In which journals do family and community physicians in Brazil publish? The Trajetorias MFC project

Introduction: Authors choose scholarly journals not only to advance their careers but also to interact with their respective scholarly communities. Objective: To describe the journals where family and community physicians in Brazil publish their work. Methods: In late 2018, we compiled a nationwide list of family and community physicians, and downloaded their curricula vitae (CV) from the Lattes Platform. We extracted data on their complete journal articles from their CV, completed these data with queries to CrossRef, VHL/LILACS, and PubMed/MEDLINE, and obtained data on the journals with queries to the United States NLM Catalog. Results: We found 3622 unique articles, published by 1014 journals. The most productive journal was RBMFC (Revista Brasileira de Medicina de Familia e Comunidade), which published 353 (9.7%) of these articles. About one in six articles were published in journals on family practice or primary health care. The proportion of articles published in journals in Brazil decreased during the study period from 83.3% to 59.0%. Conclusion: As in other countries, family and community physicians in Brazil usually publish in the national journal dedicated to their scholarly community, while also publishing extensively in journals from other disciplines. The increasing proportion of articles published in journals outside Brazil suggests primary care research in Brazil is increasingly of international relevance.


Introduction
Scholarly journals are essential infrastructure for their respective scholarly disciplines and communities. [1][2][3] While publishing in journals with high prestige or impact factor advances the careers of individual researchers, publishing in journals with the right readership communicates the scholarly output to the corresponding scholarly community, thus contributing to its intellectual conversations. [2][3][4][5] Family medicine journals are thus strategic for strengthening the capacity for primary care research, [6][7][8][9] which in turn informs primary care and reinforces its status as a scholarly discipline. [10][11][12] The status of family medicine or primary care as scholarly disciplines varies substantially across countries. 8 In Brazil, family and community medicine was recognized as a medical specialty in 1981, 13 but most physicians in primary care are not certified family and community physicians. 14 Furthermore, research funding and stricto sensu postgraduate education (master's and doctorates) don't recognize family and community medicine or primary care as a scholarly discipline, and few universities have a department dedicated to either. 15 The lack of recognition of their scholarly discipline might compel family and community physicians in Brazil to engage in "intellectual discussions" (to borrow from Wakeling et al. 3 ) with neighboring scholarly communities, such as other medical specialties or public health, rather than among themselves. In this study, our objective was to describe the scholarly journals where they publish their scholarly output, as an approach to better understanding the state of primary care research in Brazil.

Methods
This was an observational, exploratory study, using administrative data from multiple sources, as part of the Trajetórias MFC project. 16

Data sources
As previously described, 16  (after this manuscript underwent the first round of peer review), and the statistics were updated accordingly.
Throughout the data acquisition process, all data were extensively verified by three authors (LFF, MHMO and SVR). For example, we made sure every journal was always listed with the same name, consulting when necessary the corresponding journal entries in the NLM Catalog, the ISSN International Center ROAD (Directory of Open Access scholarly Resources) 21 and the VHL Serials in Health Sciences (SeCS; in Portuguese, Seriados em Ciências da Saúde). We also made sure each journal article had the same LILACS identifier, PMID, and/or DOI when listed more than once, and furthermore considered journal, volume and starting page to identify articles published in collaboration by more than one family and community physician.
We defaulted to trusting data contained in the curricula. For example, we verified the DOI retrieved from CrossRef but not the ones provided in the curricula. Sometimes, however, verifying the internal consistency of programmatically acquired data revealed mistakes in the curricula requiring data editing (for example, wrong page number) or even record removal (for example, journal articles entered more than once by the same author, conference abstracts published in supplements of scholarly journals, or articles published in clearly non-scholarly journals).

Variables and analysis
The scholarly output of family and community physicians from Brazil was described using the year of publication and the journals' titles, country, and major subjects. We made these data freely available at Zenodo. 22 Journal articles were included only if they were published during or after the year of specialization, which was taken to be the year of certification by SBMFC or conclusion of the medical residency, whichever came first. The year of publication was categorized in five-year periods, except for the first period, which included the years 1985 (first included article) through 1998 (last year before the Lattes Platform became online).
The MeSH terms for the journal's major subjects were derived from the corresponding entries from the NLM Catalog. Because the number of major subjects per journal ranged from zero to six, we weighted major subjects by the inverse of their number within a journal. For the sake of simplicity, we did not weight the journal articles by the number or proportion of family and community physicians among the authors.
The description of the most productive (that is, with more included articles) journal titles emphasized those amounting to one-third of the included articles.

Results
From  (Table 2). Surprisingly, the most productive journal in this category was Revista de APS, which is cataloged as having medicine as its only major subject, rather than primary health care as expected. The next journals in this category were, in order: Revista AMRIGS, Arquivos Catarinenses de Medicina, and PLoS One.
The second most frequent major subject was public health, with 431.4 (15.3%) articles ( The third most frequent major subject was primary health care, with 352.3 (12.5%) articles (Table 2).
Almost all articles in this group were published in RBMFC, which is cataloged with primary health care as its major subject and family practice as a secondary one. If Revista de APS were included in this group, primary health care would become the second most frequent major subject, ahead of medicine and close to public health. to 58.4% (Figure 1), while the proportion in journals from other countries rose accordingly.

Discussion
As of late December 2018, family and community physicians in Brazil had published their scholarly articles in more than a thousand journals, from tens of countries. Nevertheless, their national disciplinary journal (RBMFC) was also the most prolific one, publishing one in ten of their articles. This might come as a bit of a surprise, because RBMFC has not been favored by traditional journal-level metrics. In Brazil, such metrics are central to the advancement of individual researchers' careers, allocation of research funding, and important administrative decisions about stricto sensu postgraduate programs. In this context, the prominence of RBMFC suggests family and community physicians value the opportunity to engage in intellectual conversations with their own scholarly community. Such preference for national journals dedicated to their scholarly discipline has also been found among family physicians from the United States, 23 faculty in university departments of family medicine in the United States, 24 and primary care researchers from Spain. 25,26 This pattern seems to confirm the importance of national and regional organizations of family medicine maintaining journals to strengthen the scholarly status and research capacity of the discipline. [6][7][8][9] On the other hand, only one in six articles were published in journals about family practice or primary health care (with or without Revista de APS); the other articles were scattered in journals about public health, medicine in general, medical/health education and a myriad of other subjects. This diversity of journal subjects is well documented for family physicians and for faculty in departments of family medicine, at least in the United States. 23,24 Being at the border of clinical medicine with applied social sciences, 27,28 family and community medicine has ample opportunity for interdisciplinary research, which might or might not be published in journals about family practice or primary health care. Furthermore, journals of bordering disciplines may be more attractive for career progression because of publishing more basic research and thus having higher journal-level citation metrics. 29 Although interdisciplinary publication is an international characteristic of family medicine as a scholarly discipline, the Brazilian emphasis on public health deserves further scrutiny. Of the 14 journals accounting for a third of the articles written by family and community physicians from Brazil, six are public health journals per the NLM Catalog. One reason might be that these are actually journals about collective health, which in Latin America is a scholarly discipline comprising public health, epidemiology, and human and social sciences in health, 30 among other definitions. 31 This wider scope (compared to public health) means collective health frequently includes health research which is not strictly biomedical but would be considered medical research in other contexts.
Another reason might be that most family and community physicians in Brazil earn their master's and Ph.D. degrees in collective health, not medicine. 16 This prevalence of postgraduate degrees in collective health might explain why, as reported by Almeida et al., 28  Our findings should be interpreted with some limitations in mind, despite deriving from carefully curated data obtained from reliable sources. For instance, consistently with Brazilian norms on medical specialization, we included only physicians concluding the medical residency since 1981, as well as those undergoing special certification since 2004. When family and community medicine (then "general community medicine") was recognized as a specialty, medical doctors having already concluded their medical residencies, or having experience working as family and community physicians, could not opt to be recognized as such. We can only wonder what scholarly communities do these unacknowledged family and community physicians interact with.
Furthermore, our study was restricted to the scholarly output published in peer-reviewed journals, because of their centrality to the advancement of scholarly disciplines and relative tractability to quantitative analysis. In consequence, our findings might not apply to scholarly communication through other media, such as conferences and textbooks. Furthermore, journal articles are not a direct measurement of knowledge advancement, because not all journal articles contribute equally.
Finally, the temporal context of our data should be considered. First, journal articles were eligible only if published after specialization, to ensure the relevance of the included articles to family and community medicine. The downside is that some of the excluded articles might be relevant, too. Last, our findings are probably overrepresentative of early career researchers, because most family and community physicians concluded their medical residencies or were certificated in the last 10 years 16 and, as a result, the specialty has a fairly young demographic profile. 32 The young demographic profile of family and community medicine may constitute an opportunity for its development as an academic discipline. Primary care research has a wide spectrum, and research projects should be adequately distributed along this spectrum for primary care research to fulfill its role in the improvement of primary health care. If the sparsity of clinical research in primary care can be attributed to the stricto sensu postgraduate education received by family and community physicians in Brazil, the recently proposed master's degree in family and community medicine 15 might be pivotal in improving clinical practice in primary health care.