Due to the increasingly global and fragmented context in which they work, physicians are partially dependent on the flow of information conveyed through advertising. Marketing strategies work in two ways: on the one hand, pharmaceutical advertising simplifies messages and thus facilitates understanding of information usually requiring specialized knowledge; on the other hand, pharmaceutical advertising is both an authority figure and a stakeholder, as it is currently the most important source of financial support for medical journals as well as their main form of publicity. Although all advertisements have a commercial purpose and thus generally low credibility, medicinal drug promotion is an important filter of medical knowledge due to its indirect control over the media.1

Marketing strategies target the medical community and, to increase sales, do not always offer neutral information.2–4 To influence the perceptions and decisions of the physicians at whom such publicity is directed, reality is redefined through the use of partial and/or biased images of the target and is related to data on the benefits derived from certain properties of the drug.5 This redefinition is sometimes further accentuated by the use of information in the advertisement which is not always supported by bibliographic references.6

To provide a basis for a code of conduct in advertising, the World Health Organization (WHO) established their Ethical Criteria for Medicinal Drug Promotion.7 The pharmaceutical advertisement validity indicators used refer to the inclusion of approved scientific data, the number of bibliographic references and the accessibility of these references to physicians, as well as coherence between the epidemiological data presented in the advertising claims and the research findings.7 A tendency to increase the amount of scientific data included in medicinal drug promotion according to the WHO criteria has been observed over time.8,9

However, another marketing strategy used to achieve greater impact is the incorporation into advertising of images that segment the consumer according to socio-demographic characteristics,10 which may represent a further problem for validity11,12 when inconsistent with research findings.13 Specifically, the estimated prevalence of a disease among groups at risk occasionally seems exaggerated and thus misleading. Furthermore, the potential population base that could benefit from the medication is broadened through the use of inappropriate frames that include non-risk groups, to whom the therapeutic indications do not apply. This phenomenon is known as disease mongering.1,14

As a field of innovative research focused on the biological, physiological and sociological differences between women and men and how such differences affect diseases and their diagnoses and treatments, gender-based medicine shares with evidence-based medicine the hypothesis that inaccuracies exist in the production and dissemination of knowledge, as well as in medical practice, as regards rigor, transparency and subjective judgement.15,16 The media act as socialising agents and transmit messages that contribute to the social construction of disease. The quantitative and/or qualitative representation of both sexes in pharmaceutical advertisements is a focal point of interest in research on gender bias,11,12 since if such representations are inconsistent with reality they may reinforce the perception that certain illnesses are associated with the most frequently portrayed sex. Such is the case with advertisements for migraine treatments, in which images of women predominate.17

Despite the social changes experienced by women and men since the 1970s, the view of medicine as a gendered organization and male-dominated culture has had a powerful effect on gender imagery for women. This perspective has served to point to medicine as a cultural system tending to reinforce a gender identity based on traditional sex roles of women in society.11 The accuracy of knowledge transfer through pharmaceutical advertising is essential to avoid gender bias in medical practice and to achieve quality drug prescription according to knowledge-based evidence.18

The aim of this study was to review the scientific literature available on pharmaceutical advertising directed at health professionals (1998–2008) in order to determine whether gender bias has decreased and the quality of information in pharmaceutical advertising has improved over time. Clarification of these questions would allow health professionals to determine the current credibility of this information. As quality information criteria, we considered the findings of articles published on (a) the number, validity and accessibility of bibliographic references provided in pharmaceutical advertisements, and (b) the relationship between the sexes portrayed in the advertisements and the sex-prevalence of the diseases treated by the drugs advertised.

Original articles analyzing pharmaceutical advertising in medical journals and published between 1998 and 2008 were located using Boolean search techniques in the bibliographic databases PUBMED, Medline, Scopus, Sociological Abstracts, ERIC and LILACS. The key words differed depending on the thesaurus of the bibliographic source consulted. Thus, in PUBMED, Medline and Scopus, the key words “Advertising and Medical Journals” and “Advertising and Drug Industry” were used, while in Sociological Abstracts and ERIC, the key words were “Advertising and Medical Journals”, and “Advertising and Medication”. The search terms used in LILACS were “Publicidad y Revistas Médicas” (“Advertising and Medical Journals”) and “Publicidad de medicamentos e Industria farmacéutica” (“Medicinal drug advertising and the Pharmaceutical industry”). We selected all articles containing these key words both in their titles and in their abstracts. Articles without an abstract were excluded since they are not usually original studies and tend to be editorials, letters, practice guidelines or comments.

All original articles written in Spanish, English and Italian concerning pharmaceutical advertising aimed at health professionals and published in medical journals between 1998 and 2008 were included.

Excluded from the study were those articles focusing on advertising aimed directly at the consumer, advertising carried by media other than medical journals or popular science magazines, advertising ethics, or other marketing techniques used by the pharmaceutical industry as tools for ongoing medical training through sponsorship of courses and congresses.

A total of 1,715 articles were obtained, of which 552 did not provide an abstract. Of the remaining 1,163 articles, 1,132 were omitted in accordance with the exclusion criteria described in the preceding paragraph (fig. 1).

Figure 1.

Selection process for assessing the quality and gender equity of pharmaceutical advertising in medical journals (1998-2008).

(0.17MB).

A content analysis was carried out of the 31 original articles finally selected, for which an information compilation protocol was established and applied, comprising the following variables:

• 1)

  Scientific evidence in support of the advertising claims:

  • -

    Number of bibliographic references.

  • -

    Availability of the references in scientific bibliographic databases.

  • -

    Consistency between the contents of the claims and the scientific evidence published, measured through the agreement between the content of the message and knowledge of the drugs supported by scientific papers.

• 2)

  Sex-gender representation variables used in the analysis of advertising images developed in the original articles:

  • -

    Number of women and men portrayed.

  • -

    Occupational roles represented at work by sex.

  • -

    Family roles represented in the domestic environment by sex.

• 3)

  Medicinal drugs advertised.

• 4)

  Journals from which advertising was analyzed by the articles reviewed.

• 5)

  Specialty of the medical journal containing the advertising.

• 6)

  Date of publication of the articles analyzed.

• 7)

  Country of origin of the journals analyzed.

• 8)

  Type of research design used in the articles analyzed: quantitative and/or qualitative.

• 9)

  Period researched by the articles analyzed.

A content analysis of the above-mentioned variables was carried out.

Between 1998 and 2008, 31 original articles8,9,19–47 analyzed the pharmaceutical advertising published in medical journals over a period of 30 years (1975-2005) (Table 1). The number of advertisements analyzed per article ranged from 26 to 4,157.

Most of the advertising analyzed was of European origin,19–23,27,29,32–35,38–45 particularly northern Europe22,26,34,35 and England.29,33,34,38–40,45 Other advertising came from the USA,24–31,33,36,38–42,44–46 Brazil,9,36 and Canada.27,33,39,40 By comparing results from different articles, we noticed that the percentage of advertisements decreased from 50% in 1975 to approximately 30% in 2000.19,21,24,39,41 However, these differences varied according to the journal and country of origin.19,21,24,28,29,35,39,41,44 British journals carried significantly fewer advertisements9,39 than those from the USA, which displayed the highest levels of advertising pressure.9,39

Two varying approaches were used in advertising analysis. On the one hand, an analysis of all pharmaceutical advertisements published in medical journals and science magazines was carried out,20,29,33,36,38,39,43 while on the other, the analysis focused exclusively on medicinal drugs used within specific specialities, such as antihypertensives and antilipids for cardiovascular problems, corticosteroids to treat respiratory problems, medication for erectile dysfunction, psychotropics, analgesics and antibiotics.9,19,21–28,30–32,34,35,37,40,41,44–47

The most frequently made claim was that of the effectiveness of the drug,16–18,24,26 followed by claims related to drug safety.19,20,27

Only eight articles (25.8%) presented longitudinal comparisons of advertising over a specified period.9,20,33,36,38,39,41,43 The remaining articles presented cross-sectional studies that reviewed advertising published at a particular moment in time8,19,21–32,34,35,37,40,42,44–47 (Table 1).

The bibliographic references provided in pharmaceutical advertising were easily accessible for consultation.26,46 In 2005, 65% of bibliographic references were found in Medline, PsycInfo and Google46 while in 1999 approximately 84% were found in databases such as Medline, PubMed, WorldCat, California Digital Library and Google.26

Fifteen articles9,19–21,25–27,29,32,36–38,41,42,46 examined the number of references provided in pharmaceutical advertising (Table 2). There seems to be a temporal pattern, with an increasing tendency to include references. Nevertheless, reference frequency varied according to the journal's country of origin and medical speciality. For instance, articles analyzing advertising in Russian,42 Italian16 and Spanish21 journals included fewer references than those analyzing advertising in countries such as the UK, Netherlands, or the USA.

However, in 50% of pharmaceutical advertisements, the references provided as scientific support for the claims lacked validity19,20,25,29,32,37 (Table 2) as these claims were insufficiently supported by recorded scientific data, prescribing information, randomized controlled trials or other research.25 Neither were they supported by scientific evidence,19,20,37,44 as 44.4% of unsupported claims recommended the drug for a patient group different to that studied in the trial29 or were classified as ambiguous (vague: 54.3%; emotive: 14.2%; non-clinical: 12.4%).32

Nine articles provided information on the sex of the people appearing in the advertisements and the gender dimension analysis categories used22,23,30,31,33,36,41,43,44 (Table 3). There was a tendency to depict men in paid productive roles, while women appeared inside the home or in nonoccupational social contexts.23,33,36,43,44 Moreover, while the articles that analyzed advertising for cardiovascular disease medication coincided in their findings that men were given higher visibility,22,23,31 those that focused on psychotropic pharmaceutical advertising concluded that women were over-represented.23,33,36,43,44 The longitudinal studies revealed that the closer the study period to the present day, the greater the number of advertisements for antidepressant treatments and the higher the percentage of women portrayed.33,43

The type of information used to promote medicinal drugs varied among countries. An article that compared advertising in psychiatric journals in Brazil, the USA and the UK observed that journals in the USA and the UK contained more scientific data about the drug.9 More specifically, advertising in Brazilian psychiatric, pediatric and general medicine journals lacked information on drug interactions and adverse effects.9,36 Advertising carried by Russian journals on surgery, general medicine and cardiology also contained scant data on drug interactions, adverse effects and contraindications.42

The articles that employed forward-specific longitudinal analysis showed that from 1975 to 2005 the number of bibliographic references used to support pharmaceutical advertising claims increased.9,36,38,41 However, the accessibility, 26,46 as well as the validity of the scientific references was limited.19,20,25,29,32,37 No improvement was observed in advertising quality in developing countries or when quality was examined from a gender perspective. Articles carrying out a longitudinal analysis provided evidence that the same traditional male-female roles were represented in the first years studied (1975, 1981, 1985) as well as in the later years (1995, 2001).33,36,43 Gender bias was also identified in the studies analyzing specific years during the same period30,31,44 and in the following years (2001-2002, 2005).22,23

One of the limitations of this study is the small number of articles found. Although the review period spanned 10 years, few articles presented an empirical analysis of the information contained in pharmaceutical advertisements. Nevertheless, the articles analyzed cover 30 years of advertising (1975-2005). A further limitation is the diversity of the approaches used by the different authors in the analyses, with less than a third of the articles reporting a gender analysis or presenting results providing information on the sex of the consumer or the physicians portrayed in the advertisements. Moreover, most of the pharmaceutical advertisements analyzed were related to cardiovascular or psychotropic drugs and thus the possible existence of bias in pharmaceutical advertising for other medicinal drugs was not determined.

The advertising strategies erroneously over-represented one sex without taking into account the sex-prevalence of the disease, which contributed to maintaining a gender bias in medical knowledge and diagnostic and therapeutic decisions.22,23,31,33 This imbalance in therapeutic targets depicted in pharmaceutical advertising was observed in advertising for both mental health23,33,36,43,44 and cardiovascular disease medication.22,23,30,31 Accordingly, health professionals should maintain a critical approach when evaluating the credibility of the information contained in pharmaceutical advertising. Likewise, health authorities should actively monitor the consequences of this type of advertising strategy, which could be converting a private commercial objective into a social problem.

While a drug's effectiveness is one of the main arguments used in pharmaceutical advertising, some authors highlighted the lack of epidemiological evidence, such as data on risk reduction in absolute terms or on the number of patients treated, which could be used to support such claims.38,47 Indeed, as regards representations of male and female identity, Goffman claimed that these were commonly represented through stereotyped images of what it means to be a man or a woman.48 Such representations may reflect the sexist construction of scientific knowledge49 which, in terms of public health, has been criticized since the 1980s in the form of the Gendering the Medicalization Thesis.11

The availability of scientific data is explicitly recommended in the WHO ethical criteria for medicinal drug promotion. However, Carvalho showed that availability varied according to the country of publication of the journal carrying the advertising. Specifically, less information concerning interactions, adverse effects, precautions or contraindications was provided in developing countries than in developed countries.9

The longitudinal studies included in this review, accounting for 25.8% of the total, showed that from 1975 to 2005, pharmaceutical advertising validity increased in terms of providing scientific data and references in support of the advertising claims made. However, validity depended on the stage of development of the journal's country of origin.8,9,39,41 According to many authors, such a lack of data can contribute to the medicalization of the population's health and encourage disease mongering,1,4,50–52 both of which would occur more readily in developing countries.

The WHO medicinal drug promotion validity criteria basically refer to scientific and epidemiological data, and to the accessibility of documents supporting advertising claims.7 However, as images are one of the main tools employed in advertising to transmit information and create and reflect collective perceptions, these criteria should include the graphic representation of the consumer as a further validity criterion. This would allow determination of whether the sex or age frequency represented in the advertisement is consistent with the epidemiological data, that is, the sex and age prevalence of the health problem for which a certain drug is commercialized.

The spectrum of diseases for which the gender analysis of pharmaceutical advertising was performed and published was limited to mental health and cardiovascular disease medicinal drug promotion.22,23,30,31,33,36,43,44 These analyses confirmed that women were consistently over-represented in psychotropic drug advertising,23,33,36,43,44 while the same was true for men in cardiovascular disease medication advertising.22,23,30,31 However, scant research has been published on whether this biased and partial over-representation of one sex is repeated in pharmaceutical advertising for other types of medical treatments.

The under-representation of women in pharmaceutical advertising for cardiovascular disease treatments correlates with the generation, dissemination and application of medical knowledge on these diseases.15,16,49 This same tendency toward bias has been observed in some clinical drug trials which are subsequently cited in advertising, where either few women are included in the trials compared to the number who will eventually consume the drug once commercialized, or information is provided which has not been stratified by sex for efficacy and effectiveness.53 Thus, when concluding that the generation of new knowledge in medicine focuses on male health and merely infers knowledge of female health, Goffman's thesis on advertising is in agreement with feminist empiricism as applied to health.11

This double bias –generated by both science and advertising– could undermine medical practice54 by helping to perpetuate a perception that is inconsistent with data on the actual prevalence and mortality rate of the problem. For example, the higher representation of men than women in cardiovascular medication advertising is not consistent with the actual number of deaths from cardiovascular disease within the European Union (56% women compared with 43% men).13

Despite the diversification of the role of women over the last few years, the findings of the original papers reviewed reveal that parity between men and women in terms of the performance of social tasks or functions is not reflected in the pharmaceutical advertisements analyzed.22,23,30,31,33,36,41,43,44 There is a consistent tendency to embody stereotyped images of both sexes. Specifically, women are portrayed in submissive or passive activities, such as sleeping33 and, in 2005, were still usually shown in traditional contexts and roles, such as cooking, shopping, or at home with clothes and face covered in patches of whipped cream.22 Also, when advertising depicts men with cardiovascular disease, the disease is attributed to external factors, such as occupation.23 In contrast, pharmaceutical advertising for psychotropic drugs embodies women without a context, suggesting that such problems are generated by the woman herself.23

The findings obtained from the articles reviewed shows that the quality of information on pharmaceutical advertising has improved through the increase of bibliographic references during the study period. However, some bibliographic references mentioned in the advertising could not be accessed, and there were some difficulties related to agreement between the content of the message and the knowledge of drugs supported by scientific papers.

B. Cambronero carried out the systematic review in the main databases, participated in the study design, performed the statistical analysis and helped to draft the manuscript. M.T. Ruiz conceived the study, participated in its design and coordination and helped to draft the manuscript. N. Papí participated in drafting of the manuscript. All authors read and approved the final manuscript.

None.

None.