The results highlight a number of important points about media coverage and stakeholder participation in relation to antibiotic use and the process of regulating antibiotic sales in Mexico that could be relevant for other LMICs aiming to develop policies directed to improving the use of medicines.
Even though the theme of regulation dominated media coverage, the focus was primarily on the problems of developing the policy, and little, except for self-medication, about its objectives or its relation to public health issues. The other dominant themes in media debates mirrored the voice of the two major stakeholders involved. On one hand, the MOH defended the policy by citing the dangers of self-medication with antibiotics; on the other, pharmacy associations opposed it, citing issues of economic impact and corruption. These divergent frames were scarcely addressed by the media on the common ground of a wider public debate on rational use of medicines and pharmaceutical policy. The emphasis on the problem of self-medication probably minimized public concerns about new physician offices attached to pharmacies and contributed to the belief that, with the policy, problems in antibiotic use had been solved. This represented a missed opportunity to discuss in-depth pharmaceutical policies and the development of a national strategy on antibiotic use. Additionally, media coverage also represented a missed opportunity to sensitize the public about the problem of antibiotic resistance and the need to use antibiotics prudently, which coincides with other studies which stress the scarcity of key mobilising information on health directed to the public .
Besides the interests and resources of stakeholders, the predominant themes reflected in news coverage of antibiotic use and antibiotic regulation can also be explained by journalistic conventions. These conventions tend to favour certain generic frames such as “episodic framing”  in which social issues and public affairs are portrayed as limited to events only, and not discussed in a broader context (e.g. coverage centred on the policy enactment, rather than in the problem of antibiotic use). Other common generic frames found in our study are “conflict” (winners and losers, corruption) and “economic consequences”. Coverage might also be explained by a regional trend characteristic of Latin America of health news reporting leaning towards frames of political conflict .
The present study also sheds light on the actors and processes involved in health and pharmaceutical policy making in Mexico. The executive government, mainly the MOH and the medicines regulatory authority dominated media coverage. This agrees with studies portraying the process of public policy making in Mexico as state-centered, with low levels of pluralistic debate . The scant involvement of the legislature and the fact that coverage reflected contradictory views of individual legislators instead of that of an ad-hoc expert commission also coincides with other studies highlighting the traditionally weak role of the legislature and its advisory commissions in decision-making [36, 37].
The associations of small independent pharmacies mostly affected by the regulation were successful in organizing press round-ups to push their views in media debates and make specific demands to the government. These groups also took advantage of this momentum to advance their own agenda, such as the threat of competition by pharmacy chains. This coincides with other authors pointing out that, although Mexico has attained a media establishment relatively independent of government control, it has also been increasingly beholden to commercial interests [38, 39]. However, even when these association groups had a strong voice, their effect appears to be limited, suggesting that the influence of political factors may outweigh media framing effects on shaping policies, as other studies have pointed up .
Large pharmacy chains had a minor presence in media debates, as did the transnational pharmaceutical companies because their interests might not be greatly affected by the policy. Some large chains buffered the impact of the policy by opening new physician offices, which might have favored the prescription of branded medicines. Private enterprises largely framed the policy debate as a business issue, rather than a health issue, which resulted in a focus on maximizing profits rather than improving medicines use. This highlights the inherent complexities of pharmaceutical policies competing with both economic and public health interests.
Academic institutions and medical associations also took advantage of the policy momentum to push their own agendas, especially during the agenda setting and the early drafting periods. They used the media to stress the need to develop a comprehensive strategy on antibiotic use, a theme that had little resonance in the media, and had little impact on the final policy.
The scarcity of newspaper coverage describing dialogue or interaction between interest groups and the MOH, and the lack of governmental response to the demands of interest groups (such as developing a public information campaign), reflect a political context characterized by vertical and closed decision-making. In addition, remarks in the media and editorials speculating about the true reasons for the enforcement suggest a lack of transparency and lack of trust in the authorities.
Traditionally, civil society has had a very marginal role in decision-making in Mexico [36, 37] with the voice of these actors barely present in media coverage. The absence of opinions by pharmacist associations can be explained by the small number of pharmacists working in private sector pharmacies. The large majority of pharmacy staff lack university degrees and undergo very little professional training in pharmaceutical sciences or related areas . Hence, pharmacy associations with a business focus rather than professional associations emphasizing professional credentials dominated the media debate.
Our results coincide largely with the findings of an analysis of printed media coverage of tobacco policies in Mexico  which reported the overwhelming presence of governmental actors, and limited inclusion of academic centers and civil society organizations. However, because the tobacco study included coverage of the development of two laws, the presence of actors from the legislative branch was significantly larger. Interestingly, our results on actor participation contrast sharply with a similar study on medicines policy developed in Canada  where civil society was the dominant voice, demonstrating the strategic use of the media by these groups. Although Mexico is undergoing a democratic transition, the experience of media advocacy by academic and civil society groups is still developing.
Some lessons can be drawn from our study which are relevant for other low and middle-income countries aiming to develop policies directed at improving antibiotic use. First, active opposition by some stakeholders to the enforcement of regulations on antibiotic sales should be expected, particularly arguing the economic impact on the population. Actions seeking to counteract the policy goals include opening physician offices next to pharmacies. Second, even if the voices of governmental and business-related groups are favored in media coverage, the policy views of other groups can also achieve visibility in the media which could contribute to a more informed policy development process. Third, this study showed that even if governmental action focuses on a single problem and a single intervention (enforcement of sales regulations to avoid self-medication), other related issues will soon arise in public debate. These issues include access to medicines (a very sensitive issue for the population), the quality of medical services, the tracking of prescriptions and verification of their authenticity, and the adverse impact on commercial interests. These concerns highlight the need to discuss and develop interventions within the common framework of a pharmaceutical policy, and to engage stakeholders during the policy process. Finally, the scarcity of in-depth reporting on issues of antibiotic use and regulation in Mexico and Latin America has also been related to the limited availability of specialized journalism and independent information sources . To address these limitations and to generate political priorities on antimicrobial use and resistance, the South American Infectious Disease Initiative  has worked with the news media in three countries, resulting in an improvement on the quantity and quality of coverage of these issues , an experience that might be worth replicating in other countries.
How does the Mexican case compare with other relevant experiences? The cases of India and Brazil, both middle-income countries that undertook antibiotic policies contemporaneously to Mexico are worth discussing. During 2010, international attention turned to the spread of the New Delhi metallo-beta-lactamase 1 (NDM-1) gene, believed to have originated in India, which causes several types of bacteria to become resistant. The naming of the gene and a description of its spread from that country to Europe in an influential international medical journal was followed by an intense public debate in India . Controversy triggered action and, in 2011, the Indian government published the National Policy for Containment of Antimicrobial Resistance. The policy included, among many other interventions, the banning of over-the-counter antibiotic sales. However, later the government decided to put this regulation on hold indefinitely, arguing concerns over access to medicines, particularly in rural areas . In the case of Brazil it was the spread of the multi-resistant bacterium CRKP (Carbapenem-Resistant Klebsiella pneumoniae), also followed by the media, which gave place to the resolution RDC 44/2010 of the Brazilian regulatory agency, regulating antibiotic sales to medical prescription only, to be retained at pharmacies. The resolution, supported by medical groups, faced the opposition of pharmacy associations, citing scarce access to medical services for the poorest populations, as well as the risk of triggering a parallel antibiotic market . Taken together, the cases of these three countries underline the relevance of focusing events (influenza outbreak, NDM-1 and CRKP spread) to place issues of antibiotic use and resistance on the governmental agenda. However, these cases also show that an apparently technical issue (regulating antibiotic sales) can be very politically sensitive – because economic interests are affected and because of the relation to wider complex societal issues – underlying again the need to discuss antibiotic policies within the common framework of a pharmaceutical policy, and to mobilize political support.
Our study presents some limitations. First, our analysis used a unique coding scheme (“issue-specific” instead of “generic” frames) which limits our ability to generalize and compare the results . However, whenever possible, we discussed our coding scheme in relation to generic frames. Other limitations concern excluding an analysis of electronic media (radio, television and internet) which are important in Mexico and other countries in the region  as well as the fact that our analysis was limited to the period of only five months after policy implementation. Relevant discussions and stakeholder actions that might have appeared later in the policy process or in other media outlets were excluded. Finally, as others pointed out, we can describe the representation of issues and stakeholder participation in the media, but we can only infer their potential effects on the public and on the policy process . To provide a more comprehensive picture of stakeholder participation in the policy process it is necessary to use alternative data sources.