Although publication of research results is a widely accepted component of ethical research, in actual practice it is often neglected or substantially delayed (7). For example, one study found that only 46% of US NIH-funded trials were published in a peer-reviewed, MEDLINE-indexed journal within 30 months of completion (8). A similar finding was discovered of vaccine trials conducted in multiple countries. Manzoli et al. (9) found that "the proportion of trials published 12, 24, 36, and 48 months after completion was 12%, 29%, 53%, and 73%, respectively" (9). Similarly, a study at a university in Germany found that only half of the studies approved by the ethics committee at the institution resulted in subsequent publication (10). Non-publication of results wastes resources and denies the field of medicine the opportunity to learn from the outcomes.

One particular area of challenge is publication of negative and inconclusive results. Multiple studies have found that trials with negative results are published significantly less often and significantly less quickly than trials with positive results (11, 12). This type of publication bias can have a direct effect on patient care. An analysis in the field of psychiatry compared published literature against US FDA reviews of research involving antidepressant agents. Based on the published literature it appeared that 94% of studies were positive but the FDA analysis indicated that only 51% were positive (13). This discrepancy was because most of the negative results were not published or were published in a way that conveyed a positive outcome. The authors of the analysis concluded that the actual efficacy of the antidepressant drug class is less than would be expected based on published literature alone (13).

Higher publication of positive results is a global issue. For example, one study conducted at a hospital in Spain followed the publication outcomes of all drug-evaluating clinical trials approved by the ethics committee of the institution. The study found that trials with positive results were published significantly more often and significantly more quickly than trials with negative results (11). Additionally, a study that included manuscripts from multiple countries, including the Republic of Korea, found that the number of articles reporting positive results increased by over 20% during the years of 1990 through 2007 and that corresponding authors in Asian countries reported more positive results than those in the US and Europe (14).

The higher publication of positive results is a responsibility not only to researchers or authors but also to editors and reviewers. Some editors and reviewers devalue negative results or inconclusive findings and automatically reject the manuscript. Since human subject studies frequently produce negative or inconclusive results, the negative attitude of some editors and reviewers may be a barrier to publication. Editors should review negative results to determine if they are sound. Negative results due to bias or fault are not acceptable. The ICMJE recommends that editors make editorial decisions based on factors such as the manuscript's overall quality and relevance to the journal, and not on the conclusiveness of the results (15). The International Committee of Medical Journal Editors (ICMJE) reminds editors that negative and inconclusive results can make an important contribution to science and that findings that are not statistically significant may provide important data for a subsequent meta-analysis (15).

Clinical trial registration is another mechanism for publicly sharing information about research. The Declaration of Helsinki and the ICMJE both require registration of clinical trials in a publicly accessible database before patient enrollment (1, 16). This means that all journals that follow ICMJE guidance, including JKMS, only publish manuscripts on trials that were properly registered. Clinical trial registration is important because it improves transparency, helps to reduce research duplication, contributes to the scientific evidence base and circumvents the issues of publication bias and selective reporting (17). Studies suggest that data in registries may also be more complete than what is published in medical literature. For example, the US registry [clinicaltrials.gov] requires results reporting and one study found that the results in the registry, especially serious adverse events, were more complete than the results in published articles (18). However, the strength of registries depends on researchers' submission of timely and accurate data and evidence suggests that concerning gaps still exist (19).