official declaration.

The WHO’s narrative surrounding Pandemic Phases failed to effectively distinguish the concept of ‘pandemic’ and convey a sense of a risk. In terms of defining the end of pandemic, the WHO did not effectively distinguish the Post-Pandemic from the Pandemic Periods – the classification was not precise. This rendered the notion of pandemic, and the Phases, susceptible to contestation. The WHO’s attempts at retrospective definition merely highlighted the inadequacy of the initial Phase definitions in categorizing pandemic disease.

Unknowable boundaries: the problem with classifying pandemics

The 2009 version of the WHO’s Pandemic Alert Phases were constructed from the perspective of that institution’s historical experience with disease and the public memory surrounding the Spanish Influenza Pandemic. The wide spread of novel infectious agents generally led to outbreaks of severe disease; thus, the ‘severity’ as a characteristic was taken for granted. This led to an emphasis on geographical spread (along with the less problematic characteristics of novelty and sustained human-to-human transmission) as an important indicator of a pandemic. However, the inadequacies in this classificatory scheme became apparent to many actors when it was applied to H1N1 in 2009–2010, leading to a range of criticisms of the WHO’s actions.

The H1N1 virus represented a liminal case in the experience of novel influenza. The virus spread swiftly, through human-to-human transmission, across the globe. In this way, soon after its detection, the H1N1 virus fulfilled the criteria for the declaration of an official Phase 6 Pandemic. However, H1N1 did not fulfil the implicitly assumed characteristic that a pandemic would cause severe disease (born from the experience of the Spanish Influenza Pandemic) since the effect of the virus was different from the previous experience of such events. This fact uncovered fundamental flaws in the WHO’s construction of the Pandemic Phases, where more readily measurable criteria (spread and mode and mechanism of transmission) were emphasized, while other criteria (severity) were neglected. Given that both the manifest and implicit tasks of the Phases revolved around the concept of risk – in helping member states evaluate and manage the risk posed by the virus, and in producing an aura of certainty around the uncertain risk – this was a fundamental flaw. A high-risk event – one which would result in a higher burden on public health institutions – was not distinguished from the spread of mild disease. As part of this, several key classificatory distinctions were also missing: a distinction between pandemic and seasonal influenza and the definitions of the start and end of pandemic events. This meant that the links between Phase changes, changes in the status of risk posed, and therefore necessary public health actions, were not adequately produced.

While it can be argued that the case of H1N1 reflected the WHO’s choice of inadequate criteria in its Phase definitions, the case also raises broader questions about the application of formal scientific classificatory schemes. Epidemiological classifications construct discrete ontological categories within which to place disease events. While liminal cases can always act to contradict and dissolve the black boxes surrounding medical categories, the construction of a classificatory scheme to distinguish between (potential) pandemic states is particularly fraught with difficulty. Pandemic events are rare, meaning that any classification will be based on a limited experience of occurrences. Furthermore, as the WHO’s own statements indicate, historically, there has been a wide variation in the impact of diseases that have all been labelled ‘pandemics’. Previous influenza pandemics have produced deaths ranging from approximately 33,800 in the United States and 30,000 in England and Wales, for the least severe – Hong Kong Influenza, H3N2, 1968–1969 – to 50 million globally for the most severe – Spanish Influenza, 1918–1919 (Cox and Subbarao, 2000; Nguyen-Van-Tam and Hampson, 2003; Taubenberger and Morens, 2006). In comparison, the H1N1 Pandemic was relatively mild. However, the small number of cases upon which to base classification makes the formal definition of a ‘pandemic’ and its ‘phases’ difficult. The production of a set of discrete categories in order to sort phenomena that are both rare and diverse in nature is therefore potentially problematic given the limited evidence upon which classifications can be based. The WHO released Phase declarations, which were supposedly indicators of levels of risk and preparedness, based upon the idea that H1N1 had crossed between these (inherently indistinct) category boundaries.

With respect to the Pandemic Phases, it is clear that the construction of this classificatory scheme reflects social realities, such as the historical experience with infectious disease and institutional assumptions and blind spots, in designating what constitutes a pandemic. Nonetheless, the fact that the Pandemic Phases failed to fulfil their roles is equally significant. The Phases were held open to contestation not because they are constructed categories (as they inevitably would be), but because they failed to fulfil key functions when applied to H1N1. The categories were supposed to provide information surrounding risk, but emphasized geographical spread as the key epidemiological element, and they were supposed to produce discrete categories of phenomena, although (potential) pandemic events vary widely. This supports the argument that classifications need to be functionally stable in order to be sustainable.

Following the contestation and criticism of the Pandemic Alert Phases in their application to the case of H1N1, the WHO announced that it will revisit the way in which it handled H1N1 and applied the Phases from the IHRs (Review Committee on the Functioning of the International Health Regulations (2005) in Relation to Pandemic (H1N1) 2009 (2011). To fulfil their function more properly, the classifications would need to include a capacity to measure risk within any reinvention. However, the difficulty for the WHO in successfully reformulating the Phases lies in the nature of both classificatory schemes and pandemics themselves. Scientific classifications rely upon the ability to construct discrete, measurable and definable categories into which to sort the realities of the natural world. For the WHO, the rarity, variability and limited evidence surrounding pandemic events renders this process fraught with difficulty.

Footnotes

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Author biography

Sudeepa Abeysinghe is a Research Fellow at the London School of Hygiene and Tropical Medicine, having completed her PhD in Sociology at the Australian National University. Her research interests include the interface between knowledge production and policy production, risk and scientific uncertainty and the sociology of institutions.

References

Arber

(1954) The Mind and the Eye: A Study of the Biologist’s Standpoint. Cambridge: Cambridge University Press.

Austin

(1980) How to Do Things with Words. London: Oxford University Press.

Barry

(2004) The Great Influenza: The Epic Story of the Deadliest Plague in History. London: Penguin.

Bloor

(1982a) Durkheim and Mauss revisited: Classification and the sociology of knowledge. Studies in History and Philosophy of Science Part A 13(4): 267–297.

Bloor

(1982b) Polyhedra and the abominations of Leviticus: Cognitive styles in mathematics. In: Douglas

(ed.) Essays in the Sociology of Perception. London: Routledge & Kegan Paul, pp. 191–218.

Bowker

Star

(1991) Situations vs. standards in long-term, wide-scale decision-making: The case of the international classification of diseases. In: Proceedings of the twenty-fourth annual Hawaii international conference on social systems, Kauai, Hawaii, pp. 73–81.

Bowker

Star

(1999) Sorting Things Out: Classification and Its Consequences. Cambridge, MA: Massachusetts Institute of Technology.

Briand

(2009) WHO press briefing. 8 May. Available at: http://www.who.int/mediacentre/multimedia/swineflupressbriefings/en/index.html (accessed 14 August 2010).

Brown

(1995) Naming and framing: The social construction of diagnosis and illness. Journal of Health and Social Behavior 35: 34–52.

10.

Brown

Cueto

Fee

(2006) The World Health Organization and the transition from ‘international’ to ‘global’ public health. American Journal of Public Health 96(1): 62–72.

11.

Callon

(1986) Some elements of a sociology of translation: Domestication of the scallops and the fishermen of St Brieuc Bay. In: Law

(ed.) Power, Action and Belief: A New Sociology of Knowledge? London: Routledge, pp. 196–223.

12.

Campbell

(2010) Institutional reproduction and change. In: Morgan

Campbell

Crouch

Pedersen

Whitley

(eds) The Oxford Handbook of Comparative Institutional Analysis. Oxford: Oxford University Press, pp. 87–116.

13.

Chan

(2009) H1N1 influenza situation: statement made at the Secretary-General’s briefing to the United Nations General Assembly on the H1N1 influenza situation. Available at: http://www.who.int/dg/speeches/2009/influenza_a_h1n1_situation_20090504/en/index.html (accessed 14 June 2010).

14.

Chan

(2010) H1N1 in post-pandemic period: Director-General’s opening statement at virtual press conference. Available at: http://www.who.int/mediacentre/news/statements/2010/h1n1_vpc_20100810/en/index.html (accessed 29 September 2010).

15.

Chan

(2011) Implementation of the International Health Regulations (2005): Report of the Review Committee on the Functioning of the International Health Regulations (2005) in relation to Pandemic (H1N1) 2009. Report by the Director-General. Report made to the 64th World Health Assembly. Available at: http://apps.who.int/gb/ebwha/pdf_files/WHA64/A64_10-en.pdf (accessed 20 July 2011).

16.

Council of Europe (2010) Faked Pandemics – A threat for health. Motion of a Recommendation by the Parliamentary Assembly of the Council of Europe. Vol. Doc. 122110, 11 January. Strasbourg: Council of Europe.

17.

Cox

Subbarao

(2000) Global epidemiology of influenza: Past and present. Annual Review of Medicine 51(1): 407–421.

18.

Crosby

(1976) Epidemics and Peace, 1918. Westport, CT: Greenwood Press.

19.

David

(1994) Why are institutions the ‘carriers of history’? Path dependence and the evolution of conventions, organizations and institutions. Structural Change and Economic Dynamics 5(2): 205–220.

20.

Derksen

(2000) Towards a sociology of measurement: The meaning of measurement error in the case of DNA profiling. Social Studies of Science 30(6): 803–845.

21.

Douglas

(1969) Purity and Danger: An Analysis of Concepts of Pollution and Taboo. London: Routledge & Kegan Paul.

22.

Douglas

(1973) Natural Symbols: Explorations in Cosmology. London: Barrie and Jenkins.

23.

Douglas

(1989) How Institutions Think. Syracuse, NY: Syracuse University Press.

24.

Durkheim

(1961) The Elementary Forms of Religious Life. New York: Free Press.

25.

Durkheim

Mauss

(1963) Primitive Classification (trans. Needham

). Chicago, IL: University of Chicago Press.

26.

Fidler

(2001) The globalization of public health: The first 100 years of international health diplomacy. Bulletin of the World Health Organization 79(9): 842–849.

27.

Fleck

(1979) Genesis and Development of a Scientific Fact (trans. F Bradley and TJ Trenn). Chicago, IL: University of Chicago.

28.

Flynn

(2010a) The handling of the H1N1 pandemic: More transparency needed. Doc no. 12283 – Passed by the Council of Europe, 24 June.

29.

Flynn

(2010b) Memorandum: The handling of the H1N1 pandemic: More transparency needed. Available at: http://www.assembly.coe.int/CommitteeDocs/2010/20122329_MemorandumPandemic_E.htm (accessed 20 July 2011).

30.

Foucault

(1970) The Order of Things: An Archaeology of the Human Sciences. New York: Routledge.

31.

Freeman

Frisina

(2010) Health care systems and the problem of classification. Journal of Comparative Policy Analysis 12(1): 163–178.

32.

Friese

(2010) Classification conundrums: Categorizing chimeras and enacting species preservation. Theory and Society 39(2): 145–172.

33.

Fukuda

(2009a) WHO press briefing. 26 April. Available at: http://www.who.int/mediacentre/multimedia/swineflupressbriefings/en/index.html (accessed 14 June 2010).

34.

Fukuda

(2009b) WHO press briefing. 29 April. Available at: http://www.who.int/mediacentre/multimedia/swineflupressbriefings/en/index.html (accessed 14 June 2010).

35.

Fukuda

(2009c) WHO press briefing. 5 May. Available at: http://www.who.int/mediacentre/multimedia/swineflupressbriefings/en/index.html (accessed 14 June 2010).

36.

Fukuda

(2009d) WHO press briefing. 26 May. Available at: http://www.who.int/mediacentre/multimedia/swineflupressbriefings/en/index.html (accessed 14 June 2010) (accessed 14 June 2010).

37.

Fukuda

(2009e) WHO press briefing. 9 June. Available at: http://www.who.int/mediacentre/multimedia/swineflupressbriefings/en/index.html (accessed 14 June 2010).

38.

Fukuda

(2009f) WHO press briefing. 3 December. Available at: http://www.who.int/mediacentre/multimedia/swineflupressbriefings/en/index.html (accessed 29 September 2010).

39.

Fukuda

(2010b) WHO press briefing. 18 February. Available at: http://www.who.int/mediacentre/multimedia/swineflupressbriefings/en/index.html (accessed 29 September 2010).

40.

Fukuda

(2010c) WHO press briefing. 24 February. Available at: http://www.who.int/mediacentre/multimedia/swineflupressbriefings/en/index.html (accessed 29 September 2010).

41.

Fukuda

(2010d) WHO press briefing. 10 August. Available at: http://www.who.int/mediacentre/multimedia/swineflupressbriefings/en/index.html (accessed 29 September 2010).

42.

Gaines

(1992) From DSM-I to III-R; voices of self, mastery and the other: A cultural constructivist reading of U.S. psychiatric classification. Social Science & Medicine 35(1): 3–24.

43.

Greenwald

(2008) Organizations: Management without Control. Thousand Oaks, CA: SAGE.

44.

Halbwachs

(1992) On Collective Memory (trans. Coser

). Chicago, IL: University of Chicago Press.

45.

Haraway

(1991) Simians, Cyborgs and Women: The Reinvention of Nature. New York: Routledge.

46.

Herzlich

Pierret

(1987) Illness and Self in Society. London: John Hopkins University Press.

47.

Hesse

(1974) The Structure of Scientific Inference. Los Angeles, CA: University of California Press.

48.

Jefferson

(2010) Speech – Dr. Tom Jefferson, the Cochrane Collaboration. Available at: mms://coenews.int.vod/100329_w01_w.wnv (accessed 20 July 2011).

49.

Jutel

(2009) Sociology of diagnosis: A preliminary review. Sociology of Health & Illness 31(2): 278–299.

50.

Kuhn

(1970) The Structure of Scientific Revolutions. Chicago, IL: University of Chicago Press.

51.

Latour

(1987) Science in Action: How to Follow Scientists and Engineers through Society. Cambridge, MA: Harvard University Press.

52.

Latour

(1996) On actor-network theory: A few clarifications plus more than a few complications. Soziale Welt 47(4): 369–381.

53.

Latour

(2005) Reassembling the Social: An Introduction to Actor-Network Theory. Oxford: Oxford University Press.

54.

Law

(1992) Notes on the theory of the actor-network: Ordering, strategy, and heterogeneity. Systemic Practice and Action Research 5(4): 379–393.

55.

Lewin

(1994) Categorization and the narrative structure of science. Philosophy and Rhetoric 27(1): 35–62.

56.

MacKenzie

Wajcman

(eds) (1999) The Social Shaping of Technology. Philadelphia, PA: Open University Press.

57.

Mahajan

(2008) Designing epidemics: Models, policy-making, and global foreknowledge in India’s AIDS epidemic. Science and Public Policy 35(8): 585–596.

58.

Mahoney

(2000) Path dependence in historical sociology. Theory and Society 29(4): 507–548.

59.

Marková

Farr

(eds) (1995) Representations of Health, Illness and Handicap. Chur: Harvard Academic Publishers.

60.

Martin

(2004) Can’t any body count? Counting as an epistemic theme in the history of human chromosomes. Social Studies of Science 34(6): 923–948.

61.

Martin

Lynch

(2009) Counting things and people: The practices and politics of counting. Social Problems 56(2): 243–266.

62.

Messeri

(2010) The problem with Pluto: Conflicting Cosmologies and the Classification of Planets. Social Studies of Science 40(2): 187–214.

63.

Monto

(1987) Influenza: Quantifying morbidity and mortality. American Journal of Medicine 82(6 Suppl. 1): 20–25.

64.

Moscovici

(1988) Notes towards a description of social representations. European Journal of Social Psychology 18(3): 211–250.

65.

Nguyen-Van-Tam

Hampson

(2003) The epidemiology and clinical impact of Pandemic influenza. Vaccine 21(16): 1762–1768.

66.

Olick

Robbins

(1998) Social memory studies: From ‘collective memory’ to the historical sociology of mnemonic practices. Annual Review of Sociology 24: 104–140.

67.

Prout

(1996) Actor-network theory, technology and medical sociology: An illustrative analysis of the metered dose inhaler. Sociology of Health & Illness 18(2): 198–219.

68.

Review Committee on the Functioning of the International Health Regulations (2005) in Relation to Pandemic (H1N1) 2009 (2011) Implementation of the International Health Regulations (2005): Report of the Review Committee on the functioning of the International Health Regulations (2005) in relation to Pandemic (H1N1) 2009. 64th World Health Assembly. Provisional agenda item 13.2, 5th May. World Health Organization.

69.

Roth

(2005) Making classifications (at) work. Social Studies of Science 35(4): 581–621.

70.

Rudwick

(1982) Cognitive styles in geology. In: Douglas

(ed.) Essays in the Sociology of Perception. London: Routledge & Kegan Paul, pp. 219–241.

71.

Sontag

(1978) Illness as Metaphor. Harmondsworth: Penguin.

72.

Steingart

(2012) A group theory of group theory: Collaborative mathematics and the ‘uninvention’ of a 1000-page proof. Social Studies of Science 42(2): 185–213.

73.

Stichweh

(1992) The sociology of scientific disciplines: On the genesis and stability of the disciplinary structure of modern science. Science in Context 5(1): 3–15.

74.

Strand

(2011) Where do classifications come from? The DSM-III, the transformation of American psychiatry, and the problem of origins in the sociology of knowledge. Theory and Society 40(3): 273–313.

75.

Taubenberger

Morens

(2006) 1918 influenza: The mother of all pandemics. Emerging Infectious Diseases 12(1): 15–22.

76.

Taylor

(2004) Governing the globalization of public health. Journal of Law, Medicine & Ethics 32(3): 500–508.

77.

Tognotti

(2003) Scientific triumphalism and learning from facts: Bacteriology and the ‘Spanish Flu’ challenge of 1918. Social History of Medicine 16(1): 97–110.

78.

Treacher

Wright

(eds) (1982) The Problem of Medical Knowledge: Examining the Social Construction of Medicine. Edinburgh: Edinburgh University Press.

79.

Vaihinger

(1949) The Philosophy of ‘As If’: A System of the Theoretical, Practical and Religious Fictions of Mankind (trans. Ogden

). London: Routledge & Kegan Paul.

80.

Webster

(2002) The importance of animal influenza for human disease. Vaccine 20 (Suppl. 2): S16–S20.

81.

Weindling

(1981) Theories of the cell state in imperial Germany. In: Webster

(ed.) Biology, Medicine and Society, 1840–1940. Cambridge: Cambridge University Press, pp. 99–155.

82.

WHO (2009) Pandemic Influenza Preparedness and Response: A WHO Document. Geneva: Global Influenza Programme, Health Security and Environment Cluster, WHO.

83.

WHO (2009–2010) Situation updates – Pandemic (H1N1) 2009. Update 1 – 112. Available at: http://www.who.int/csr/disease/swineflu/updates/en/index.html (accessed 9 September 2010).

84.

WHO (2011a) Global Influenza Programme. Geneva: World Health Organization. Available at: http://www.who.int/influenza/en/ (accessed 14 June 2010).

85.

WHO (2011b) International Health Regulations: External review of pandemic response. Available at: http://www.who.int/ihr/review_committee/en/index.html (accessed 20 July 2011).

86.

WHO (2011c) Weekly seasonal influenza updates. Available at: http://www.who.int/influenza/surveillance_monitoring/updates/GIP_surveillance_2011_archives/en/ (accessed 9 November 2011).