This study examined current practices in case reports and other studies publishing identifying information, with the hope of identifying issues around obtaining consent.
Obtaining consent or consent forms was considered necessary by approximately 40% of journals under review, and the submission of consent forms was required by 30% of the journals. Approximately 10% provided their own consent forms. While there may be differences between specialties and publishing groups in each field, these rates are higher than those reported in a previous study , suggesting an increasing interest in case reports and studies publishing identifying information.
The frequency of statements regarding obtaining consent varied depending on the specialist field. Statements were found most frequently in pediatrics journals (58.0%), and least frequently in genetics and heredity journals (21.1%). Similarly, statements regarding substitute consent were found most frequently in pediatrics journals (29.0%), and least frequently in genetics and heredity journals (10.9%). Specific mentions of pedigrees in instructions for authors were found least frequently in genetics and heredity journals (2.0%). With genetics being a field in which psychological and financial (insurance and employment) risks have been highlighted in discussions of the need to consider the influence on subjects and their families of publication , special attention is required when obtaining consent for the publication of pedigrees. Similarly, in the field of pediatrics, in which it has been considered more appropriate, but is actually difficult to obtain consent when subjects grow up sufficiently to make their own judgments , instructions for considering the influences of publication on a child’s future were observed in only one journal. To address this problem, some journal editors responding to the questionnaire pointed out the need to store consent forms for a longer period of time. There were also statements regarding mental impairment in some journals; in a previous study , substitute consent was not regarded as appropriate for subjects with impaired judgment. Despite the need for sensitivity in this area, few of the journals covered by this study clearly presented their policies.
In the consent forms provided by journals, a large number of statements were found regarding subjects’ signatures, journal names, non-paper-based publication, and how it would not be possible to completely anonymize subject data. These were regarded as necessary contents. The ICMJE  recommends that authors disclose to their subjects whether the manuscript includes identifiable personal data available online, and 26 out of 27 journals’ consent forms adhered to this requirement. As described in the Declaration of Helsinki , which requires that potential subjects must be informed of any potential risks involved in taking part in a study, more than 80% of consent forms described the risk of the possible identification of the subject. However, statements about the study objective, subjects’ right to refuse consent, and provisions for withdrawal of consent were found infrequently, although these points were considered to be important by editors responding to the questionnaire. Thus, there is a clear inconsistency between the contents of existing consent forms and opinions expressed by survey respondents about what those forms should contain. The study objective, subjects’ unconditional right to refuse consent, and withdrawal of consent at any time without reprisal, were considered as necessary contents for informed consent in the Declaration of Helsinki . Similarly, in studies that publish identifiable personal information, explanation of the study’s scientific or medical significance may determine whether a subject consents to their information being published. Giving assurance of subjects’ right to refuse consent may also give rise to the need for the subject to be able to give consent discretionally. As some journals provided information about withdrawal of consent, providing this information on withdrawal of consent may affect a subject’s decision to give consent, and the need to add this to the forms is suggested.
Problems with consent forms were also pointed out in some previous studies. To avoid the necessity of repeatedly obtaining consent when manuscripts are rejected by journals accepting only their original consent form, the development of common forms has been proposed [11, 12, 19]. However, a large number of editors responding to the questionnaire said that their journals required the submission of consent forms with manuscripts, and editors from eight of the journals providing a consent form stated that they did not accept other forms. Only a few of the journals reviewed for this study provided multilingual consent forms. Difficulty in using such forms by non-native English speakers was pointed out in a previous study . It is suggested therefore that several issues remain unresolved.
Although the ICMJE  recommends that journals should establish their own policies with local legal guidance, they also recommend that consent forms should be archived with the journal, the authors, or both. The criticism that subjects’ identifiable personal data are transferred to the journal by the submission of signed consent forms has been raised . On this issue, the ICMJE suggests that, to protect patient confidentiality, journals may decide that the author archives the consent form and provides the journal with a written statement that they have received and archived written patient consent. In our study, approximately 30% of journals required submission of a signed consent form, and 56.5% of journals stored submitted consent forms indefinitely. In their report on patient confidentiality and publishing, Bal et al  state that consent forms should be explicit that identifiable personal information will be transferred to a journal for publication. One journal required statements about the submission of signed consent forms to the publisher as part of the consent form. However, a statement about submitting signed consent forms to the publisher was not found in half of journals’ consent forms reviewed in our study, despite including statements about submission of consent forms in their instructions for authors. We suggest that it is possible that subjects may not notice or be made sufficiently aware that their information will be transferred to a journal for publication.
The questionnaire results indicated that the most common reason for requiring such a submission was to confirm that authors had obtained consent. To confirm the contents of the consent form was listed as the reason in fewer than half the journals reviewed in our study. In current practice, archiving signed consent forms is recommended by the ICMJE , and it is expected that the journal producing its own consent forms and requiring submission of signed consent forms is a way to ensure research ethics are complied with. In practice however, the issue of consent is between the author and the subject, and whether the publisher should take responsibility for subject consent, beyond supporting author’s ethical practice, requires further discussion. Although it may be viewed that author’s declaration to journals that they have obtained consent is sufficient , it is necessary to discuss the roles of editors and researchers in guaranteeing ethics.
Regarding points to consider when obtaining consent, statements allowing subjects to review manuscripts before publication or regarding non-paper-based publications were occasionally noted. Statements advising consideration of the effects on subjects of publishing identifying information were rarely observed. Examples of the difficulty of maintaining anonymity appeared in some consent forms; however, it is questionable whether such examples alone are enough to lead subjects to consider the effects on themselves. Case reports and studies publishing identifying information mainly differ from major clinical and epidemiological studies in the possibility of individual identification and consequent effects on subjects and their families. Studies have pointed out the necessity of explaining benefits and risks in detail when obtaining consent  and the presence of uncontrollable risks after consenting (inappropriate use of information and media follow-up) . As such effects depend on subjects’ backgrounds, it may be necessary for authors or researchers to implement different consent procedures appropriate to each case.
Difficulty has been reported in obtaining consent from individuals with impaired judgment, such as children and patients with mental disorders, and in cases of medical malpractice, unavailability of subjects, their deaths, and possible interference with the physician-patient relationship [10, 14]. In line with this, obtaining consent has not been regarded as absolutely necessary on some occasions [15, 16], while other researchers have insisted on the necessity of considering it even when it is difficult  and of giving importance to the decision-making process . It has also been pointed out that relatives’ consent should be obtained after a subject’s death . On analysis of instructions for authors, statements that identifying information should be deleted when consent is not available were found. In fact, obtaining consent is needed because complete anonymity is impossible. Therefore it is necessary to include only scientifically necessary information and maintain anonymity of identifiable information , and furthermore, to determine whether to obtain consent or to tolerate publishing without consent despite potential identifiability. Although these points interact with each other, further discussion may be necessary to focus separately on issues such as the content of consent and points to consider when obtaining it, and the appropriateness of obtaining or not obtaining consent.
This study had several limitations. The review of instructions for authors studied only those texts and, to a limited degree, linked documents, and therefore may not have reflected journals’ policies as described elsewhere, although it is likely to have discovered the most important items. In addition, the questionnaire survey of journal editors was not anonymous, to allow comparison of editors’ responses with their journal’s instructions for authors, observation of any differences between fields. However, between-field comparisons could not be accomplished because the response rate was low. Furthermore, it may be difficult to generalize its results, considering the possible influences of the questionnaire’s design on editors’ responses and response rate. The limited number of journals under study also did not necessarily represent general tendencies among academic journals. However, these surveys focused on key fields in which consent-related problems have been pointed out, while taking diversity into consideration.