34 © 2012 IMIA and Schattauer GmbH
Business Process Modelling is an Essential Part of a Requirements Analysis Contribution of EFMI Primary Care Working Group
S. de Lusignan1,3, P. Krause2, G. Michalakidis2, M. Tristan Vicente3, S. Thompson1, M. McGilchrist4, F. Sullivan4, P. van Royen5, L. Agreus6, T. Desombre1, A. Taweel7, B. Delaney8 1 Department of Health Care Management and Policy, University of Surrey, Guildford, Surrey, UK 2 Department of Computing, University of Surrey, Guildford, Surrey, UK 3 Primary Care Informatics, Division of Population Health Sciences and Education, St. George’s – University of London, London, UK 4 Division of Clinical & Population Sciences and Education, The Mackenzie Building, Dundee, Scotland 5 Dept of Primary and interdisciplinary care, University of Antwerp, Antwerpen (Wilrijk), Belgium 6 Dept of Neurobiology, Care Sciences and Society, Karolinska Institutet, Huddinge, Stockholm, Sweden 7 Department of Informatics and Public Health, King’s College London, London, UK 8 Department of Primary Care and Public Health Sciences, London, UK
[11, 12]. Cancer registries are also well Summary Introduction established [13, 14] and more recently Objectives: To perform a requirements analysis of the barriers to con- Requirements analyses are part of soft- have become international and used to
ducting research linking of primary care, genetic and cancer data. ware engineering with the purpose of study of a wide range of conditions Methods: We extended our initial data-centric approach to include reducing the chances of failure or the [15], and the effectiveness of treat- socio-cultural and business requirements. We created reference models need for frequent expensive changes ments [16]. Primary care data are also of core data requirements common to most studies using unified mod- further along the development path- widely used in research. These data are elling language (UML), dataflow diagrams (DFD) and business proc- way. A requirements analysis is the rich, collected longitudinally and in ess modelling notation (BPMN). We conducted a stakeholder analysis process of determining what a system countries with registration systems pro- and constructed DFD and UML diagrams for use cases based on simu- must do, acknowledging the at times vide a population denominator [17, 18]. lated research studies. We used research output as a sensitivity analysis. conflicting needs of the various Research networks have been devel- Results: Differences between the reference model and use cases iden- stakeholders [1, 2]. Conventionally oped within primary care, at regional tified study specific data requirements. The stakeholder analysis conducting a requirements analysis was and national level and have improved identified: tensions, changes in specification, some indifference from an early step in the waterfall approach the capability and capacity for research data providers and enthusiastic informaticians urging inclusion of to software development [3, 4, 5]. in primary care; including recruitment socio-cultural context. We identified requirements to collect informa- However, "agile" approaches which in- into trials [19, 20]. There are many tion at three levels: micro- data items, which need to be semantically clude user and stakeholder involvement models, sometimes they are part of interoperable, meso- the medical record and data extraction, and throughout development [6] and recog- National or regional sentinel networks macro- the health system and socio-cultural issues. BPMN clarified nise the need for flexibility and change [21], or of single vendor linked data complex business requirements among data providers and vendors; [7] have been used more and also com- collection schemes [15]. and additional geographical requirements for patients to be repre- bined with the waterfall approach. There have been a number of at- sented in both linked datasets. High quality research output was the Medical research is a complex proc- tempts to link these different types of norm for most repositories. ess and although there are many com- data for research [22, 23, 24, 25], but Conclusions: Reference models provide high-level schemata of the puterised data repositories, they are not the successes have largely been lim- core data requirements. However, business requirements’ modelling commonly linked in research studies ited to the use of specific methods in identifies stakeholder issues and identifies what needs to be ad- [8]. These data repositories include: bioinformatics [26, 27]. dressed to enable participation. genetic databases, disease registries, and, routinely collected primary care Keywords data. Genetic data has been collected General practice; medical records systems, computerized; research into biobanks since the completion of Objective design; registry; records as topic the mapping of the human genome [9, 10], though biological specimens have We carried out this requirements analy- Yearb Med Inform 2012:34-43 been used in forensic science for longer sis to identify how to maximise the use
IMIA Yearbook of Medical Informatics 2012 35 Business Process Modelling is an Essential Part of a Requirements Analysis
of IT in conducting complex research with elsewhere within the project [34]. (1) End users of the linked datasets; spe- projects linking aggregated primary We developed a schema of the proc- cifically those looking to run the use care data to a disease registry or a ge- esses associated with a linked data re- case defined linked-data studies. netic data repository. search project as a reference model (2) Direct and indirect providers of these (Figure 1) [35] based on our experi- data. Direct suppliers include da- ence of extracting, aggregating and tabase owners or other direct data processing routinely collected data [36, providers to a study. We spoke in Methods 37]. Additionally we recognised that depth to ten. Indirect data providers linked research required the same in- include those recording or aggregat- 1 Overview dividuals to be represented in both ing data earlier in the process. linked databases; as cases present in (3) The TRANSFoRm study team and We adopted an agile approach to al- only one database could not be used their funders, and low for any alterations in the require- for linked research. We called this link (4) External experts, principally from ments as other parts of the project de- a "geographical requirement" (Figure the International Medical Infor- veloped [28]. We simultaneously 2) and included it within our business matics Association (IMIA) and Eu- conducted a literature review and held requirements. ropean Federation for Medical expert consensus workshops [29, 30]. We created UML and data flow dia- Informatics (EFMI) primary care We developed a reference model and grams (DFD) for the reference model informatics working groups. models for data flow and of the busi- at a high level of abstraction, so we ness process involved. We conducted could contrast them with individual a stakeholder analysis that included study use cases [38, 39, 40]. interacting with potential data provid- 5 Use Case Testing ers and electronic health record (EHR) We analysed the TRANSFoRm use cases system vendors. 4 Stakeholder Analysis to define the data requirements and the
integrity of the use cases, constructing a We identified the following stakeholder DFD and UML models for each. The 2 TRANSFoRm Research groups and summarised our findings in use cases were in two clinical domains, a mind map and stakeholder analysis type-2 diabetes and acid-reflux related Programme grid (Figure 3, Table 2): oesophageal disease. The type-2 diabe- This investigation was part of the TRANSFoRm (Translational Research and Patient Safety In Europe) research programme [31]. TRANSFoRm is de- signed to reduce the barriers to con- ducting research using routine health data and promote interoperability be- tween different health computer sys- tems so that international data can be collected about the quality of care and for research [32]. One element of TRANSFoRm was a requirements analysis to assess the feasibility of con- ducting research studies set out in the form or use cases.
3 Creating a Reference Model for the Research Process and Studies We separately modelled the use case defined research studies, using the Unified Modelling Language (UML) [33]. We excluded consent and data privacy issues as they were being dealt Fig. 1 Schema of the process involved in conducting a linked data research study
IMIA Yearbook of Medical Informatics 2012 36 de Lusignan et al.
(Table 2). There were challenges in getting direct data providers to engage with the process. We made 316 con- tacts with data providers we thought might participate and only 56 (17.7%) supplied the comprehensive informa- tion defined by the requirements analysis. The proportion of valid re- sponses from primary care data pro- viders was 26.4% (29/110); from can- cer registries 16.9% (15/89); and from genetic data bases 10.3% (12/117). Many of the data repositories were re- gional not national and already sus- tained by a viable programme of re- search; most were producing high grade research output. Their data Fig. 2 BPMN schema of the key business decisions which inform whether a data repository is likely to participate in a TRANSFoRm research study processing systems had grown or- ganically as EHR systems had be- come more sophisticated. They were tes cohort study explored the risk of com- The stakeholder analysis informed us generally proprietary, not using plications and response to oral medica- these were important and they were ne- standard metadata, and functioning tion, a cohort study linking primary care glected in our initial programme design. as "black boxes" between data entry and genetic data. The second domain, and export for research. However, all gastro-oesophageal reflux disease could output data in a range of stand- (GORD), included two use cases link- 8 Sensitivity Analysis ard formats. ing primary care and cancer registry We identified and contacted 17 data. One was a randomised control- We collected data about peer review pub- EHR vendors identified through our led trial of on-demand compared with lications produced by data repositories initial searches and a further nine continuous use of proton pump inhibi- being considered for inclusion in the flagged by stakeholders. EHR ven- tors (PPIs) a class of anti-indigestion TRANSFoRm research studies. We did dors were reluctant to engage unless medicine. The other was a nested case- this because that track record may be there were some chance of benefits control study to explore the association predictive of future success. Finally, we realisation including a priori publica- of GORD symptoms and the use of PPIs conducted a review of what elements of tion of criteria for inclusion in future with developing Barrett’s disease and our requirements analysis contributed to TRANSFoRm studies. One primary cancer of the oesophagus. our final specifications. care EHR system reported it has an open application programme interface (API). The lack of open APIs makes 6 Defining the Scope of the 9 Ethical Considerations this approach to interaction with EHRs more challenging as it might require This investigation did not involve any Information Needed from separate licenced APIs to be developed direct contacts with patients or access Candidate Databases with each vendor. to medical records. No specific ethical TRANSFoRm investigators had We defined the scope of the informa- consent was deemed necessary. tion we would need to collect from can- different perspectives reflecting their didate data providers; though asking varying backgrounds; they were stakeholders to define what they con- largely clinical or information sys- sidered to be essential requirements. tems researchers. These differences Results were nuanced, but the former gener- ally described the project as a com- 1 Stakeholder Analysis plex clinical process, while the lat- 7 Business Process Modelling We identified key stakeholders and ter looked to rationalise things into Late on in our process we introduced used a mind map to capture key con- machine processable activities. The Business Process Modelling Notation cepts (Figure 3) and summarised their TRANSFoRm senior investigators (BPMN) to define business requirements. views in a stakeholder analysis grid were trying to co-ordinate the "big
IMIA Yearbook of Medical Informatics 2012 37 Business Process Modelling is an Essential Part of a Requirements Analysis
Fig. 3 Mind map of stakeholders and issues
Table 1 Summary stakeholder analysis
Project Stakeholder Specific Information Needs Best Source of Information Planned Method of Delivery Timing Considerations Needed
1 End users / researchers Will data deliver study? 1. Own study protocol Require a third party to link data Want to avoid any delays in
1. Unique linked patient records; 2. Relevant use case research process 2. Case definition (Coding); 3. Track record of provider 3. Outcome variables; 4. Inclusion & exclusion criteria 5. Study specific - controls, randomisation, patient link 2 Direct Data providers 1. Return on investment of time - Who will get a funded study? Engage once benefits Have standard processable (Data repositories and studies, funding Track record of researcher outputs readily available networks) 2. Strategic interest Network / repository owner Indirect Data providers 1. About use of their data Largely via network / repository Challenging to change (Citizens, Patients, GPs) 2. Maintenance of confidentiality/privacy arrangement 3 Study team & funders 1. Progress towards achieving milestones 1. TRANSFoRm study protocol 1. Connecting data via TRANSFoRm Critical nature of dependencies in project 2. Work package outputs engine from previous work packages 2. Requirements of next steps in project 2. Provenance engine 3. Meeting own institutions objectives 4 External experts Academic interest in how and whether Simulations & studies - proof of Informal input / advice re: projects Variable this can be done concept
picture" and focused on project mile- in scope of provenance extended be- and socio-cultural factors all influence stones and reports to funders. There yond what is defined within the use the process and should be included were also inevitable changes in cases as it included influences prior in the requirements analysis [23, 24]. project scope. Important changes for to recording. Their view was that contextual prov- the requirements analysis were the in- The Informatics working group enance whilst important could only clusion of cancer registries and adop- members also broadened the scope apply to data that were recorded and tion of contextual provenance with of the requirements. They argued that a broader contextual overview the requirement to describe "How that human factors and the wider might better explain why data were that data came to be." This change context of the health system, legal present or not.
IMIA Yearbook of Medical Informatics 2012 38 de Lusignan et al.
2 Use Case Analysis We compared the use cases developed for the TRANFoRm project with the reference model schema we constructed for a generic research study to identify specific data requirements. The diabetes use case varied little from the reference model schema other than we needed to add a prescribing database. The UML model reflected that this was a cohort study where a subset of the population is identified (Figure 4(a)). The GORD use cases were more complex, also reflecting the study type (Figures 4(b) and 4(c)). 4(a) In this use case genetic and primary care data about type-2 dia- betes (T2D) are linked; <
IMIA Yearbook of Medical Informatics 2012 39 Business Process Modelling is an Essential Part of a Requirements Analysis
Discussion 1 Principal Findings The requirements for linking com- plex heterogeneous databases are ex- tensive, and failure to foresee the business requirements resulted in stake-holder disengagement. In ad- dition to the anticipated core data re- quirements we have identified geo- graphical and socio-cultural requirements. The use case, data driven model, helped define whether a database was suitable to conduct a particular research study. However, we also required information about the geographical overlap between data repositories and readiness to par- ticipate. We used a reference model to identify generic as well as study specific requirements. Local lan- guage and the regional nature of some data repositories are additional challenges to conducting interna-
tional research. Many of the success- ful data collection systems were pro- prietary and saw no reason to move to more open methods and systems; they believed this entailed risk and expense but with no obvious benefit. Only after we included BMPN in our approach did we manage to ration- alise and frame these requirements.
Fig. 5 DFD for the TRANSFoRm use cases 2 Implications of the Findings The data requirements from the use case analysis identified most data needed to conduct research, but may 5 Sensitivity Analysis maining four, two estimated 11 to 20 miss important business issues which and two estimated 2 to 5 publications. may constrain involvement. Whilst We used publication of peer-reviewed generic reference models for re- studies as an index of a functional busi- search studies make sense of the ness process. Eight of the ten data 6 Contributions to the Requirements processes and data flows; the busi- providing stakeholders who provided ness requirements of the data reposi- full information could provide evi- Analysis tory owners and EHR vendors also dence of work produced from their Finally, we report how the different need to be modelled and met. Re- data and indexed in the bibliographic inputs contributed to the final require- searchers who design data projects database Medline. Four data providers ments analysis (Table 2). With the ex- that involve linking databases need estimated between 30 and 100 peer- ception of socio-cultural issues, all the to make sure that they include in- reviewed publications had been pub- requirements were identified by more centives for participation otherwise lished in the last five years. Of the re- than one source. they risk low levels of engagement.
IMIA Yearbook of Medical Informatics 2012 40 de Lusignan et al.
Fig. 6 Detailed BPMN diagram to demonstrate the steps in ascertaining if a data source meets the business requirements
Table 2 Scope of data collection defined by the requirements analysis
Micro- Meso- Macro- Study level level level specific
Dara source / Data Method & Record system Organisational Socio-cultural Use-Case Data type Interoperability control of data & information level factors compatibility level extraction model
Data extracted Diabetes study Demographics ID Use case/protocol and Primary Health Wider social Genetic data HL7 Health Level Primary care & Coded Data EPR architecture Care System influences Seven - www.hl7.org Free-text genetic data Data quality On data availability Fragmentation / consent & access of care Primary Care / CDISC Clinical Data Taxonomy of errors PCROM to data Cancer Registry Exchange Consortium associated with data Primary Care Separate providers data - www.cdisc.org extraction Research Model Coverage Socio-cultural Barrets Univesal/limited dimensio to what is oesophagus & disease & presented BRIDG Biomedical medicatoin ISO 11179 OpenEHR to primary care Research Integrated Management of All types of data Metadata Archetypes Domain Group - http//metadata- www.openehr.org Ambulatory care www.bridg.org standards.org/11179 sensitive conditions
One strategy to overcome this might 3 Comparison with the Literature is a working example of the use of be creating a research network infra- ontologically driven systems [49]. Such structure to register researchers and Linking heterogeneous data is challeng- ontologically driven systems might best data providers wishing to conduct ing [46]. Use cases and UML have used be developed using object-orientated ap- linked data studies that might pro- in research reporting adverse events in proaches as has been used in specialist vide a forum to facilitate and broker clinical trials [47] and medical image cancer management [50] and for check- solutions. analysis pilots [48]. The OntoRAT toolkit ing healthcare domain models [51].
IMIA Yearbook of Medical Informatics 2012 41 Business Process Modelling is an Essential Part of a Requirements Analysis
Although widely used in software to the life sciences to help generate engineering, the modelling tools we UML models [59]; and other tailored Conclusions propose to be used have been little packages have been used to generate We defined generic user requirements used in the health domain; despite UML diagrams for bioinformatics for research studies linking health calls that they should be made more research [60] and to generate genotype- data based on a reference model, accessible to health researchers [52]. phenotype maps [61]. which can be extended for specific DFD have been used little in health Our requirements analysis pro- studies; using UML use case dia- care though they have been proposed duced a non-executable set of mod- grams and DFD. Additionally we as a component of object orientated els and methods for sorting candi- have identified geographical require- health information management system date databases into those potentially ments and the importance of socio- [53]. UML has been used the most. It useable in linked data research. We cultural context. However, the key has been proposed as a tool for avoid- did not define these steps in suffi- learning from this exercise is the im- ing integration problems [54] and for cient detail to be able to set them out portance of modelling the business modelling research studies [55], but using business process executable process and including this within a re- thus far with little evidence of ben- language (BPEL) specifications [62], quirements analysis. Failure to include efit [56]. BPMN is now considered BPEL has been used within health- and recognise the importance of mod- an appropriate tool for modelling care [63], and might help reduce elling the business process as part of digital pathology and care pathways barriers to potential researchers and requirements analysis risks lack of en- but has not been proposed for mod- data providers. gagement by intended participants and elling health research [57, 58]. increases the risk of project failure.
Acknowledgements 4 Limitations of the Method 5 Call for Further Research IMIA and EFMI for supporting their primary care informatics working
We could have just used UML. The The user requirements developed need groups. Elena Crecan for her contribu- UML alternatives are either to de- to be tested prospectively. Where the tion to the research. TRANSFoRm is velop a domain model or use UML required activities can be published with part-financed by the European Com- state-machine or finite state machine clearly defined interfaces it will be mission - DG INFSO (FP7 2477). diagrams [26-8]. Business architec- possible to produce an executable model Antonis Ntasioudis for assistance with ture models (BAM) have been applied such as BPEL. the diagrams and modelling.
Data sharing agreement Table 3 Contribution of use cases and stakeholder analysis to final requirements The limited quantitative data collected as part of this study have been placed on a database at University of Dundee Micro Meso Macro Study Specific Sensitivity Data level analysis with the Work Package lead (FS); ini- / type tially for further analysis within the Patient Coding Metadata Record Health Socio- Diabetes Non- Track record TRANFoRm project. The generic mod- level system & & system & service cultural & GORD use peer review els developed as part of this study are data Interoper- Extraction information organisa- & legal studies case publications freely available for use from this pa- ability model tional per or from the clinical informatics Use cases x xx x xx website:http://www.clininf.eu/
Stakeholder refmodel/ - users are requested to cite
analysis: this paper.
End users xx xx x x xx xx xx Data xx xx xx x Providers References
Study Team xx x x x x x 1. Cadle J, Paul D, Turner P. Business analysis techniques, 72 Essential Tools for Success. External x x xx xx xx xx xx Swindon: BCS; 2010. experts 2. Alexander I, Stevens R. Writing better requirements. London: Addison-Wesley; 2002. 3. Krause P, de Lusignan S. Procuring interoperability Key: x=contribution to requirements; xx=major contribution to requirements at the expense of usability: a case study of UK
IMIA Yearbook of Medical Informatics 2012 42 de Lusignan et al.
National Programme for IT assurance process. Stud Candy B, de Lusignan S. Towards a conceptual uri=CELEX:52004DC0356:EN:NOT Health Technol Inform 2010;155:143-9. framework for evaluating primary care research 33. Kumarapeli P, de Lusignan S, Ellis T, Jones B. 4. Hood C, Wiedemann P, Fichtinger S, Pautz U. networks. Br J Gen Pract 2000;50(457):651-2. Using Unified Modelling Language (UML) as a The interface between requirements development 20. Peterson KA, Fontaine P, Speedie S. The Electronic process-modelling technique for clinical-research and all other systems engineering processes. Primary Care Research Network (ePCRN): a new process improvement. Med Inform Internet Med Berlin; Springer-Verlag; 2010. era in practice-based research. J Am Board Fam 2007 Mar;32(1):51-64. 5. Reddy M, Pratt W, Dourish P, Shabot M. Med 2006;19(1):93-7. 34. de Lusignan S, Chan T, Theadom A, Dhoul N. Sociotechnical requirements analysis for clinical 21. Boffin N, Bossuyt N, Vanthomme K, Van Casteren The roles of policy and professionalism in the systems. Methods Inf Med 2003;42(4):437-44. V. Readiness of the Belgian network of sentinel protection of processed clinical data: a literature general practitioners to deliver electronic health review. Int J Med Inform 2007 Apr;76(4):261-8. 6. Montabert C, McCrickard D, Winchester W, Perez- record data for surveillance purposes: results of 35. Blobel B. Ontologies, knowledge representation, Quinones M. An integrative approach to survey study. BMC Fam Pract 2010;11(1):50. artificial intelligence – hype or prerequisites for requirements analysis: How task models support 22. Friedman C, Hripcsak G, Johnson SB, Cimino International pHealth interoperability. Stud Health requirements reuse in a user-centric design JJ, Clayton PD. A generalized relational schema Technol Inform 2011;165:11-20. DOI: 10.3233/ framework. Interact Comput Aug 2009;21(4):304- for an integrated clinical patient database. 978-1-60750-735-2-11 15. DOI: 10.1016/j.intcom.2009.06.003 Proceedings 14th Annual Symposium on 36. van Vlymen J, de Lusignan S, Hague N, Chan T, 7. Surendra N. Using an ethnographic process to Computer Applications in Medical Care. Los Dzregah B. Ensuring the quality of aggregated conduct requirements analysis for agile systems Alamitos, CA: IEEE Computer Society Press; general practice data: lessons from the Primary development. Inf Technol Manag 2008;9(1):55- 1990. p. 335-9. Care Data Quality Programme (PCDQ). Stud 68. DOI 10.1007/s10799-007-0026-6 23. Prokosch HU, Ganslandt T. Perspectives for Health Technol Inform 2005;116:1010-15. 8. Matteson S, Paulauskis J, Foisy S, Hall S, Duval medical informatics. Reusing the electronic 37. van Vlymen J, de Lusignan S. A system of metadata M. Opening the gate for genomics data into clinical medical record for clinical research. Methods Inf to control the process of query, aggregating, research: a use case in managing patients’ DNA Med 2009;48:38-44. cleaning and analysing large datasets of primary samples from the bench to drug development. 24. Burgun A, Bodenreider O. Accessing and care data. Inform Prim Care 2005;13(4):281-91. Pharmacogenomics 2010;11(11):1603-12. integrating data and knowledge for biomedical 38. Podeswa H. UML™ for the IT Business Analyst: 9. International Human Genome Sequencing research. Yearb Med Inform 2008:91-101. A Practical Guide to Object-Oriented Requirements Consortium.Initial sequencing and analysis of the 25. Ohmann C, Kuchinke W. Future developments of Gathering. Boston, MA: Thompson Course human genome. Nature 2001;409(6822):860–921 medical informatics from the viewpoint of technology PTR; 2005. 10. International Human Genome Sequencing networked clinical research. Interoperability and 39. Fakhroutdinov K. UML, Use Case Diagrams. Consortium. Finishing the euchromatic sequence integration. Methods Inf Med 2009;48(1):45-54. URL: http://www.uml-diagrams.org/use-case- 26. Mirhaji P, Zhu M, Vagnoni M, Bernstam EV, diagrams.html of the human genome. Nature 2004;431(7011): 931-45. Zhang J, Smith JW. Ontology driven integration 40. Ambler S. UML2 Use Case Diagrams. URL: 11. Hirtzlin I, Dubreuil C, Préaubert N, Duchier J, platform for clinical and translational research. http://www.agilemodeling.com/artifacts/ Jansen B, Simon J, et al; EUROGENBANK BMC Bioinformatics 2009;10 Suppl 2:S2. useCaseDiagram.htm Consortium. An empirical survey on biobanking 27. Wang X, Liu L, Fackenthal J, Cummings S, 41. Object Management Group (OMG) business Olopade OI, Hope K, et al. Translational integrity management initiative. Business process of human genetic material and data in six EU and continuity: personalized biomedical data modelling notation (BPMN), version 2. URL: countries. Eur J Hum Genet 2003;11(6):475-88. integration. J Biomed Inform 2009;42(1):100-12. http://www.bpmn.org/ 12. McEwen J. Forensic DNA data banking by state 28. Huo M, Verner J, Zhu L, Ali Barbar M. Software 42. Dolin RH, Alschuler L. Approaching semantic crime laboratories. Am J Hum Genet 1995; quality and agile methods. Proceedings of 28th interoperability in Health Level Seven. J Am Med 56(6):1487-92. Annual International Computer Software and Inform Assoc 2011 Jan 1;18(1):99-103. 13. Schneider B. Progress in cancer control through Applications Conference (COMPSAC’04) 43. Michalakidis G, Kumarapeli P, Ring A, van cancer registries. CA Cancer J Clin 1958; 2004;Vol 1: 520-5. http://doi.ieeecomputersociety. Vlymen J, Krause P, de Lusignan S. A system for 8(6):207-10 org/10.1109/CMPSAC.2004.1342889 solution-orientated reporting of errors associated 14. Navarro C, Martos C, Ardanaz E, Galceran J, 29. de Lusignan S, Peace C, Shaw N, Liaw S-T, with the extraction of routinely collected clinical Izarzugaza I, Peris-Bonet R, et al.; Spanish Michalakeidis G, Vincente M, et al. What are the data for research. Stu Health Technol Inform Cancer Registries Working Group. Population- barriers to conducting International research using 2010;160:724-8. based cancer registries in Spain and their role routinely collected primary care data? Stud Health 44. Nadkarni, P.M., Brandt, C. Data Extraction and Ad in cancer control. Ann Oncol 2010;21 Suppl Technol Inform 2011;165:135-40. DOI: 10.3233/ Hoc Query of an Entity-Attribute-Value Database. J 3:iii3-13. 978-1-60750-735-2-135 Am Med Inform Assoc 1998; 5:511-7 15. Gatta G, Capocaccia R, Trama A, Martínez-García 30. de Lusignan S, Liaw S-T, Krause P, Curcin V, 45. Santos MR, Bax MP, Kalra D. Building a logical C; RARECARE Working Group. The burden of Vincente M, Michalakidis G, et al. Key concepts EHR architecture based on ISO 13606 standard rare cancers in Europe. Adv Exp Med Biol to assess the readiness of data for International and semantic web technologies. Stud Health 2010;686:285-303. research: Data quality, lineage and provenance, Technol Inform 2010;160(Pt 1):161-5. 16.Francisci S, Capocaccia R, Grande E, extraction and processing errors, traceability, and 46. Castano S, De Antonellis V. Global viewing of Santaquilani M, Simonetti A, Allemani C, et curation. Yearb Med Inform 2011;6(1):112-20. heterogeneous data sources. IEEE Transactions al.; EUROCARE Working Group. The cure of 31. TRANSFoRm (Translational Research and on Knowledge and Data Engineering cancer: a European perspective. Eur J Cancer Patients safety In Europe). URL: http:// 2001;13(2):277-97. DOI: http://dx.doi.org/ 2009 Apr; 45(6):1067-79. www.transformproject.eu 10.1109/69.917566 17. de Lusignan S, Chan T. The development of 32. European Union, Europe’s Information Society. 47. London JW, Smalley KJ, Conner K, Smith JB. primary care information technology in the United Communication from the Commission to the The automation of clinical trial serious adverse Kingdom. J Ambul Care Manag 2008 Jul- Council, the European Parliament, the European event reporting workflow. Clin Trials 2009; Sep;31(3):201-10. Economic and Social Committee and the 6(5):446-54. 18. de Lusignan S, van Weel C. The use of routinely Committee of the Regions - e-Health - making 48. Estrella F, Hauer T, McClatchey R, Odeh M, collected computer data for research in primary healthcare better for European citizens: an action Rogulin D, Solomonides T. Experiences of care: opportunities and challenges. Fam Pract plan for a European e-Health Area engineering Grid-based medical software. Int J 2006;23(2):253-63. {SEC(2004)539}; 2004. URL: http://eur- Med Inform 2007;76(8):621-32. 19. Clement S, Pickering A, Rowlands G, Thiru K, lex.europa.eu/LexUriServ/LexUriServ.do? 49. Al-Hroub Y, Kossmann M, Odeh M. Developing
IMIA Yearbook of Medical Informatics 2012 43 Business Process Modelling is an Essential Part of a Requirements Analysis
an Ontology-driven Requirements Analysis Tool process improvement. Med Inform Internet Med and modeling with UML. Methods Mol Biol (OntoRAT): A Use-case-driven Approach. 2007;32(1):51-64. 2010;637:23-45. Proceedings of second international conference on 56. Vasilakis C, Lecnzarowicz D, Lee C. Application 61. Baker EJ, Jay JJ, Philip VM, Zhang Y, Li Z, the applications of digital information and web of Unified Modelling Language (UML) to the Kirova R, et al. Ontological Discovery technologies (ICADIWT 2009) 2009:130-8. Modelling of Health Care Systems: An Environment: a system for integrating gene- 50. Weber R, Knaup P, Knietitg R, Haux R, Merzweiler Introduction and Literature Survey. In Ed. Tan J. phenotype associations Genomics 2009;94 A, Mludek V, et al. Object-oriented business Developments in Health Information Systems and (6):377-87. process analysis of the cooperative soft tissue Technologies: Models and methods. Vancouver, 62. Zheng Y, Zhou J, Krause P. An Automatic Test sarcoma trial of the german society for paediatric BC: IGI-global; 2010,275-87. DOI: 10.4018/978- Case Generation Framework for Web Services. oncology and haematology (GPOH). Stud Health 1-61692-002-9.ch019. Journal of Software 2007;2(3):64-77. http:// Technol Inform 2001;84(Pt 1):58-62. 57. Scheuerlein H, Rauchfuss F, Dittmar Y, Molle R, citeseerx.ist.psu.edu/viewdoc/download? 51. Baksi D. Model checking of healthcare domain Lehmann T, Pienkos N, et al. New methods for doi=10.1.1.94.6186&rep=rep1&type=pdf models. Comput Methods Programs Biomed clinical pathways-Business Process Modeling 63. Tan W, Missier P, Foster I, Madduri R, Goble C. 2009;96(3):217-25. Notation (BPMN) and Tangible Business Process A Comparison of Using Taverna and BPEL in 52. Jun GT, Ward J, Morris Z, Clarkson J. Health care Modeling (t.BPM). Langenbecks Arch Surg 2012 Building Scientific Workflows: the case of caGrid. process modelling: which method when? Int J Feb 24. [Epub ahead of print] Concurr Comput 2010;22(9):1098-1117. Qual Health Care 2009;21(3):214-24. 58. Rojo MG, Daniel C, Schrader T. Standardization 53. Krol M, Reich DL. Object-oriented analysis and efforts of digital pathology in Europe. Anal Cell design of a health care management information Pathol (Amst) 2012;35(1):19-23. system. J Med Syst 1999;23(2):145-58. 59. Becnel Boyd L, Hunicke-Smith SP, Stafford Correspondence to: 54. Benson T. Prevention of errors and user alienation GA, Freund ET, Ehlman M, Chandran U, et al. Simon de Lusignan in healthcare IT integration programmes. Inform The caBIG(R) Life Science Business Department of Health Care Management and Policy Prim Care 2007;15(1):1-7. Architecture Model. Bioinformatics 2011 Mar University of Surrey 55. Kumarapeli P, De Lusignan S, Ellis T, Jones B. 29. [Epub ahead of print] Guildford, Surrey, UK Using Unified Modelling Language (UML) as a 60. Yan Q. Bioinformatics for transporter pharmaco- Tel +44 (0) 1483 683 089 process-modelling technique for clinical-research genomics and systems biology: data integration E-mail: [email protected]
IMIA Yearbook of Medical Informatics 2012