ICBO 2014 Proceedings Development of a BFO-Based Informed Consent Ontology (ICO) Yu Lin1*, Marcelline R. Harris2, Frank J. Manion3, Elizabeth Eisenhauer2, Bin Zhao4, Wei Shi3, Alla Karnovsky5, Yongqun He1* 1 Unit for Laboratory Animal Medicine, Department of Microbiology and Immunology, and Center for Computational Medicine and Bioinformatics, University of Michigan Medical School, Ann Arbor, MI 48109, USA 2 Division of Systems Leadership and Effectiveness Science, University of Michigan School of Nursing, Ann Arbor, MI 48109 3 Comprehensive Cancer Center, University of Michigan Medical School, Ann Arbor, MI 48109 4 University of Michigan School of Information, Ann Arbor, MI 48109 5 Department of Computational Medicine and Bioinformatics, University of Michigan Medical School, Ann Arbor, MI 48109 Abstract — An Informed Consent Ontology (ICO) was Investigations (OBI) [5] and Information Artifact Ontology developed to support informed consent data integration and (IAO). reasoning. ICO is aligned with the Basic Formal Ontology (BFO), To further extend ICO, we started with manually and logically represent the terms and their relations related to identifying and extracting a list of candidate terms by concept informed consent processes and contents. ICO contains 471 terms including 137 ICO-specific terms and the other terms imported from extraction from the informed consent templates used at the existing reliable ontologies. The ontology is available at http://ico- University of Michigan, which covers clinical research study, ontology.googlecode.com/. behavioral research study and biobank areas. The candidate terms were then mapped to several pre-identified resources, Keywords— Informed consent; ontology; ICO; OBO Foundry, especially the National Cancer Institute Thesaurus (NCIt) [6]. Basic Formal Ontology (BFO), OBI ontology This manual mapping process allowed us to reuse terms and definitions vetted by others and judged to be consistent with I. INTRODUCTION the use of the term in consent documents. The candidate terms The informed consent process is one of the fundamental were   then   categorized   and   organized   based   on   BFO2’s   pillars of human research [1]. A human subject research study structure. Authors held face to face discussions to review must undergo informed consent process, which cannot be those terms and their definitions. If appropriate, logical conducted unless approved by a regulatory body. The entire axioms were defined to provide restrictions to these terms. informed consent process involves giving a subject adequate ICO was generated using the Web Ontology Language information of the study, providing adequate opportunity for (OWL2). Protégé-OWL 4.2 was used for the ontology the subject to consider all options, ensuring that the subject authoring and editing. New terms were generated using new has comprehended this information, obtaining the subject’s ICO   IDs   with   the   prefix   of   “ICO_”   followed   by   seven   auto- voluntary agreement to participate and, continuing to provide incremental digital numbers. information as the subject or situation requires. The signed documents will then be archived for possible future usage. II. RESULT Adoption of electronic consent documents has been appealing A. ICO availability and statistical summary to the clinical research community [1]. Yet, there is no ICO is released under creative commons by 3.0 License. It coherence of representing informed consent in various has been deposited into the Ontobee program [7]: electronic systems, which impede productive data integration http://www.ontobee.org/browser/index.php?o=ICO. Ontobee and sharing when interoperability among those systems become mandatory. To tackle this challenge, we initiated a is the default program for dereferencing ICO ontology terms. community-driven effort to develop an Informed Consent ICO has also been deposited into NCBO BioPortal: http://bioportal.bioontology.org/ontologies/ICO. Ontology (ICO), to enable Semantic Web technology that As of Aug. 14, 2014, ICO contains 471 terms (137 ICO- allows integration, sharing and meaningful information specific), including 385 classes (131 ICO-specific), 55 object extraction while keeping related consent information properties (3 ICO-specific), 30 annotation properties (3 ICO- distributed, dynamic and diverse in different systems. specific), and one datatype property. ICO was developed using a combination of top-down and bottom-up approaches. The   ‘backbone’   of   ICO   is   formed   by   B. BFO-aligned ICO hierarchy adopting the Basic Formal Ontology (BFO) 2 as the upper As shown in Fig. 1, the ICO informed consent related level ontology [2]. Using the tools Ontodog [3] or OntoFox terms are ultimately placed under ‘continuant’ and ‘occurent’ [4], related ontology terms were imported from reliable branches of BFO2. The top level ICO terms are placed under ontologies including the Ontology for Biomedical either OBI terms (e.g. informed consent process related terms) or IAO terms (e.g. informed consent document related terms). * Co-Corresponding authors. 84 ICBO 2014 Proceedings Various types of informed consent forms and their elements purpose, study protocol, benefits and risks of participating, are represented in ICO. sampling procedure, sample usage, data derived from the document informed consent signed informed study, data usage, and record confidentiality. Finally, the (IAO) form (ICO) consent form (ICO) information …  ... ‘signed informed consent form’   will   be   archived   (‘archiving content entity (IAO) document part of informed signature section in signed informed consent form’) for future use (Fig. 2). consent form informed consent part (IAO) (ICO) form (ICO) continuant (BFO) specimen …  ... III. DISCUSSION (OBI) investigator (ICO) We presented a BFO-based informed consent ontology. material organism entity (BFO) (OBI) human (OBI) human subject (ICO) Importing BFO and selected OBI and IAO terms provided a organization informed consent Institutional Review basic syntactic and semantic framework for further ICO regulatory body (OBI) (ICO) Board (NCIt) development. ICO intends to capture both the depth and entity (BFO) breadth of informed consent in clinical study research domain, informed consent process (OBI) and to be used in the following example applications: 1) informed consent form approval (ICO) Automatic generation of electronic informed consent forms; 2) process (BFO) planned process (OBI) informed consent form validation (ICO) Informed consent validation; 3) Biobank biospecimen storage, signing informed consent form (ICO) processing, and data release. The development of ICO is in its occurrent (BFO) early stage. We look forward to future collaborations with permission cancellation (ICO) other related ontologies and research efforts [8, 9]. temporal temporal informed consented study period (ICO) region (BFO) interval (BFO) ACKNOWLEDGMENT Fig. 1. The top level hierarchical structure and key terms of ICO. We thank Dr. Nicholas H. Steneck and Blake J. Roessler C. Modeling informed consent workflows for their valuable discussions and feedback. This research was An example of a workflow starting form an ‘informed supported by a University of Michigan interdisciplinary consent form design’ to the archiving of a ‘signed informed research award (MCubed) and by the National Center for consent form’ is modeled using ICO (Fig. 2). Advancing Translational Sciences of the National Institutes of Health (NIH) under Award Number 2UL1TR000433-06. The content is solely the responsibility of the authors and does not necessarily represent the official views of the funding sources. REFERENCES [1] M. M. Issa, E. Setzer, C. Charaf, A. L. Webb, R. Derico, I. J. Kimberl, et al., "Informed versus uninformed consent for prostate surgery: the value of electronic consents," J Urol, vol. 176, pp. 694-9; discussion 699, Aug 2006. [2] P. Grenon and B. Smith, "SNAP and SPAN: Towards Dynamic Spatial Ontology," Spatial Cognition and Computation, vol. 4, pp. 69-103, 2004. [3] J. Zheng, Z. Xiang, C. J. Stoeckert, Jr., and Y. He, "Ontodog: a web-based ontology community view generation tool," Bioinformatics, Feb 1 2014. [4] Z. Xiang, M. Courtot, R. R. Brinkman, A. Ruttenberg, and Y. He, "OntoFox: web-based support for ontology reuse," BMC Res Notes, vol. 3, p. 175, 2010. [5] R. R. Brinkman, M. Courtot, D. Derom, J. M. Fostel, Y. He, P. Lord, et al., "Modeling biomedical experimental processes with OBI," J Biomed Semantics, vol. 1 Suppl 1, p. S7, 2010. Fig. 2. ICO modeling of an informed consent process. [6] NCI. (2014). NCI Thesaurus (NCIt). Available: After an informed consent form is drafted, it must be http://ncit.nci.nih.gov/ approved by the   ‘informed consent regulatory body’. An [7] Z. Xiang, C. Mungall, A. Ruttenberg, and Y. He, "Ontobee: A approved informed consent form will then be subjected to the linked data server and browser for ontology terms," in The 2nd ‘informed consent form process’, which includes: 1) International Conference on Biomedical Ontologies (ICBO), ‘explaining to participant candidate about informed consent Buffalo, NY, USA, 2011, pp. Pages 279-281. study’  indicated  in  the   informed  consent  form;;  2)  ‘assessment [8] A. Grando and R. Schwab, "Building and evaluating an ontology-based tool for reasoning about consent permission," of   participant   candidate’s   understanding’   of   the   explanation; AMIA Annu Symp Proc, vol. 2013, pp. 514-23, 2013. 3)   ‘participant candidate making voluntary decision of [9] M. Brochhausen, M. N. Fransson, N. V. Kanaskar, M. Eriksson, acceptance’;;   and 4) ‘signing the informed consent form’.   To R. Merino-Martinez, R. A. Hall, et al., "Developing a ensure the participant make voluntary decision, it is necessary semantically rich ontology for the biobank-administration to explain to him/her adequate information. Those were domain," J Biomed Semantics, vol. 4, p. 23, 2013. captured in ICO as parts of   ‘explaining to participant candidate about informed consent study’,   including: study 85 ICBO 2014 Proceedings Development of a BFO-Based Informed Consent Ontology (ICO) Yu Lin, Marcelline R. Harris, Frank J. Manion, Elizabeth Eisenhauer, Bin Zhao, Wei Shi, Alla Karnovsky, Yongqun He University of Michigan, Ann Arbor, MI 48109, USA Abstract ICO modeling of informed consent workflow An Informed Consent Ontology (ICO) was developed to support informed consent An  example  of  an  informed  consent  workflow  starting  form  an  ‘informed consent data integration and reasoning. ICO is aligned with the Basic Formal Ontology form design’  to  the  archiving  of  a  ‘signed informed consent form’  is  modeled  using   (BFO), and logically represent the terms and their relations related to informed ICO  (Fig.3).    This  model  signifies  the  ‘informed consent process’  by  partitioning  it   consent processes and contents. ICO contains 471 terms including 137 ICO- into  partial  processes:  ‘explaining to participant candidate about informed consent specific terms and the other terms imported from existing reliable ontologies. The study’,‘assessment of  participant  candidate’s  understanding’   and  ‘participant ontology is available at http://ico-ontology.googlecode.com/. candidate making voluntary decision of acceptance’.  To ensure the participant make voluntary decision, more detailed partial processes of explaining study Introduction protocol, study purpose, benefit and risks, The informed consent process is one of the fundamental pillars of human research sample collection and [1]. The entire informed consent process involves giving a subject adequate usage, data information of the study, providing adequate opportunity for the subject to consider collection and usage, all options, ensuring that the subject has comprehended this information, obtaining record confidentiality the subject’s voluntary agreement to participate and, continuing to provide have been arranged information as the subject or situation requires. The signed documents will then be as subclasses of archived for possible future usage. Adoption of electronic consent documents has ‘planned  process’.  It been appealing to the clinical research community [1]. Yet, there is no coherence of is also modeled that representing informed consent in various electronic systems, which impedes a  ‘informed consent productive data integration and sharing when interoperability among those systems process’  has   becomes mandatory. To tackle this challenge, we initiated a community-driven effort specified input to develop an Informed Consent Ontology (ICO), to enable Semantic Web ‘informed consent technology that allows integration, sharing and meaningful information extraction form’  and  specified   while keeping related consent information distributed, dynamic and diverse in output  ‘signed different systems. informed consent ICO overview form’  follows  OBI’s   Besides a full import of BFO 2 as our framework, we design pattern [2] started from identifying reusable component from OBO (Fig.4). Fig.4 ICO modeling a typical informed consent in general. foundry ontologies, especially OBI and IAO ontologies. Based on this top-down procedure, we further Discussion expanded the ontology by extracting related terms from ICO intends to capture both the depth and breadth of informed consent in clinical three informed consent templates used in University of study research domain, and to be used in the following example applications: 1) Michigan. The terminology expertise then mapped the Automatic generation of electronic informed consent forms; 2) Informed consent extracted terms to other existing resources: UMLS®, validation; 3) Biobank biospecimen storage, processing, and data release. NCIt, BRIDG, OCRe, CHV, UCSD permission ontology, ICO is in its early stage of NCBO Bioportal repository and Ontobee repository. development. we are aware of The definitions and relations of ICO terms were several other groups are finalized by manual review during ICO developers’ Fig.1 Portions of imported classes and working on informed consent meeting. ICO classes in current ICO. related web application or The current ICO (v.53) is written in RDF/OWL syntax. It contains 471 terms (137 eConsent, such as OMIABIS, ICO-specific), including 385 classes (131 ICO-specific), 55 object properties (3 d-acts ontology, OMRSE and ICO-specific), 30 annotation properties (3 ICO-specific), and one datatype property. some non-BFO aligned ontologies. As shown in Fig.5, BFO aligned ICO hierarchy the content of current ICO are rich in informed consent and document (IAO) informed consent form (ICO) signed informed consent form (ICO) informed consent form related Fig.5 Word cloud of ICO terms indicates that ICO is rich in informed information content entity …  ... terms. It requires substantial consent and informed consent forms related terms. ICO terms are (IAO) document part of informed signature section in evaluation against workflows and decision making by multiple parties involved. ultimately placed part (IAO) consent form (ICO) informed consent form (ICO) We look forward to future collaborations with other related ontologies and research under ‘continuant’ continuant (BFO) specimen …  ... efforts to enrich and expand ICO. and ‘occurent’ (OBI) investigator (ICO) branches of material organism Acknowledgements human (OBI) human subject (ICO) BFO2. The top entity (BFO) (OBI) We thank Dr. Nicholas H. Steneck and Blake J. Roessler for their valuable discussions and level ICO terms organization informed consent Institutional Review feedback. This research was supported by a University of Michigan interdisciplinary research regulatory body are placed under (OBI) (ICO) Board (NCIt) award (MCubed) and by the National Center for Advancing Translational Sciences of the either OBI terms National Institutes of Health (NIH) under Award Number 2UL1TR000433-06. The content is entity (BFO) solely the responsibility of the authors and does not necessarily represent the official views of (e.g. informed informed consent process (OBI) the funding sources. consent process informed consent form approval (ICO) References related terms) or process planned process informed consent form validation (ICO) 1. M. M. Issa, E. Setzer, C. Charaf, A. L. Webb, R. Derico, I. J. Kimberl, et al., "Informed versus uninformed consent for prostate surgery: the IAO terms (e.g. (BFO) (OBI) value of electronic consents," J Urol, vol. 176, pp. 694-9; discussion 699, Aug 2006. 2. R. R. Brinkman, M. Courtot, D. Derom, J. M. Fostel, Y. He, P. Lord, et al., "Modeling biomedical experimental processes with OBI," J Biomed informed consent occurrent signing informed consent form (ICO) Semantics, vol. 1 Suppl 1, p. S7, 2010. document related (BFO) List of Acronyms permission cancellation (ICO) terms) (Fig.2). OBO: Open Biological and Biomedical Ontologies, BFO: Basic Formal Ontology, OBI: Ontology for Biomedical Investigations; IAO: temporal temporal Information Artifact Ontology, UMLS®: Unified Medical Language System, NCIt: National Cancer Institute Thesaurus, NCBO: National Center for informed consented study period (ICO) region (BFO) interval (BFO) Biomedical Ontology, BRIDG: Biomedical Research Integrated Domain Group, OCRe: Ontology of Clinical Research , CHV: Consumer Health Vocabulary, UCSD:University of California San Diego, OMIABIS: Ontologized Minimum Information About BIobank data Sharing, OMRSE: Ontology of Medically Related Social Entities- Fig.2 Hierarchy  of  ICO  terms  placed  under  BFO’s  framework. 86