Introduction

In the last decade, global health researchers have increasingly been calling attention to the authorship patterns for articles published about low- and middle-income countries (LMICs), noting in particular that authors from high-income countries often play the dominant role. See Abimbola (2019) for a commentary and Mbaye, et al. (2019) for some empirical findings. The Commission on Health Research reported in 1990 that 90 percent of the global disease burden was in global South. However, the South receives only 10 percent of the global funding to research those diseases. This lack of funding creates a significant research production gap, which exacerbates the representation of authors from the global South (Schneider and Maleka, 2018.) Brown (2019) finds that the authorship share from LMICs for impact evaluations conducted in LMICs is actually better in the global health fields than in the social sciences. Roughly half of the authors of development impact evaluations published in global health journals have LMIC affiliations, but the authorship share is only one quarter for those published in social science outlets.

We, along with our colleagues Ebelechukwe Monye and Otto Chabikuli, are analyzing the authorship patterns for articles published by staff in our company during 2018 and 2019 as a baseline for advancing our own practices. For this research, we are looking at patterns both in terms of the gender of the authors and whether they have a professional affiliation in the country they are researching. Our research builds on a growing body of evidence, and in this post, we summarize two articles from our background reading that focus on the lead authorship of articles for health research conducted in LMICs. Together the articles cover publications from 2000 to 2016 and generally find that only roughly half of these articles have local or LMIC first authors.

Local and foreign authorship of maternal health intervention research

Matthew Chersich, et al. (2016) conducted a large-scale systematic mapping of the literature on maternal health interventions in LMICs from 2000-2012 to analyze the alignment of research on maternal health to the main challenges individual countries face. In a companion article they analyze the authorship pattern for the sample of 2,292 included studies in their mapping research (Chersich, et al., 2016). For each study they code the country or countries of study and the country of countries of the first author’s professional affiliation as listed in the article.

They find that 49.1 percent of the 2292 articles were led by LMIC affiliated authors; 43.4 percent by authors from high income countries (HIC); and 7.5 percent by authors with multiple affiliations including both LMIC and HIC affiliated institutions (p. 4.) First authors affiliated with institutions in United States and United Kingdom share the highest number of publications, accounting for one third of the total. Chersich, et al. define a “local lead author” as a first-listed author with an affiliation in at least one of the countries where the research was conducted, and they find that of the empirical, non-review studies in their sample, 60.0 percent have local lead authors.

They also explore local first authorship by article characteristics. For example, of the maternal health topics covered in the included studies, hypertension has the highest share of local first authorship at more than 80 percent. Malaria and STDs other than HIV both have greater than 60 percent local first authorship. The lowest first authorship shares are for emergency obstetric care and traditional birth attendance, although these topics also have a relatively small number of included studies. Dividing the set by research conducted in low-income, lower-middle-income, and upper-middle-income countries, the authors find that 80 percent of those conducted in upper-middle-income countries have local first authors compared to less than 50 percent for low-income countries. They also look at the prevalence of local first authors by journal impact factor (IF) in three categories: IF ≥ 5, 2 ≤ IF < 5, and IF < 2. Not surprisingly, they find that the share of local first authors is lowest for journals in the higher impact factor group, although only 40 percent of their population of studies came from journals with impact factors.

The authors also discuss “data ownership” and “research capacity” in LMICs, but they define both these concepts with the same metric – local lead authorship. They rightly note in the limitations section that this metric may not be a good proxy for those more complex concepts. There is no doubt about inequities in research capacity and data ownership between high-income countries and LMICs, but it is important to keep in mind the well-known flaws with academic publishing that make it hard for researchers without the “right” affiliations to publish, even holding capabilities equal.

Chersich, et al. make several recommendations to address inequities between LMIC and HIC researchers. These include revisions in authorship and attribution guidelines. They note, in fact, that some journals now requires inclusion of a local authors for studies conducted in LMICs. They also discuss mentorship, research infrastructure, and funding dedicated to improving academic writing.

Patterns of authorship in community health worker publications

Helen Schneider and Nelisiwe Maleka (2018) conduct similar research to Chersich, et al. on a set of studies published from 2012 through 2016, identified for a systematic scoping review about community health workers (CHWs) in LMICs. Schneider and Maleka look at both lead and last authors, recognizing that last authorship in health research typically designates the senior author. Their primary analysis focuses on the region of the authors’ first-listed country of affiliation: high income, middle income, or low income. They do report, however, that “the vast majority (+95%) of LMIC lead/last authors were also local lead/last authors,” (p. 6.)

Schneider and Maleka find in their sample of 649 articles that 53 percent have LMIC first authors and 46 percent have LMIC last authors. Not surprisingly, the lowest percentage is for low-income country lead and last authors, at 15 percent or lower for both categories in all five years. They also examine which countries within these groups have the highest and lowest representation. Similar to Chersich et al., Schneider and Maleka also find that the United States and the United Kingdom account for the highest shares of lead and last authors.

Schneider and Maleka also coded the number of authors for each study and whether each author is local to the country(ies) of study. They find that the proportion of a study’s authors who are local authors is higher for studies conducted in middle-income countries than for those conducted in low-income countries. They also find, “the proportion of local authors in authorship teams was strongly associated with the provenance of the last author…publications with HIC last authors had the lowest mean proportion of local authors (0.34) compared with 0.63 and 0.78 for publications with LIC and MIC last authors, respectively (p<0.001)” (pp. 5-6.)

In their conclusions, Schneider and Maleka highlight the importance of funding and research practices in the global North. For researchers in HICs to better support and promote their LMIC counterparts, the performance metrics and incentives in their home institutions need to change.

Conclusion

Both studies present much more analysis and thoughtful discussion than we summarize here. They both point out that although there has been a dramatic increase in global health research conducted in LMICs over the last few decades, the leadership of LMICs researchers in this work has lagged. Among the concerns about these inequities is the quality of the research – the histories, culture, contexts and particular needs of each country cannot be adequately addressed without local voice and leadership. To accelerate improvements in authorship representation, efforts are needed across multiple fronts, including research funding, academic publishing, and institutional incentives.

Photo Credit: Jirapong/Adobe Stock

Sharing is caring!