Posted in Flagyl on May 5, 2015

‘); } //]]>‎Figure 1. The structure of a statement in Wikidata. Source: []. View this shape


Drug-Drug Interaction Data

We reviewed frankly available DDI knowledge bases, including Drugbank [], the National Drug File-Reference Terminology (NDF-RT) [], a suddenly list of clinically important and stale DDIs maintained by the CredibleMeds nonprofit organized being [], the Office of the National Coordinator in the place of Health Information Technology (ONC) “extreme priority” list of potential DDIs suggested like high priority to alert clinicians in at all care environment [], and the ONC “noninterruptive” list of potential DDIs not requiring interruptive alerting in some care environment []. We decided to contain only clinically relevant potential DDIs that had boastful priority and were curated by teams of experts in direct to avoid overloading the knowledge base with large numbers of interactions that were not hazardous or not backed by sufficient testimony []. We selected the ONC high priority list for these reasons, reserving the possibility of further extension of the lore base at a later point in time.

Adding Data to Wikidata and Wikipedia

The protuberance for adding potential DDI data to Wikidata consisted of independent steps and community interactions. First, we registered notwithstanding accounts on the Wikidata system. In ill-defined, every person with or without every account can directly edit data pages via the Wikidata website. Complex contribution tasks are usually organized end WikiProjects (eg, WikiProject Medicine), which are the in the ~ place points of contact for people who partake a specific goal in Wikidata expansion. If users want to be part of a WikiProject, they solely have to add themselves to the cast participant list. We joined the Wikidata Medicine protrude [].

We researched if appropriate Wikidata properties publicly existed or if it would be necessary to create a new Wikidata property. We fix no existing property for capturing possible DDIs, so we drafted a proffer for a new property and submitted it on this account that community discussion and voting (discussion archived at []). Three discussants participated in this disputation. After the name, description, and semantics were civilized by us based on community feedback and the profit and novelty of the property were ascertained, the proposed property was accepted and added to the Wikidata method []. Major discussion points were the inclusion of forage-drug interactions, whether the property should exist symmetric or should be nonsymmetric to behold precipitant and objects drugs, and whether existing properties instead of physical interactions between objects could subsist reused. The entire process took 1 week. Minor backwards-compatible refinements to the property explanation were conducted at later stages of the suit based on requirements that became plain during data loading. After a property is created, any user can translate it. Additionally, property necessity can be set that help to obtain mistakes in the data when of that kind constraints are violated.

We created a Wikidata “bot” (a program that autonomously adds or edits appease) for automatically adding and updating interactions from the larger ONC loftily priority list. The implementation and appeal of the bot was discussed with the Wikidata community until consensus steady its utility and the scope of the imported facts was reached (discussion archived at []). Eight discussants participated in this discussion.

To operate the bot, we registered a sunder bot account and created a solicit for permission (as stated in the Wikipedia bot management []). The request’s text included a narrative of our team and the bots main functionality and goals. The resulting debate included 8 other Wikipedians (mainly members of the WikiProject Medicine) and spanned other thing than a month before the bot was approved ~ dint of. a Wikidata “bureaucrat” (the official title of members responsible for similar decisions in the Wikipedia/Wikidata community). The bot was created by using the Pywikibot carcass, which is a collection of Python scripts in quest of maintaining different MediaWiki sites []. We derived the bot from a basic template script [], that is part of the Pywikibot software budget and used specific functionality geared ready Wikidata editing. The code of the bot we created is available on GitHub [].

We created modified templates of drug infoboxes for the English and German Wikipedia that turn to account potential DDI data from the Wikidata apprehension base; Infoboxes are templates that typically be associated with a Wikipedia article and are usually displayed at the be superior-right of a page. They typically keep in check the most important facts about a Wikipedia ingress. We copied selected Wikipedia articles into an online development environment and integrated the modified deaden with narcotics infobox templates so that potential DDI given conditions from Wikidata were automatically shown in the resulting Wikipedia articles. This workflow can potentially be implemented for any of the else than 288 languages of Wikipedia similar to long as the property name and the statement’s values are translated. These articles were used to the degree that a proof of concept and to foretoken how information from Wikidata would have existence displayed in Wikipedia with current technologies.

For this direct study, we implemented these changes in a segregated Wikipedia expansion environment instead of the main regularity because changes to widely used templates (eg, the medicine infobox) usually require extensive discussion. Given that the Wikidata assiduity program interface (API) is still not full stable, large-scale adoption of Wikidata-based infoboxes are not however approved by the community. The Wikipedia unfolding environment is not technically different from erect Wikipedia, but it is less observable (eg, Web search engines do not usually aim to pages in the development environment).


We conducted a direct analysis to explore if adding the ONC overbearing priority data would significantly enhance the denunciation currently available on Wikipedia in terms of coverage of significant potential put ~s into interactions. We developed a simple script that analyzed English Wikipedia articles of substances in the ONC verging on taint priority list and checked if interacting substances were mentioned in the current Wikipedia branch through string matching.

Finally, an person specially versed clinical pharmacist (JH) conducted a detailed evaluation of the current Wikipedia DDI complaint available for 2 examples of drugs up~ the body the ONC list (ramelteon and warfarin). A model reference for clinically important DDIs coauthored through the expert was used as lowest part for the evaluation [].


We derived 1150 peculiar interactions from the ONC high priority list. The coverage of active ingredients in Wikidata was exhaustive and we did not urgency to create new entities. A screenshot of one example of the data represented in Wikidata is shown in and screenshots of the same data as represented in an infobox in the German and English Wikipedia are shown in .

Integration of the possible DDI data from Wikidata into the Wikipedia infobox proved to have existence straightforward and yielded useful results, mete also highlighted potential shortcomings of the current combination of parts to form a whole. We recognized difficulties with making far-reaching lists of interacting drugs accessible through the infoboxes because some drugs had interactions by up to 30 other drugs. This enigma could be addressed by hiding spun out lists of interacting drugs behind expandable user interface elements, which is already done for other use cases (eg, gene ontology term annotations put ~ protein articles), or by removing other accusation that might be of lower benefit. Still, the existence of these for a ~ time lists of interactions for some drugs might lead to unsatisfactory usability or reduced findability of influential information. An alternative could be the integration of the Wikidata notice into the main text of the Wikipedia item as exemplified in .

Another shortcoming of the current integration of Wikidata into Wikipedia that was uncovered ~ means of our prototype is the way in what one. literature citations/evidence is rendered. Although backing make manifest from Wikidata can be displayed through endnote references in Wikipedia, the Wikipedia hypothesis currently does not recognize that multiple citations may end to a single reference, leading to the appointment of redundant reference list entries. Furthermore, the current theory is only able to display the titles of belles-lettres references, but does not provide a formatted endnote by all bibliographic details (eg, authors, diary, and publication year). These current limitations make the ability of Wikipedia readers to efficiently counter the evidence behind data from the Wikidata attainments base. However, given the novelty of Wikidata and the ongoing progressive growth of the database interfaces, this restraint will soon disappear.

‘); } //]]>‎Table 1. Statistics forward coverage of drug-drug interaction (DDI) premises in existing English Wikipedia articles compared to the ONC pre-eminent priority list for 1150 DDI pairs pure.View this table

A review of randomly selected articles showed that divers contained implicit information about drug interactions by providing information about interacting drug classes or interactions by enzymes. However, in many cases, readers potency not have the background knowledge essential to infer actionable information from these statements (eg, the admonition that a drug significantly inhibits the cytochrome P450 3A4 [CYP3A4] enzyme requires the perception that a potentially coadministered drug is metabolized ~ the agency of CYP3A4 to be of practical benefit). We also observed that many articles that did not comprise explicit mentions of interacting drugs from the ONC primeval priority list did mention interactions with other drugs that were not attached the list, suggesting a significant overall inconsistency between drug interaction information on Wikipedia and the ONC admirable priority list.

It is also momentous to keep in mind that the ONC strip is not intended to be including both of all clinically relevant DDIs because it was developed from a severe list of interacting drug pairs and in that case expanded to include related drugs that in like manner interact. This makes the lack of DDIs in Wikipedia strange to say more disturbing because many important DDIs are alone missing.

Detailed Analysis of Exemplary Wikipedia Drug Articles

The following detailed analyses of DDI complaint in Wikipedia articles for the drugs ramelteon and warfarin show by example some of the limitations of the current Wikipedia article coverage of DDIs from the vista of a professional pharmacologist.

As eminent in , ramelteon is primarily metabolized by CYP1A2. CYP3A4 and CYP2C9 also give to its metabolism. Thus, any unsalable article that alters the activity of CYP1A2, CYP3A4, or CYP2C9 is suitable to alter the elimination of ramelteon. Ramelteon moreover has a very low bioavailability (total of oral drug reaching systemic diffusion) of 1.8%. This means that granting that a 10 mg dose of ramelteon is taken vocally, less than 0.2 mg inclination reach the systemic circulation and cause a pharmacologic response. If another physic inhibits the metabolism of ramelteon, the amount of ramelteon reaching the systemic motion in a circle will be increased, potentially to a actual large extent.

Currently, Wikipedia notes exclusive drugs that do not interact through ramelteon; however, it is not not to be mistaken if the drugs have no purport on ramelteon or that ramelteon has no effect on the listed drugs. Actually, the two outcomes are true. None of the drugs listed in the same proportion that noninteracting with ramelteon would be expected to interact since they do not affect any of the metabolic pathways that metabolize ramelteon. The Wikipedia entry further states:

A drug interaction study showed that there were no clinically meaningful effects or every increase in adverse events when ramelteon and the SSRI Prozac (fluoxetine) were coadministered. Ramelteon and fluvoxamine should not have ~ing coadministered. Ramelteon should be administered with caution in patients taking other CYP1A2 inhibitors, able to endure CYP3A4 inhibitors such as ketoconazole, and intoxicating CYP2C9 inhibitors such as fluconazole. Efficacy may exist reduced when ramelteon is used in compound with potent CYP enzyme inducers such as rifampin, since ramelteon concentrations may exist decreased.

Although no references are on condition for the preceding statements, they stand in judgment to be taken from the ramelteon fruit label. We agree that ramelteon should have existence avoided in patients taking fluvoxamine for the reason that the label notes a 190-fold increase in ramelteon levels resulting from the interaction. Fluvoxamine inhibits CYP1A2 for example well as CYP2C9 and CYP3A4. As remarkable previously, CYP1A2 is the primary enzyme that metabolizes ramelteon. Other drugs that bar CYP1A2 include atazanavir, ciprofloxacin, amiodarone, enoxacin, mexiletine, tacrine, thiabendazole, cimetidine, ticlopidine, zileuton, and vemurafenib, notwithstanding none of these are mentioned in the Wikipedia thing. Likewise, there are many drugs that restrain both CYP3A4 and CYP2C9 other that the 2 drugs listed in the thing. For example, atazanavir, ciprofloxacin, and cimetidine check 2 of the metabolic pathways of ramelteon and would potentially end large increases in ramelteon plasma concentrations.

Warfarin is primarily metabolized by CYP2C9 with CYP1A2 and CYP3A4 in like manner contributing to its elimination. Although moreover lengthy to reproduce fully here, Wikipedia notes drugs that “be able to displace warfarin from serum albumin and action an increase in the international normalized rate (INR)” can interact with warfarin. This is theoretically steady; however, because the clearance of warfarin from the blood is limited to drug that is not attached to albumin, displaced warfarin is metabolized and the total of active warfarin at steady state does not change so the INR is in like manner unchanged.

Antibiotics such as metronidazole and macrolides are illustrious to “greatly increase the general of warfarin” by reducing its metabolism. Metronidazole bequeath reduce the metabolism of warfarin, being of the kind which will several other antibiotics and antifungal agents that are not mentioned in the turning-point and are likely to be other commonly used and produce larger changes in warfarin concentrations than the drugs cited. Macrolides bring about not appear to alter warfarin metabolism. They regard been associated with enhanced warfarin efficiency, but this is likely due to other causes including the poison itself or altered diet. Several cases of altered warfarin rejoinder associated with thyroid activity are besides cited in the article. It has been demonstrated that the the ministry of thyroid supplementation does not vary patients’ response to warfarin. The Wikipedia portion notes that excessive use of highly rectified spirit can increase the response to warfarin. This is honest for binge drinking; however, chronic alcohol consumption is likely to increase the metabolism of warfarin leading to reduced response.

Several paragraphs in the thing recount selected reports of herb-warfarin interactions. Most of the reports cited execute not meet even minimal standards instead of evidence of an interaction and none of the trials showing no or minimal furniture of the herbals on warfarin are included. This is each example of selection bias in DDI reporting.

As by the ramelteon article, the warfarin p~ of logical quantity omits many important DDIs. No cursory reference is made of potent inhibitors of CYP2C9, such as amiodarone, sulfamethoxazole, fluconazole, voriconazole, or fluoxetine. No cursory reference is made of drugs that can increase the risk of bleeding in patients seizure warfarin, such as the nonsteroidal antiinflammatory drugs or acetaminophen.


Principal Results

We implemented a continuous experiment for enriching medical data in the Wikidata cognition base and demonstrated that automated updating of medical content in Wikipedia through Wikidata is a viable election, albeit further refinements and community-distant consensus building are required before integration into the world Wikipedia is possible. Adding data to Wikidata and Wikipedia is a lengthened process that requires lots of community interactions and familiarity with customs and requirements of the particular communities. We expect that actual integration into the self-governed Wikipedias in harvested land language will require further refinement and efficient dialog with the different WikiProject Medicine communities.

Better given conditions quality in Wikidata can reduce the sustenance work required in Wikipedia, giving editors else time to focus on the station of articles. If an article exists in totality 288 languages of Wikipedia, keeping it up to era or adding a piece of data with Wikidata amounts to a upright edit compared to 288 edits without Wikidata. This helps to improve the completeness and currentness of medical content on Wikipedia, a resource that has become central to health information seeking mixed patients and health professionals on a global scale.

We had the experience that the Wikidata community was very open toward novel participants and granted constructive feedback and assistance with the integration of new data into the complex Wikidata notice base. We decided to invite any Wikipedia member who provided significant hold to become a coauthor of this writing (TS). Based on our experiences, we eagerly recommend the inclusion of long-mete Wikidata and Wikipedia community members in philosophical or medical projects such as this human being.

Wikidata is a recent addition to the Wikipedia ecosystem and its strengths and weaknesses in wont widespread use for serving complex data to Wikipedia or as a ill-defined-purpose knowledge base have yet to be determined. Although centralized data management in Wikidata be possible to improve efficiency of data management and standing in Wikipedia, its integration into Wikipedia efficacy also be a source of problems. Of appropriate concern is the fact that not every one of data in Wikidata are necessarily displayed in Wikipedia and that the “sundry eyeballs” principle that helps to reform errors in Wikipedia might not put to some of the content in Wikidata. The inclusion of data without long-term plans of maintenance or inclusion in visible Wikipedia articles potency lead to a problematic accumulation of outdated data.

The review of the DDI entries in favor of warfarin and ramelteon reveals the rigorous limitations of the current system to produce clinically important and useful DDI denunciation. This problem might be partially mitigated by the strong community interaction and feedback that helps to decide that data to include or not comprise into Wikidata to keep the judgment base manageable. Furthermore, the usage of automated bots to imply and map data from primary sources into Wikidata efficiency play an important role. The bots can also make routine checks to resolve if the different Wikipedias use the similar data and if the data are a subset of the premises in Wikidata. Wikidata is also developing tools to barrier for inconsistencies in the data. Furthermore, the common is not only tending to the Wikipedia-expanded data but is also encouraged to detail mistakes to the source database, by that means improving databases that are willing to apportioned lot their data.

When we compared positive mentions of potential drug interactions in Wikipedia with interactions from the ONC high priority list, we found a large result of missing information in English Wikipedia. This potency be even more pronounced in other speech versions of Wikipedia that have, in inaccurate, fewer editors and worse coverage than English Wikipedia. This finding resonates with prior research that form in a mould substantial differences in drug interaction pairs captured in DDI perception bases [,]. The inclusion and exclusion criteria of a DDI knowledge base are vital to its adapted to practice utility. Although failing to include in a high degree. significant DDI can have obvious negative consequences, the suffered inclusion of large numbers of DDI backed ~ dint of. insufficient evidence or of low clinical meaning can have a negative impact to the degree that well because it can lead to cognitive overlade and the inobservance of truly indicative interactions [], frustrating clinicians [] and leading to unadapted responses []. The inclusion of clinically of great weight DDIs based on critical evaluation of radical DDI evidence or established expert-curated DDI wealth should be a goal of the Wikipedia/Wikidata community.


The publication of medical premises on a public website that includes readers from the commander-in-chief public requires special attention to ensure proper understanding of the data and its implications. In this regard, more remote work is required to reduce the jeopardy of improper utilization of the given conditions (eg, making it clear that nonoccurrence in the like does not imply that a undoubted drug combination does not carry risks of unprosperous events). Furthermore, when the current implementation notes a in posse interaction, it does not provide more remote information on the mechanism of the interaction or possible actions to mitigate patient risk.

A major limitation of our current implementation is the disjoin presentation of potential DDI data from Wikidata and possible DDI information in the main verse of the Wikipedia article. The in posse redundancies and differences between the 2 information sources might further add to this jumble and automatic checks comparing the Wikipedias to Wikidata are not in put at interest yet for drug interactions. In fit condition to have all potential DDI advice in one place, integrating potential DDI premises as a table into main verse under a “drug interactions” title might be preferable to the inclusion in the infobox forward the upper right hand of the serving-boy. Furthermore, editing policies need to exist set up to clarify the respect between structured data from Wikidata and unstructured theme in Wikipedia. Routine checks by bots, perhaps after every edit to a remedy page, could potentially determine if the interactions listed are a subset of the facts in Wikidata. We will investigate the automated inclusion of structured premises into the main article once the requisite features in both Wikipedia and Wikidata be delivered of reached sufficient maturity, which was not now the case at the time of this document.

A limitation of our evaluation methodology was that the file matching approach used for identifying DDI mentions power have missed mentions that used deaden with narcotics class references rather than individual substances.

Comparison With Prior Work

The Wikidata WikiProject Medicine that we are participating in is moreover involved in other endeavors, such being of the kind which managing sitelinks for medical topics or connecting sanatory topics with their corresponding identifiers in curative databases. Furthermore, a WikiProject for molecular biology content on Wikidata was lately established [].

Wikidata might also become each interesting platform for large-scale, graph-based learning integration tasks in the biomedical ~ of battle that have been realized with Semantic Web technologies in late years, such as Linked Open Drug Data [] or Bio2RDF []. Further careful search and pilot projects are needed to fathom the potential of Wikidata to be changed to a centralized repository of large-mount medical and life science data.


We confident that this work provides a groundwork for a long-term endeavor to improve the healing information in Wikipedia through structured given conditions representation and automated workflows. It leave strengthen the collaboration with the medical Wikipedia community to bring high-character information on drug safety into fruit use as part of Wikipedia in different languages. We will also seek to align our drudge on drug safety information in Wikidata through projects in related domains, such similar to biomedical research and the life sciences [,].

Finally, we are publicly preparing a collaboration with international experts in clinical pharmacology and mix with ~s safety to establish and maintain a supernatural agency-readable, open-source representation of significant DDIs based on the ONC verging on taint priority list as well as other sources. The organization for work of such a resource could not but benefit the quality of drug close custody data in Wikipedia, but also improve the temper of clinical decision support interventions or equal drug product labels [].

To have a sustained collision, it is vital that the Wikipedia common carries this work further in as well-as; not only-but also; not only-but; not alone-but the structured world of Wikidata and the textual nature of Wikipedia. We invite interested readers to join this endeavor.


We thank other members of WikiProject Medicine and Wikidata because of their feedback and support. We thank Serkan Ayvaz (Kent State University) and Lisa Hines (University of Arizona) by reason of assistance with accessing the ONC DDI dataset. The investigation leading to these results has accepted funding from the Austrian Science Fund (FWF; P 25608-N15) and the United States National Library of Medicine (1R01LM011838-01).

Authors’ Contributions

MS devised the study and wrote major parts of the manuscript; AP and TS led Wikidata common interaction, design decisions, and technically implemented the conductor system; RB gave advice about DDI datasets and overall generalship; and JH conducted an in-middle review of 2 Wikipedia drug articles. All authors contributed to writing the manuscript.

Conflicts of Interest

None declared.


API: putting on program interface

DDI: drug-drug interactions

INR: between nations normalized ratio

ONC: Office of the National Coordinator on the side of Health Information Technology

Edited by G Eysenbach; submitted 21.12.14; pry-reviewed by B Good, N Farič, J Sanz-Valero; comments to original 07.03.15; revised version admitted 12.03.15; accepted 14.03.15; published 05.05.15


©Alexander Pfundner, Tobias Schönberg, John Horn, Richard D Boyce, Matthias Samwald. Originally published in the Journal of Medical Internet Research (, 05.05.2015.

This is ~y open-access article distributed under the conditions of the Creative Commons Attribution License ( /2.0/), which permits unrestricted conversion to an act, distribution, and reproduction in any pervading substance, provided the original work, first published in the Journal of Medical Internet Research, is properly cited. The complete bibliographic information, a copula to the original publication on, in the same proportion that well as this copyright and permit information must be included.


<!– /templates/newDesign/hackneyed/footer.tpl –>

Journal of Medical Internet Research ISSN 1438-8871


Editorial Policies




Journal Sponsorship

Site Map

About this Publishing System


Copyright © 2015 JMIR Publications Inc.

Japanese ~ of health, labour and welfare gain positively done the marital study forward bpa.