Public health journals’ requirements for authors to disclose funding and conflicts of interest: a cross-sectional study

For this study, we adopted the Institute of Medicine’s definition of COI, provided above [1]. We defined a COI policy as one that required COI disclosure by at least one of the authors. We used a cross-sectional design to review the journals’ instructions for authors on disclosing COI and funding.

The study focused on journals categorized as “Public, Environmental & Occupational Health”, in Social Science Citation Index (SCIE) edition of the Journal Citation Reports (JCR) of the Web of Science database (https://​jcr.​incites.​thomsonreuters.​com). We included in our study all of the 173 journals listed in the SCIE in January 2017. We did not limit our selection by language of publication.

We used three sources to abstract relevant data:

Three teams of two reviewers collected data in a duplicate and independent manner using a standardized and pilot tested data abstraction form (Fig. 1). They compared results and resolved disagreements through discussion or with the help of a third reviewer (MBH or EAA) when needed. All reviewers conducted calibration exercises prior to data collection. We collected and managed data using the Research Electronic Data Capture (REDCap) tools hosted at the American University of Beirut [19]. REDCap is a secure, web-based application for supporting data capture for research studies.

We collected data regarding the characteristics of the journals, the COI disclosure policies, and funding policies (the processes were limited to explicit journal policies):

General characteristics of the journal:

Funding reporting policy:

COI disclosure policy:

Journal requirements for financial COI disclosure:

Journal requirements for non-financial COI disclosure:

We checked for any missing data or erroneous entries. For the descriptive analyses of journals’ characteristics, we used frequencies and percentages for categorical variables. We conducted regression analyses to identify relationships between [1] existence of a funding disclosure policy, and explicit requirement for disclosure of COI and the following independent variables: COPE, ICMJE and impact factor. None of these analyses was significant. We conducted analysis using SPSS version 23.0 (SPSS INC., Chicago, IL, USA).

Table 1 provides the results of the general characteristics of the 173 included journals. Twenty-one percent of those journals were members of the ICMJE, 67% were members of COPE, and 75% were affiliated to a professional organization. The most common publisher was Elsevier (15%) associated with public health journals.

Out of the 173 included journals, 155 (90%) had a policy regarding the reporting of information regarding funding of the research. Table 2 summarizes the characteristics of those policies. Ninety-two percent of the policies were specific to the journal. No policy explicitly included procedures to deal with non-reporting or under-reporting of funding. The most frequent two methods for reporting of funding information were online as part of the submission process (74%) and in the body of the manuscript (63%).

Table 3 provides the elements of funding required by the policies of the 155 journals requiring the reporting of funding information. Respectively, 88% and 87% of policies did not require the reporting of the phase(s) of the research to which the funding applied, or the types of funding sources. Twelve percent of policies required the disclosure of whether a not-for-profit funding source is supported by a for-profit entity. Forty-three percent of policies did not specify the types of funding to be reported. Monetary support was the most frequent type of funding the policies required to be reported (53%). However, none of these required the reporting of the amount of monetary support.

Out of the 155 journal policies for reporting of funding information, 36 (23%) required reporting on the role(s) of funder. Figure 2 shows the percentages of policies requiring the disclosure of the role of the funder for each of the steps of the research process. The most frequently required steps were data analysis, study design, and data collection (83%, 81% and 75% respectively). Four policies clearly requested authors to report on having independent and full control in at least one of the steps of the research process: study design (n = 1), data collection (n = 1), data management (n = 2), data analysis (n = 1), decision to submit for publication (n = 2), access to the data (n = 1), and freedom to conduct the research (n = 1) (Table 4).

Out of the 173 eligible journals, 163 (94%) had a disclosure policy regarding individual authors’ COI. Table 5 summarizes the characteristics of those policies. Seventy-seven percent of the policies were specific to the journals while 18% were specific to both the publishers and the journals. The majority of policies required the disclosure of COI at the time of manuscript submission (98%), in a form of narrative statement (91%), during the online submission process (59%), or in the body of the manuscript (58%). COI disclosures were stated to be made available to the public within the published manuscript in 50% of policies. Seven percent of the policies mentioned the possibility of verifying COI disclosures following revision or acceptance. While 11% of journals would contact authors for more details on disclosed COI, 19% of policies describe the effect of disclosures on the editorial process (e.g., editors and reviewers may decide to disqualify review from publication or reject the manuscript); and 10% clearly mention a potential impact of not disclosing COI on the editorial process (e.g., editors may correct, reject or retract the publication). No policy described procedures to verify authors’ COI disclosures.

Table 6 describes the requirements for financial COI disclosure of the 163 journals with a COI disclosure policy. Out of journals having COI disclosure policies, 25% and 10% respectively require the disclosure of the financial relationships of authors’ institutions and family members. The most commonly required forms of financial COI are as stock ownership (51%), grant/research support (45%), direct employment (45%) and serving as an advisor, consultant, or public advocate (43%). Eighty-seven percent of the policies do not specify the duration for which to disclose the financial COI. Among those in which it was necessary for the previous number of months to be reported (2%), 90% specified a duration of 36 months. Although 39% of journals’ policies explicitly required reporting the source of financial benefit, only one required reporting of the monetary value.

Ninety-two of the 163 COI journals’ policies (56%) required the disclosure of at least one form of non-financial COI. The top 4 descriptors of non-financial COI were: “personal relationship” (33%), “personal” (20%), “academic competition” and “non-financial COI” (14%) (Table 7).

Of 173 public health journals, 155 (90%) had a policy for reporting study funding information. Out of these, a majority did not require reporting of the phase of the study for which funding was received (88%), nor the types of funding sources (87%). Of the 173 journals, 163 (94%) had a policy requiring disclosure of authors’ COI. However, the majority of these journals did not require financial conflicts of interest disclosures relating to institutions (75%) nor to the author’s family members (90%) while 56% required the disclosure of at least one form of non-financial COI.

This is the first study to investigate, using systematic methodology, public health journal policies for reporting of funding information. Our study assessed this topic in a detailed way (e.g., potential effect of disclosures and incomplete or imprecise COI disclosures on the editorial process). One limitation is that we might have missed funding and COI policies that journals implemented during later steps of the editorial process, especially that we did not contact the journals.

Shawwa et al. reported that 99% of Core Clinical Journals had a policy on COI disclosure in comparison to 94% of public health journals [15]. Clinical journals were more likely to require specifying financial and non-financial COIs (100% and 57% respectively) compared to public health journals (89% and 18% respectively) [20]. Also, a higher percentage of clinical journals, compared to public health journals, specified measures to deal with inaccurate or incomplete disclosures statements (23% vs. 10%) [20]. Furthermore, no public health journal reported procedures to verify the accuracy or completeness of authors’ disclosed COIs, compared to 17% of clinical journals [20].

The issues of conflicts and their potential influence on what articles are published, their results, and the framing of the conclusions, suggest that editors and publishers of public health journals should reevaluate their policies regarding reporting of funding and disclosure of authors’ COI. Such a reevaluation should be based on the principles related to honesty and objectivity of the editorial process [21].

A comprehensive policy for reporting funding should require the specification of the phase of research study to which funding applied (planning, conduct or reporting), the specific type of funding source (governmental or not, for profit or not) as well as the type of funding (monetary support, provision of supplies, assistance in writing) and its amount.

While improvement in editorial policies can include the verification of financial COI of authors’ institutions and their family members, findings in Table 7 also suggest that policies are limited in their disclosure of non-financial COIs. These might include intellectual COI related to authorship, academic affiliations, institutional, personnel or professional COIs, political and religious affiliations.

Ideally, journals would implement a systematic approach with acceptable validity and reliability that compels authors to reveal financial and non-financial interests underlying their research work. Although a significant fraction of public health journals included policies offering considerable description on the study’s funding and authors’ own COIs, COIs related to their family members and affiliated departments are seldom addressed. There is a need for well-defined operational definitions to simplify both the disclosure of COI and funding information by authors and their verification assessment and control by the editorial board. Future research should also emphasize credible evidence for the interventions and guidelines to manage funding information and COI.