Date Published: August 02, 2017
Publisher: The American Society of Tropical Medicine and Hygiene
Author(s): Guillaume Fournié, Erling Høg, Tony Barnett, Dirk U. Pfeiffer, Punam Mangtani.
Almost all human infections by avian influenza viruses (AIVs) are transmitted from poultry. A systematic review was conducted to identify practices associated with human infections, their prevalence, and rationale. Observational studies were identified through database searches. Meta-analysis produced combined odds ratio estimates. The prevalence of practices and rationales for their adoptions were reported. Of the 48,217 records initially identified, 65 articles were included. Direct and indirect exposures to poultry were associated with infection for all investigated viral subtypes and settings. For the most frequently reported practices, association with infection seemed stronger in markets than households, for sick and dead than healthy poultry, and for H7N9 than H5N1. Practices were often described in general terms and their frequency and intensity of contact were not provided. The prevalence of practices was highly variable across studies, and no studies comprehensively explored reasons behind the adoption of practices. Combining epidemiological and targeted anthropological studies would increase the spectrum and detail of practices that could be investigated and should aim to provide insights into the rationale(s) for their existence. A better understanding of these rationales may help to design more realistic and acceptable preventive public health measures and messages.
All four of the influenza virus strains that resulted in pandemics in the last century have had an avian origin. While the 1918–1919 H1N1 pandemic strain was entirely derived from an avian virus,1 the subsequent pandemic strains of H2N2 in 1957, H3N2 in 1968, and H1N1 in 2009 all acquired gene segments from avian viruses by reassortment.2,3 Within the last 20 years, a variety of avian influenza virus (AIV) subtypes affecting domestic poultry—especially H5N1, H7N9, and H9N2—has resulted in human infections in mainly Asia and Egypt.4–6 Although these zoonotic transfers are sporadic and their transmission is not sustained within human populations, they also show a potential for reassortment with human viruses7; a very few nucleotide substitutions in some circulating strains might allow them to be transmissible between humans.8 It is widely feared that ongoing circulation of zoonotic AIVs within poultry populations and their transfer to humans could result in emergence of a novel human pandemic strain. As almost all human cases result from exposure to poultry or to environments contaminated by poultry,9–11 mitigation measures intended to prevent zoonotic infections and reduce the risk of adaptation of these viruses to human hosts must be carefully targeted, not only toward the poultry populations sustaining these viruses12 but also toward practices exposing people to infected poultry and contaminated environments. Mitigation measures have to take into account the complexity and difficulty of behavior change strategies and techniques, recognizing that “behavior” should not be construed as exclusively “individual” but as located within a socioeconomic and cultural milieu.
Both direct and indirect exposures to poultry in households, farms, or markets were associated with human infection by AIVs in most of the reviewed risk factor studies. The strength of this association seemed stronger for H7N9 than for H5N1, for sick and dead compared with healthy poultry, and in markets compared with any other setting. Several studies also suggested that the odds of infection further increased with the proximity between humans and poultry, the size of the poultry population to which humans were exposed, and the frequency of exposure. Direct exposure was not associated with higher odds of infection than indirect exposure. This apparent association between AIV infection and indirect exposure to poultry, and the possible role of handwashing and environmental disinfection as protective factors suggest that contacts with contaminated environments followed by ingestion, intranasal or conjunctival self-inoculation of the virus may be a major mode of AIV transmission. Infected poultry shed a high viral load, which may survive in the environment for a few days under favorable conditions.53 In households, virus survival in the environment may represent an infection pressure to which people may have prolonged contact, in particular when poultry are kept inside home, including in bedrooms. Even when environmental exposure is of a shorter duration, such as in the case of people visiting markets, the frequent introduction of infected poultry in markets and the associated viral circulation among marketed poultry54 means that humans may be exposed to high virus loads. However, high uncertainty remains regarding the actual modes of transmission involved. Contributions of aerosols and large droplets cannot be ruled out, as investigated exposures may be associated with several modes of transmission.55 H5N1 was shown to be transmitted between poultry by aerosols.56 Some practices, such as mechanical defeathering, may generate contaminated aerosols and large droplets, and result in the infection of people visiting markets.57 These results suggest that interventions aiming to reduce virus load in markets,58 and behavioral change strategies leading to higher biosafety standards when handling poultry, especially sick or dead specimens, could substantially reduce human exposure to AIVs. The adoption of risky and protective practices varied greatly across studies, and was frequently explained as motivated by financial constraints and religious beliefs. These variations could also result from temporal changes in people’s perception of their risk of infection. As these factors were expected to vary across the heterogeneous socioeconomic and cultural landscape covered by the reviewed articles, risk mitigation interventions should be tailored to these local contexts. However, none of the studies reviewed here aimed to assess the rationale behind practices at risk of human exposure to AIVs. Reviewed knowledge, attitude, and practice studies investigated questions related to awareness and knowledge, and a few studies did touch on some reasons behind specific practices and discussed these as post hoc hypotheses, but neither in sufficient detail nor at the appropriate level of conceptualization.
Our review was exposed to recall bias, as exposures were captured in all risk factor studies, and most practice prevalence studies, through structured interviews of study participants, or proxies when study participants have died. Bias in the measurement of exposures was more pronounced in risk factor studies using serology to define prevalent cases: AIVs being endemic in most settings, there was uncertainty about whether the reported exposures preceded, or not, the infection. However, this bias might be limited as most investigated exposures were daily, routine practices, which might not greatly change in the medium-term in AIV-endemic settings.
To address these limitations, epidemiological surveys could benefit from being combined with anthropological investigations. Anthropological studies may help to identify practices that would be better described with observations and in-depth interviews to develop a more accurate and detailed understanding. Such practices may, for instance, include handwashing and the use of PPEs. Moreover, the development of structured interview protocols would greatly benefit from a prior anthropological exploration of both the conceptualization of “practices” and of practices of interest. Whether “practices” are of interest may alter in the light of more detailed description and contextualization. Practices that are not systematically investigated, but which may reveal to be of epidemiological importance, may thus be characterized. Further description of practices could include a characterization of the contacts involved, and a measure of their intensity and frequency. The more detailed and grounded into the local contexts these descriptions are, the less comparable they may be across settings characterized by heterogeneous populations. On the other hand, the more general these descriptions are, the less likely it would be possible to tailor interventions to the local contexts that shape those practices. Detailed descriptions are required to identify the most relevant practices and populations at risk that should be targeted by risk mitigation strategies. Nevertheless, the small number of cases identified in most risk factor studies may limit the exploration of the association between AIV infection and specific, detailed practices. If these practices were only adopted by a small fraction of the population, the statistical power would be low, and even if the actual association with AIV infection was strong, the measured strength of association would be uncertain.