Representation of low and middle income countries in traumatic stress journals from 2006 to 2015

Abstract

Although the majority of trauma survivors reside in low and middle income countries, these regions have historically been poorly represented in traumatic stress research. However, a recent review of geographic representation within traumatic stress journals is lacking. This study reviewed all articles published between 2006 and 2015 in six leading traumatic stress journals. When articles (n = 2530) were categorised by region, less than one tenth (9.76%; n = 247) were representative of low and middle income countries. All articles categorised as being from a low and middle income country were then coded for the regional and country representations of samples, author affiliations, and funding sources, and for type of research methodology. The majority of primary author teams (56.28%) and funding sources (55.87%) for articles based in low and middle income countries were located in high-income countries. The majority of low and middle income country articles (71.66%) used structured symptom questionnaires of which more than two thirds (70.6%) assessed symptoms of posttraumatic stress disorder. These findings indicate that knowledge production about traumatic stress in domain-specific journals reflects a state of ongoing geographic inequality and that research published from low and middle income countries reflects predominantly etic methodologies centred on measuring posttraumatic stress disorder. The implications are discussed and recommendations are offered for developing a research base in domain-specific journals that better represents the experiences and needs of trauma survivors in low and middle income countries.

Keywords

Knowledge production low and middle income countries review traumatic stress

There are currently a number of academic journals that are specifically dedicated to traumatic stress research and therefore constitute the core platforms through which knowledge production on traumatic stress is disseminated among researchers. However, the extent to which these domain-specific journals are contributing to the development of an internationally representative trauma knowledge base remains unclear. More than 80% of the world’s population resides in low and middle income countries (LMICs) (Patel & Kim, 2007), where the risk of experiencing both violence (Wolf, Gray, & Fazel, 2014) and unintentional injury (Chandran, Hyder, Peek-Asa, 2010) is inflated compared to high-income countries (HICs) due to elevated rates of poverty, inequality, conflict, and mass displacement. As the majority of trauma survivors in the world today reside in LMICs, these regions should be well represented in traumatic stress journals. Although articles on traumatic stress in LMICs are published in a range of mental health journals, traumatic stress journals are key curators of traumatic stress research and need to be geographically representative. The absence of LMIC-based traumatic stress research from domain-specific journals would signal a form of knowledge exclusion, reflecting what Fricker (2007) has termed ‘a structural identity prejudice in the collective hermeneutical resource’ (p. 155). The representative inclusion of LMIC studies in the central sites of knowledge production about traumatic stress therefore forms part of an epistemic justice agenda.

Yet, few studies have tracked this geographic representation over time. Bedard, Greif, and Buckley (2004) examined the author affiliations of over 13,000 trauma publications in the Published International Literature on Traumatic Stress (PILOTS) database between 1987 and 2001, and Figueira and colleagues (2007) surveyed trauma publications on the Web of Science database between 1983 and 2002. While both studies found a trend towards increasing internationalisation of trauma research over time, this mainly reflected a shift towards a greater proportion of publications from HICs other than the United States, rather than an increase in publications from LMICs. Both studies are now more than a decade out of date. More recently, Fodor and colleagues (2014) found that 87% of a random sample of traumatic stress articles on the PsychInfo, Pubmed, and PILOTS databases in the year 2012 (N = 1000) involved research in HICs and 88% had authors from HICs. While this indicates that traumatic stress research from LMICs may not be well represented in academic publications generally, it does not clarify the representation of LMICs in traumatic stress journals.

Theories and clinical conceptualisations of traumatic stress that have emerged from high-income settings tend to assume universality in trauma exposure, impact, and treatment (Fassin & Rechtman, 2009). However, they may have limited applicability to LMIC settings characterised by specific socio-economic constraints, different forms of trauma exposure than are typically found in HICs, and varying culturally based mental health norms and practices (Fodor et al., 2014; Hinton & Lewis-Fernández, 2011). Researchers, therefore, need to exercise caution when relying on etic, quantitative approaches to collecting data on traumatic stress in LMICs while neglecting emic approaches that qualitatively explore local trauma constructs and meanings (Rasmussen, Keatley, & Joscelyne, 2014). For example, it has been argued that there is an over reliance on posttraumatic stress disorder (PTSD) to conceptualise the psychological impact of traumatic events globally and that the endorsement of PTSD symptoms across different cultures and contexts does not necessarily confirm that PTSD is the sine qua non of traumatic stress responses universally (Hinton, Kredlow, Pich, Bui, & Hofmann, 2013; Nicolas, Wheatley, & Guillaume, 2015). Evidence to support this argument comes from several lines of research. First, the psychological effects of prolonged and continuous trauma exposure, common in contexts of civil conflict and mass displacement, extend beyond those captured in the PTSD diagnosis (Eagle & Kaminer, 2013). Second, culture-specific traumatic stress responses (such as somatic symptoms) have been documented in a number of LMICs in response to a variety of events (Rasmussen et al., 2014; Zheng & Gray, 2015). Third, even where PTSD symptoms are reported in LMIC contexts, their meaning and salience vary widely. For example, the prevalence of avoidance symptoms is particularly inconsistent across cultures (Hinton & Lewis-Fernández, 2011). It therefore remains unclear whether the construct of PTSD should be used to assess and identify traumatic stress universally.

Characterising the existing quantity and form of traumatic stress research from LMICs is a necessary first step in developing strategies to achieve a comprehensive understanding of the global phenomenon of trauma as it is experienced across diverse cultures and societies. This study aims to review the representation of LMICs in articles published in the leading traumatic stress journals over the 10-year period (2006-2015). The review aimed to identify the percentage of traumatic stress articles in domain-specific journals that have originated from LMICs as well as the geographical distribution of the samples, primary authors, and sources of funding within these articles, and the most frequently used methodologies in LMIC-based research published in domain-specific journals.

Method

Sample

A search was conducted on the 2015 release of the Thomson Reuters Journal Citation Reports (JCR; Thomson Reuters, 2015) for all journals with the term ‘trauma’ (also truncated as ‘trauma*’) in their titles. Once journals focusing on medical trauma (e.g., Journal of Neurotrauma) were excluded, there were six journals that focused specifically on traumatic stress and had been assigned an impact factor, indicating that in 2015, they were the leading traumatic stress journals with regard to global impact. One traumatic stress journal had not been assigned an impact factor and was not included in the study. While the usefulness of impact factors as a bibliometric measure has at times been contested, they remain a key indicator of international scientific impact (Gutierrez, Beall, & Forero, 2015). The JCR does not include predatory journals, and focuses on excluding pseudo-journals, and is therefore a valid index of the impact of genuine academic journals (Beninger, Beall, & Shumway, 2016).

The six journals included in the study were the European Journal of Psychotraumatology (EJPT); the Journal of Loss and Trauma (JOLT); the Journal of Trauma and Dissociation (JOTD); the Journal of Traumatic Stress (JOTS); Psychological Trauma: Theory, Research, Practice and Policy (PTTRPP); and Trauma, Violence and Abuse (TVA). The impact factor of these journals in the 2015 release of the JCR ranged between 0.70 and 2.67. PTTRPP and the EJPT were first published in 2008 and 2010, respectively; therefore, the analysis for these two journals began from their first publications. Articles published in JOLT for 2015 were not accessible to the researchers due to publisher restrictions and were thus excluded from the analysis. A total of 240 issues were included in the review.

Data collection

Data were collected during 2016, for the period 2006-2015. All articles in the six sampled journals were categorised according to whether they originate from a LMIC or a HIC. Categorisation was based on the World Bank’s (2015) classification of countries by income, which is based on a country’s gross national income (GNI). Countries classified by the World Bank as low-income, lower-middle-income, or upper-middle-income were all categorised as LMIC, and the remaining countries were in the World Bank’s high-income category. Articles were coded as LMIC or HIC based on the location of the study sample or the country of origin in the case of samples of immigrants or refugees. Articles that did not contain any sampling (such as editorials or theoretical articles) were categorised according to the affiliations of their first three authors as being HIC (the primary affiliation of all three first authors is an institution in a HIC), LMIC (the primary affiliation of all three authors is an institution in a LMIC), or mixed (the affiliations of the first three authors included institutions in both HICs and LMICs). All articles categorised as LMIC were then coded with regard to journal title; year of publication; type of article (editorial, research article, or brief report); country in which the sample resides or, in cases of immigrant or refugee samples, country of origin; the location of the primary institution of each of the first three authors (HIC or LMIC and the specific country); the location of sources of funding, where reported (HIC, LMIC, or mixed, and the specific country); and, where applicable, the method of data collection. Articles were coded by two of the authors (BR and LC), and interrater reliability was .76 or higher for all of the above coding categories, with an average interrater reliability of .94.

Data analysis

Frequency counts (ns and percentages) were generated for all categories using SPSS (Version 23.0).

Results

Representation of LMICs

During the 10-year period between 2006 and 2015, a total number of 2530 articles were published in the six peer-reviewed traumatic stress journals that were sampled in the study. Of these, 2283 (90.24%) were representative of HICs (i.e., the sample was based in or originated from a HIC or, where there was no sample, the first three authors were all based in institutions in HICs) while 247 (9.76%) of the articles were representative of LMICs (i.e., the sample was based in or originated from a LMIC or, where there was no sample, the first three authors were all based in institutions in LMICs). LMIC articles included full articles (88.87%; n = 219), brief reports (10.53%; n = 26), and editorials (0.81%; n = 2). As shown in Table 1, the percentage of LMICs articles was consistently low, ranging from 6.08% to 11.73% across the 10 years.

Table 1.

Frequency of LMIC articles in traumatic stress journals 2006-2015.

Year	Journals	Total no. of articles	No. of articles from LMICs	% of articles from LMICs
2006	JOLT, JOTD, JOTS, TVA	162	19	11.73
2007	JOLT, JOTD, JOTS, TVA	181	11	6.08
2008	JOLT, JOTD, JOTS, TVA, PTTRPP	166	12	7.23
2009	JOLT, JOTD, JOTS, TVA, PTTRPP	205	21	10.24
2010	JOLT, JOTD, JOTS, TVA, PTTRPP, EJPT	247	20	8.10
2011	JOLT, JOTD, JOTS, TVA, PTTRPP, EJPT	266	24	9.02
2012	JOLT, JOTD, JOTS, TVA, PTTRPP, EJPT	303	34	11.22
2013	JOLT, JOTD, JOTS, TVA, PTTRPP, EJPT	341	38	11.14
2014	JOLT, JOTD, JOTS, TVA, PTTRPP, EJPT	373	35	9.38
2015	JOTD, JOTS, TVA, PTTRPP, EJPT	286	33	11.54
Total		2530	247

JOLT = Journal of Loss and Trauma; JOTD = Journal of Trauma and Dissociation; JOTS = Journal of Traumatic Stress; TVA = Trauma, Violence and Abuse; PTTRPP = Psychological Trauma: Theory, Research, Practice and Policy; EJPT = European Journal of Psychotraumatology.

Location of LMIC research on traumatic stress

Sample country

Overall, 54 LMICs were represented. As summarised in Table 2, the most commonly studied country in the sample was Turkey, which constituted 8.91% (n = 22) of the LMIC sample, followed by China (8.10%; n = 20), the Palestinian territory of Gaza (5.26%; n = 13), and Rwanda (5.67%; n = 14). Half of all the represented countries (n = 27) only appeared once across the 10-year review period. Just under a quarter of the LMIC studies (n = 58; 23.48%) sampled refugees or immigrants who had relocated from their country of origin. In 23.08% (n = 57) of the studies sampled, either no participants were used (in 32 cases, the article was an editorial or a theoretical contribution) or no country was specified (e.g., some of the refugee or immigrant studies did not specify the country of origin, and in some studies, where the first three authors were from LMICs, the location of the sample was unspecified).

Table 2.

Number and frequency of articles from each LMIC represented in traumatic stress journals (2006-2015).

Countries	% of n	n
Turkey	8.91	22
China	8.10	20
Gaza	5.26	13
Rwanda	4.86	12
Democratic Republic of Congo	2.83	7
India	2.83	7
South Africa	2.83	7
Iraq (refugees)	2.43	6
Somalia (refugees)	2.43	6
West Bank	2.43	6
Cambodia (refugees)	2.02	5
Turkey (refugees)	2.02	5
Bosnia (refugees)	1.62	4
Burundi	1.62	4
China (refugees)	1.62	4
Colombia	1.62	4
Democratic Republic of Congo (refugees)	1.62	4
Kosovo	1.62	4
Mexico	1.62	4
Romania	1.62	4
Sierra Leone (refugees)	1.62	4
Uganda	1.62	4
Cambodia	1.21	3
Indonesia	1.21	3
Iran (refugees)	1.21	3
Lebanon	1.21	3
Sri Lanka	1.21	3
Tanzania	1.21	3
Vietnam (refugees)	1.21	3
Afghanistan (refugees)	0.81	2
Angola (refugees)	0.81	2
Bosnia	0.81	2
El Salvador	0.81	2
Ethiopia	0.81	2
Ethiopia (refugees)	0.81	2
Ghana (refugees)	0.81	2
Guinea (refugees)	0.81	2
Iran	0.81	2
Iraq	0.81	2
Kenya	0.81	2
Kosovo (refugees)	0.81	2
Rwanda (refugees)	0.81	2
Sri Lanka (refugees)	0.81	2
Sudan (refugees)	0.81	2
Thailand	0.81	2
Vietnam	0.81	2
Afghanistan	0.40	1
Brazil	0.40	1
Cameroon (refugees)	0.40	1
Côte d’Ivoire	0.40	1
Côte d’Ivoire (refugees)	0.40	1
Gambia (refugees)	0.40	1
Gaza (immigrants)	0.40	1
Gaza (refugees)	0.40	1
Georgia	0.40	1
Haiti	0.40	1
Herzegovina	0.40	1
Iraq (immigrants)	0.40	1
Jordan	0.40	1
Kazakhstan (refugees)	0.40	1
Lebanon (immigrants)	0.40	1
Liberia (refugees)	0.40	1
Malaysia	0.40	1
Morocco (refugees)	0.40	1
Mozambique	0.40	1
Nigeria (refugees)	0.40	1
North Korea (refugees)	0.40	1
Pakistan	0.40	1
Pakistan (refugees)	0.40	1
Peru	0.40	1
Serbia	0.40	1
Serbia (refugees)	0.40	1
Sierra Leone	0.40	1
Somalia	0.40	1
Syria	0.40	1
Syria (refugees)	0.40	1
Timor-L’este	0.40	1
West Bank (immigrants)	0.40	1
No sample included	13.77	34
LMIC authors but country not specified	1.21	3
Country of origin for immigrant sample not specified	2.43	6
Country of origin for refugee sample not specified	5.67	14
Total		300^a

There were instances where articles had more than one sample country.

Sample countries were also classified by region according to the World Bank’s (2015) regional classification system. The most commonly studied region was Sub-Saharan Africa (30.77%; n = 76 articles), followed by Europe and Central Asia (19.43%; n = 48), the Middle East and North Africa (17.81%; n = 44), and East Asia and the Pacific (18.22%; n = 45), which were fairly equally represented. Articles from South Asia (6.88%; n = 17) and Latin America and the Caribbean (5.26%; n = 13) were rare. Almost a quarter of articles (23.98%; n = 57) could not be classified by region due to either not using any participants or failing to specify the country of origin in the case of immigrant or refugee samples currently residing in HICs. The combined percentage for regions exceeds 100% as some articles sampled across more than one region.

Host country for immigrant and refugee samples

In 44.83% (n = 26) of the articles that used refugee or immigrant samples, the United States was the host country in which the research was conducted. Following well behind were, in order, Germany (10.34%; n = 6), the Netherlands (8.62%; n = 5), Australia and Uganda each with 6.90% (n = 4), the United Kingdom and Israel each with 5.17% (n = 3), and Switzerland (n = 2; 3.45%). Egypt, France, Hungary, Norway, South Korea, and Thailand were each only represented once as host countries.

Location of primary authors

In over half of the sample of articles from LMICs (56.28%; n = 139), the primary author team (up to the first three authors) were all affiliated with institutions in countries from HICs; in 20.24% (n = 50) of the sample, the primary authors were all affiliated with institutions in LMICs; and in 23.48% (n = 53) of the sample, the primary author team was affiliated with institutions in both LMICs and HICs.

Examining each of the primary authors separately, in 69.64% (n = 172) of the sample, the first author was exclusively affiliated with institution(s) in HICs; in just over a quarter of the sample (27.53%; n = 68), the first author was exclusively affiliated with institution(s) in LMICs; and in 2.83% (n = 7) of the sample, the first author was affiliated with institutions in both a LMIC and a HIC. As summarised in Table 3, the most commonly recorded first author country was the United States, which constituted 33.60% (n = 84) of the sample, followed by Germany (11.34%; n = 28) and Turkey (9.31%; n = 23). Where articles had more than one author, a similar pattern was found for the second authors (exclusively affiliated with institutions in HICs in 59.11% of cases) and third authors (47.77% of cases), with the same pattern of country of affiliation.

Table 3.

Affiliation countries of first three authors of LMIC articles in traumatic stress journals (2006-2015).

Countries	First Author		Second Author		Third Author
Countries	% of n	n	% of n	n	% of n	n
The United States	33.60	84	29.55	74	22.27	55
Germany	11.34	28	10.53	26	8.50	21
Turkey	9.31	23	8.50	21	6.07	15
China	6.07	15	4.86	12	2.83	7
The United Kingdom	4.86	12	4.05	10	3.64	9
The Netherlands	4.45	11	3.64	9	3.24	8
Australia	3.64	9	2.83	7	2.83	7
Switzerland	2.83	7	2.43	6	1.21	3
Israel	2.83	7	3.64	9	2.02	5
South Africa	2.43	6	2.43	6	2.02	5
India	2.02	5	1.62	4
Norway	2.02	5	1.62	4	0.81	2
Italy	1.21	3	0.81	2	0.81	2
Lebanon	1.21	3	1.62	3	0.81	2
Malaysia	1.21	3	0.40	1	0.40	1
Romania	1.21	3	0.81	2
Belgium	0.81	2			0.40	1
Brazil	0.81	2	0.40	1	0.81	2
Canada	0.81	2	0.81	2	0.40	1
France	0.81	2	0.40	1
Hong Kong	0.81	2	0.40	1	0.40	1
Hungary	0.81	2	0.40	1
Iran	0.81	2
Kenya	0.81	2	0.40	1	0.40	1
Bosnia	0.40	1	0.40	1
Burundi	0.40	1	0.40	1	0.81	2
Denmark	0.40	1	1.21	3	0.81	2
Finland	0.40	1	0.40	1	0.40	1
Gaza	0.40	1	0.40	1	1.62	4
Georgia	0.40	1	0.40	1	0.40	1
Herzegovina	0.40	1
Ireland	0.40	1
Jordan	0.40	1
Mexico	0.40	1	0.40	1	0.81	2
Nigeria	0.40	1
Rwanda	0.40	1	0.81	2	1.21	3
South Korea	0.40	1	0.40	1	0.40	1
Spain	0.40	1	0.40	1	0.40	1
Sri Lanka	0.40	1	0.81	2
Uganda	0.40	1			0.81	2
West Bank	0.40	1	0.81	2
Thailand			1.21	3	0.40	1
Cambodia			0.40	1
Colombia			0.40	1	0.40	1
Croatia			0.40	1	0.40	1
Democratic Republic of Congo					0.40	1
Egypt			0.40	1	0.40	1
El Salvador			0.40	1
Haiti			0.40	1	0.40	1
Indonesia					0.40	1
Iraq					0.40	1
Jamaica			0.40	1
New Zealand			0.40	1
Serbia					0.40	1
United Arab Emirates			0.40	1
NA^a			7.69	19	30.36	75
Total		257^b		251^b		251^b

LMIC = low and middle income country.

Instances where cases did not have a second author or third author.

There were instances where authors were affiliated with institutions in more than one country.

Source of funding

In 55.87% (n = 138) of the sample, the source of funding was not specified. Of the sample that did report the source of funding, the vast majority of studies (73.39%; n = 80) were funded by sources in HICs, while 19.27% (n = 21) received funding from sources in LMICs and the remainder (7.34%; n = 8) received funding from both. The United States provided the funding in almost half of all cases where funding source was reported (45.87%; n = 50), followed by Germany (5.67%; n = 14), China (3.24%; n = 8), the United Kingdom (2.83%; n = 7), and South Africa (2.43%; n = 6).

Methods of data collection

The most commonly used method of data collection in the LMIC articles was the structured symptom questionnaire (71.66%; n = 177) and PTSD was measured in over two-thirds (70.06%; n = 124) of the articles that used a structured symptom questionnaire. Other types of structured questionnaires were used in just over half the LMIC articles (54.25%; n = 134), for example, assessing social support or life events. A minority of studies (14.17%; n = 35) used qualitative interview techniques for individuals or focus groups. The use of mixed methods (structured questionnaires plus interviews) was rare across the LMIC studies (8.91%; n = 22).

Discussion

Reviews of geographic representation in academic journals can contribute towards exploring possible inequities in knowledge production. The current study found that just under one-tenth of all articles published in the leading traumatic stress journals over the period 2006–2015 represented knowledge produced in or by LMICs. This replicates the 10/90 divide in mental health research that was reported by Saxena, Paraje, Sharan, Karam, and Sadana (2006) over a decade ago. The low percentage of LMIC articles was consistent across all 10 years under review, never exceeding 11% of articles published per year, and in some years dropping to as low as 6%. There was a very slight increase in the percentage of LMIC articles between the first 5 years (an average of 8.68%) and the last 5 years (an average of 10.46%), which may indicate a slow upwards trend. Overall, however, knowledge production about traumatic stress in the most globally influential journals dedicated to this topic remains massively geographically skewed. Despite the fact that the majority of the world’s population resides in LMICs, and that these regions experience higher rates of both violence exposure and unintentional injury than HICs, the experiences and needs of trauma survivors in LMICs remain enormously under-represented. These findings closely resemble those of Fodor et al.’s (2014) review of a random sample of all traumatic stress articles published in 2012. The results of the current study confirm that this underrepresentation was a consistent trend across a decade of research published in the leading traumatic stress journals.

Africa was the best represented LMIC region in the sample. While this is somewhat surprising, given that this region is particularly scarce in research resources, it is also apparent from our findings that most LMIC research to date has been conducted and funded by Northern researchers. The most commonly studied African countries were Rwanda, Democratic Republic of Congo, and South Africa, which all experienced protracted political conflicts and post-conflict transitions that attracted international interest during the 1990s and early 2000s, and thus may have come to the attention of trauma researchers in HICs. Turkey and China were the most commonly studied individual countries in the sample. Both are upper-middle-income countries and thus among the better resourced of the LMIC countries. Furthermore, the Chinese government has deliberately focused on increasing scientific research output to improve their economic growth (Xie, Zhang, & Lai, 2014).

Many LMICs were only represented once over the 10-year period and therefore have had very limited participation in developing the global knowledge base of traumatic stress. Furthermore, there has been a notable paucity of research on traumatic stress from South Asia and from Latin America and the Caribbean. It is possible that trauma studies from these regions are more likely to be published in non-English journals given that language might be a barrier for researchers residing there.

The majority of LMIC articles were produced by a group of primary authors from HICs. Researchers from the United States in particular have had a dominant role in knowledge production about traumatic stress in LMICs in the past decade: a third of all articles examining traumatic stress in LMICs were led by first authors based in the United States. This likely reflects the abundant research funding capacity of the United States across all scientific fields, compared to most other countries, and the consequent reliance on United States research resources to capacitate research in regions where research resources are constrained (Lansang & Dennis, 2004). While such sharing of funding and research expertise is certainly of benefit to LMICs, the nature of these partnerships can skew modes and content of knowledge production. If HIC researchers, as the principal investigators and primary authors, unilaterally determine the research questions and methods for trauma research based on their existing conceptualisations, while LMIC partners are responsible largely for providing research sites and overseeing fieldwork, there is a danger of simply reproducing Northern knowledges (Boshoff, 2009; Hodes & Morrell, 2016). The reproduction of Northern conceptualisations of trauma is evident in the content of LMIC research in traumatic stress journals over the past decade: most studies have utilised structured symptom questionnaires for PTSD, while there has been little use of more emic, explorative methods to characterise the psychological impact of trauma beyond PTSD or beyond Diagnostic and Statistical Manual of Mental Disorders (DSM) categories more generally. While PTSD symptoms may well be endorsed when asked for, it remains unclear whether this diagnosis offers the best ‘fit’ to capture the complexity of trauma impact across diverse economic and cultural settings, particularly when trauma exposure may be part of daily life and embedded in multiple adversities (Eagle & Kaminer, 2013; Nicolas et al., 2015). An emerging body of emic research in LMICs suggests that PTSD has only limited congruence with local traumatic stress concepts across different cultural settings (Rasmussen et al., 2014); however, the results of the current study indicate that much of this research has been published outside of traumatic stress journals. This may result from the epistemological biases of traumatic stress journals, the publication preferences of LMIC trauma researchers, or both.

Conclusion

This study has a number of limitations. The review was confined to articles published in six journals that focus specifically on traumatic stress and have been assigned an impact factor. Traumatic stress research is also published in general mental health journals, and it is possible that much LMIC-based trauma research is contained in these sources. However, Fodor et al.’s (2014) bibliometric review found a similar under-representation of trauma research from lower income settings across a number of journal databases, suggesting that this is a more general pattern not confined to domain-specific journals. Furthermore, it is critical to address issues of geographic representation in the domain-specific journals that have the greatest impact on a specific field. The underrepresentation of LMIC research in leading traumatic stress journals suggests a form of knowledge exclusion that needs to be recognised and addressed (Keet, 2014).

Another potential limitation is that articles published in JOLT for 2015 were not accessible due to publisher restrictions and were thus excluded from the analysis. However, there is no reason to believe that a different trend would have emerged should articles from this 1 year have been included.

The review also only provides an overview of the quantity and content of LMIC publications on trauma in the journals sampled; it has not examined trends in the findings of these studies or the degree of impact of these publications on the international trauma literature (e.g., through citation analysis, quality, and usefulness of the research). Future research addressing these issues will allow better characterisation of the global contribution of LMIC research, as low representation does not necessarily equate to low impact or to an absence of clear patterns of findings.

While bearing these limitations in mind, the persistent geographic bias in knowledge production about traumatic stress remains starkly apparent from this review: the regions where most trauma survivors reside continue to have little representation in the leading traumatic stress journals. This points to something of an epistemological crisis in the traumatic stress field and urgent, concrete remedies are needed.

Capacitating LMIC trauma researchers with regard to funding, training, facilities, and journal access, so that they can lead traumatic stress research from their own regions, remains a priority. This will enable the development of research that is more closely embedded in local contexts of trauma and that provides more contextualised knowledge about traumatic stress to supplement more universal aspects. There does seem to be some movement towards more contributions by LMIC researchers: for 20% of articles in this study, all the primary authors were based in LMICs, while just under a quarter were produced through LMIC–HIC collaborations. Furthermore, China and South Africa are among the five countries that have most frequently funded LMIC trauma research, although their contribution remains well below that of the United States. The inclusion of mental health in the United Nations Sustainable Development Agenda in 2015 may catalyse increased investment in mental health research in lower income countries, as it has done previously for research on health conditions such as HIV/AIDS. Ongoing LMIC–HIC collaborations will enable greater sharing of inequitable research resources and cross-fertilisation of knowledge; however, it is critical that LMIC partners should have a leading role in developing the research agenda and research methodologies for trauma research in LMIC settings.

International traumatic stress organisations can also play a role; for example, in 2010, the International Society for Traumatic Stress Studies (ISTSS) adopted the ‘Global Initiative’ to facilitate global collaborations among trauma researchers (Schnyder, 2013). It is vital to ensure that such partnerships allow for the generation of new, contextualised conceptualisations of traumatic stress, rather than only replicating those developed in HICs. Furthermore, traumatic stress journals could consider strategies to promote the inclusion of LMIC researchers in systems of knowledge production, such as providing more open access to journals and considering ways to facilitate language translation support.

Finally, it is recommended that reviews similar to the current study be conducted every 10 years, so that trends as well as gaps in the knowledge base on traumatic stress in LMICs can be tracked.

The above strategies will contribute to the development of an internationally representative traumatic stress research base that is truly responsive to the needs of trauma survivors in a multiplicity of contexts.

Footnotes

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Bedard

Greif

Buckley

(2004). International publication trends in traumatic stress literature. Journal of Traumatic Stress, 17, 97–101.

Beninger

P. G.

Beall

Shumway

S. E.

(2016). Debasing the currency of science: The growing menace of predatory open access journals. Journal of Shellfish Research, 35(1), 1–5.

Boshoff

(2009). Neo-colonialism and research collaboration in Central Africa. Scientometrics, 81, 413–434.

Chandran

Hyder

Peek-Asa

(2010). The global burden of unintentional injuries and an agenda for progress. Epidemiologic Reviews, 32, 110–120.

Eagle

Kaminer

(2013). Continuous traumatic stress: Expanding the lexicon of traumatic stress. Peace and Conflict: Journal of Peace Psychology, 19, 85–99.

Fassin

Rechtman

(2009). The empire of trauma: An inquiry into the condition of victimhood. Princeton, NJ: Princeton University Press.

Figueira

Luz

Braga

Cabizuca

Coutinho

Mendlowicz

(2007). The increasing internationalization of mainstream posttraumatic stress disorder research: A bibliometric study. Journal of Traumatic Stress, 20, 89–95.

Fodor

Unterhitzenberger

Chou

Kartal

Leistner

Milosvljevic

. . .Alisic

(2014). Is traumatic stress research global? A bibliometric analysis. European Journal of Psychotraumatology, 5(23269), 1–7.

Fricker

(2007). Epistemic injustice: Power and the ethics of knowing. Oxford, UK: Oxford University Press.

10.

Gutierrez

F. R. S.

Beall

Forero

D. A.

(2015). Spurious alternative impact factors: The scale of the problem from an academic perspective. BioEssays, 37, 474–476.

11.

Hinton

Kredlow

Picj

Bui

Hoffman

(2013). The relationship of PTSD to key somatic complaints and cultural syndromes among Cambodian refugees attending a psychiatric clinic: The Cambodian Somatic Symptom and Syndrome Inventory (CSSI). Transcultural Psychiatry, 50, 347–370.

12.

Hinton

Lewis-Fernández

(2011). The cross-cultural validity of posttraumatic stress disorder: Implications for DSM 5. Depression and Anxiety, 28, 783–801.

13.

Hodes

Morrell

(2016). South African social science in the global HIV/AIDS knowledge domain (CSSR Working Paper No. 382). Cape Town, South Africa: Centre for Social Science Research. Retrieved from http://www.cssr.uct.ac.za/pub/wp/382

14.

Keet

(2014). Epistemic ‘othering’ and the decolonisation of knowledge. Africa Insight, 44, 23–37.

15.

Lansang

M. A.

Dennis

. (2004). Building capacity in health research in the developing world. Bulletin of the World Health Organisation, 82(10), 764–770.

16.

Nicolas

Wheatley

Guillaume

(2015). Does one trauma fit all? Exploring the relevance of PTSD across cultures. International Journal of Culture and Mental Health, 8, 34–45. doi:10.1080/17542863.2014.892519

17.

Patel

Kim

(2007). Contribution of low- and middle-income countries to research published in leading general psychiatry journals, 2002 – 2004. The British Journal of Psychiatry, 190, 77–78.

18.

Rasmussen

Keatley

Joscelyne

(2014). Posttraumatic stress in emergency settings outside North America and Europe: A review of the emic literature. Social Science & Medicine, 109, 44–54.

19.

Saxena

Paraje

Sharan

Karam

Sadana

(2006). The 10/90 divide in mental health research: Trends over a 10-year period. The British Journal of Psychiatry, 188, 81–82.

20.

Schnyder

(2013). Trauma is a global issue. European Journal of Psychotraumatology, 4(20419), 1–2.

21.

Thomson Reuters. (2015). Journal citation reports: Journals in the 2015 release of JCR. Retrieved on November 27, 2015, from http://about.jcr.incites.thomsonreuters.com/full-titles-2015.pdf

22.

Wolf

Gray

Fazel

(2014). Violence as a public health problem: An ecological study of 169 countries. Social Science & Medicine, 104, 220–227.

23.

The World Bank. (2015). Country and lending groups. Retrieved from http://data.worldbank.org/about/country-and-lending-groups

24.

Xie

Zhang

Lai

(2014). China’s rise as a major contributor to science and technology. Proceedings of the National Academy of Sciences of the United States of America, 111(26), 4–6.

25.

Zheng

Gray

(2015). Posttraumatic coping and distress: An evaluation of Western conceptualization of trauma and its applicability to Chinese culture. Journal of Cross-Cultural Psychology, 46, 723–736.