=Paper=
{{Paper
|id=Vol-2237/medracer-paper-3
|storemode=property
|title=Automatically Identifying Drug Conflicts in Clinical Practice Guidelines
|pdfUrl=https://ceur-ws.org/Vol-2237/medracer-paper-3.pdf
|volume=Vol-2237
|authors=António Silva,Tiago Oliveira,Ken Satoh,Paulo Novais
|dblpUrl=https://dblp.org/rec/conf/kr/00040SN18
}}
==Automatically Identifying Drug Conflicts in Clinical Practice Guidelines==
Automatically Identifying Drug Conflicts in Clinical Practice Guidelines António Silva 1 , Tiago Oliveira 2 , Ken Satoh 2 and Paulo Novais1 1 Algoritmi Centre, University of Minho, Braga, Portugal, asilva@algoritmi.uminho.pt, pjon@di.uminho.pt 2 National Institute of Informatics, Tokyo, Japan, {toliveira,kstaoh}@nii.ac.jp Abstract is especially problematic for patients with multimorbidity. Clinical Practice Guidelines (CPGs) are documents devel- Multimorbid patients have complex treatment plans and face oped in a systematic way that aim to improve the quality of a high burden of the disease since they suffer from multi- health care, reduce variations in medical practice, and reduce ple diseases at the same time. There are also several prob- health care costs. However, when concurrently apply them, lems regarding the application of treatment plans of mul- this can lead to adverse drug-drug interactions that can im- tiple disease-specific CPGs to multimorbid patients. Such pair the patient’s condition. Several efforts have been made includes adverse drug events, increased treatment complex- in order to provide systems capable of identifying these con- ity, and cost of treatment (Tinetti, Bogardus Jr, and Agostini flicts. However, the current approaches for this purpose have 2004) (Boyd et al. 2005). Thus, the application of multiple some limitations. This paper presents a solution that rep- CPGs individually can result in complex multiple drug regi- resents CPGs as Computer-Interpretable Guidelines (CIGs) and allows for the automatic drug conflict identification and mens (polypharmacy) with the potential for harmful combi- resolution. Also, we provide the identification of improve- nations of drugs (Dumbreck et al. 2015). ments to include in a future model. Moreover, this system Therefore, new needs arise in order to provide computer- provides clinical recommendations in an agenda, being capa- assisted tools that automatically identify the common poten- ble of identifying drug interactions when drugs are prescribed tial conflicts or interactions that can happen when merging simultaneously and provide conflict-free alternatives. CIGs, namely, those that happen when there are drug-drug interactions. 1 Introduction The work described herein presents a system that auto- To improve the utilisation of clinical practice guidelines matically identifies recommendation interactions, conflicts, (CPGs) at the point of care, there have been numerous efforts and alternatives using existing terminology services such as to computerise CPGs in ontologies and incorporate them the RxNorm API. Thus, the contributions featured in this within Clinical Decision Support Systems (CDSSs). There work are: characterisation of main approaches to handle the are several guideline description languages such as Arden combination of CIGs, especially for multimorbid patients Syntax (Samwald et al. 2012), Guideline Interchange For- and a solution to address this problem. mat (GLIF) (Peleg et al. 2000), Asbru (Balser, Duelli, and The paper is organised as follows. Section 2 describes re- Reif 2002), EON (Musen et al. 1996), PROforma (Vier et al. lated work regarding systems for combining CPGs. Section 1997) and Guideline Acquisition, Representation and Exe- 3 presents an architecture for combining CIGs as well as cution (GLARE) (Bottrighi et al. 2006) that are aimed at the contributions for the deployment of CPGs in CDSSs. Fi- the representation of CPGs as computer-interpretable guide- nally, Section 5 presents the conclusions drawn so far with lines (CIGs) in order to provide computer-assisted tools that the development of the system and future directions for the help health care professionals make decisions. Through the work. formalisation of CIGs in CDSSs, a new range of operations can be performed with the knowledge they enclose. Such in- 2 Existing Systems for Identification of cludes automated reasoning for the generation of recommen- Conflicts Between Concurrently Executed dations, automatic identification of conflicts between differ- ent CIGs, consistency checking within the same CIG and CPGs across different CIGs, and merging CIG knowledge with When treating multimorbid patients, health care profession- contextual information such as patient and physician pref- als need to retrieve clinical recommendations from multiple erences or available health care resources. So, the objective chronic disease CPGs. From the combination of these rec- of these approaches is to provide support to the processes of ommendations, several problems can happen, for instance diagnosis and planning of the clinical treatments, as well as when a drug, prescribed for one condition, has an adverse ef- to promote the use of the best clinical practices. fect on another condition (Boyd et al. 2005). With the grow- However, these systems lack the flexibility to support ing number of multimorbid patients, identification of these cases where multiple protocols need to be combined; this inconsistencies becomes increasingly essential (Wilk et al. 2011). Computerised CDSSs have been used to alert health actions). OWL is a W3C standard for web ontologies, for care professionals to adverse drug events at the point of which CPG concepts are converted to RDF triplets and XML care (De Clercq, Kaiser, and Hasman 2008). In this section, file (Jafarpour and Abidi 2013). This model defines four we will provide a literature review of the existing CDSSs types of constraints for concurrent execution of tasks from for identification of conflicts between concurrently executed multiple guidelines: workflow constraints, operational con- CIGs. straints, temporal constraints and medical constraints. Workflow constraints are rules that specify whether tasks 2.1 Constraint Logic Programming should be combined with, substituted by, executed simul- Wilk et al. propose an approach that combines logic pro- taneously with or executed before or after a task from an- gramming with constraint satisfaction problems (Wilk et al. other guideline. Operational constraints refer to limitations 2011). They use CIGs as an activity graph and use constraint for combining tasks at a specific medical Institute; temporal logic programming to identify and mitigate possible adverse constraints specify the time required between the first and interactions between CIGs, it means, to identify conflicts second task of two guidelines. Medical constraints are di- associated with potentially contradictory and adverse ac- vided in Task Substitutes (a substitute for a task of protocol tivities resulting from applying two CPGs to the same pa- A that does not conflict with a task of protocol B) and use tient. Although this approach provides automatic identifica- results constraints ( a rule that specifies expiry date of task tion of conflicts and solutions, it depends on the availability results). of knowledge bases containing information about both dis- They also built a merging representation ontology to cap- eases and the whole work of combining CIGs remains man- ture merging criteria in order to achieve the combination of ual. So, in order to provide automatic identification conflicts CIGs. Semantic Web Rule Language (SWRL) rules were and solutions need to be defined in a medical background used to identify potential conflicts during the merging pro- knowledge as protocol-dependent rules/constraints. cess. All conditions related to the merging process need to be described by the rules, increasing the effort to maintain the 2.2 Rule-based Combinations system up-to-date, and reducing the possibility of sharing The RBC approach provides identification and reconcilia- knowledge. However, some related problems were not yet tion of drug conflicts between recommendations of two con- (completely) addressed in their work, for instance, potential currently executed CPGs (López-Vallverdú, Riaño, and Col- contradictions between rules, the scalability of the merging lado 2013). They use a standard terminology called ATC model to combine several CIGs, and how the ontology/rules (Anatomical Therapeutic Chemical Classification System are maintained up-to-date. for drugs) in order to provide as output, a final treatment plan without interaction, i.e., a set of ATC-codes of medicines that should be prescribed. 2.4 Transition-based Medical Recommendations For the identification of all possible drug conflicts that Model can occur when combining two specific CPGs, they use the knowledge from health care professionals and knowledge engineers in order to manually build knowledge units for the The TMR4I model has been developed for the automatic in- pairwise combination of three diseases: hypertension, dia- ference of interactions between recommendations (Zambor- betes mellitus and heart failure. These knowledge units rely lini et al. 2016). Its scope is currently limited to conflicts be- on the existence of drug-drug interactions, the presence of tween CPG statements on drug prescription, but it could be a drug which is adverse to a specific disease (drug-disease used for non-pharmacological treatment recommendations interaction) and the absence of a necessary drug for a com- as well. bination of diseases. This model defines meta-rules for identification and Although this approach can only combine pairs of CPGs, reconciliation of three categories of drug conflicts using a final treatment plan based on two CPGs could again be SPARQL queries (SPARQL is a W3C-standard for semantic combined in a pairwise manner with a new CPG. queries). The meta-rules define how a conflict is identified, and how drugs with similar effects but without conflicts can 2.3 OntoMorph be selected from CPG-knowledge. The categories of con- The objective of the OntoMorph (Jafarpour 2013) approach flicts within CPGs are repetition interactions, contradiction is to propose a treatment plan, consisting of several tasks, interactions and alternative interactions. that do not conflict and that are time and resource effi- A web-tool for execution of guidelines was developed. In cient. Jafarpour et al. (Jafarpour and Abidi 2013) used on- this tool, clinicians firstly enter all guideline recommenda- tologies to develop systems to merge two concurrent CPGs tions applicable to a patient. The execution engine creates a into a comorbid personalised guideline. They extracted clin- new, merged guideline with all recommendations. With the ical tasks from the CPGs and converted them to CIGs with SPARQL meta-rules, interactions are identified and classi- an OWL-based CPG ontology. An ontology is a method- fied. Then, the engine consults the alternative recommenda- ology for CPG representation. It consists of rules to repre- tions, in order to choose solutions for the conflicts. Finally, sent declarative knowledge (medical statements and propo- a list of conflicts and recommended solutions is presented to sitions) and procedural knowledge (workflow structures and the clinician. 3 Architecture to Automatically Identify expressed in the form of conditions on the patients state, Drug Interactions and Conflicts such as TriggerConditions, PreConditions and Outcomes. Moreover, it provides a model of temporal representation There are some limitations regarding the systems aforemen- (Oliveira et al. 2017) that aims to represent the temporal tioned which should be taken into account. These limitations constraints placed on clinical tasks. This model represents regard with the necessity of manually defining all guideline- temporal constructors on the execution of tasks such as Du- dependent rules, limitations in the number of CIGs that can rations, Repetitions, Periodicities, Waiting Times and Rep- be combined and necessity of all conflicts and solutions to etition Conditions and temporal constraints about the state be available in a knowledge base. of a patient. To acquire and represent CPGs we use the Thus, the present work not only aims to provide recom- CompGuide plugin which provides information step-by-step mendations to support medical decision-making but also to on how to fill the data for the guideline entries (Gonçalves represent automatically the conflicts and interactions that et al. 2017). This plugin performs the role of managing the can happen when merging CPGs. To accomplish the goal creation and editing of CIGs. mentioned above, as shown in Figure 1, we provide a solu- The final output will be a CIG that will be saved in tion in three levels: representation of CPGs in CIGs, iden- the Guideline Repository. This component is responsible tification of recommendation interactions and provision of for keeping different CIGs represented according to the recommendation alternatives in case that some recommen- CompGuide ontology. The Guideline Handler is responsi- dations, when applied together, are adverse. The following ble for managing the access to recommendations of CIGs sections provide an explanation of the architecture in the dif- in the Guideline Repository, providing the clinical tasks and ferent stages regarding the different level mentioned before. constraints placed on the tasks to the Guideline Execution Engine. 3.2 Identification of Recommendation Interactions The Guideline Execution Engine with the information of the clinical tasks provided by the Guideline Handler inter- prets all the scheduling constraints on the tasks and produces enactment times. The applications implemented to interact with the health care professionals are then responsible for verifying starting and ending times of the tasks. This component is also responsible for calling the RxNorm service in order to identify the interactions and recommendation conflicts. RxNorm (Liu et al. 2005) in- tegrates the Unified Medical Language System and offers normalised names for clinical drugs and links its names to many of the drug vocabularies commonly used in pharmacy management and drug interaction software, including those of First Databank, Micromedex, MediSpan, Gold Standard Drug Database, and Multum. The RxNorm interaction API uses two sources for its interaction information - ONCHigh and DrugBank. The RxNorm interaction API provides in- formation such as source name, severity and description of the interaction. Thus, the Guideline Execution Engine pro- cesses all the clinical tasks that are being executed, retrieves all drugs and for each pair of drugs calls the RxNorm In- teraction API to obtain the severity and description of the interaction. 3.3 Determine Alternative Recommendations Figure 1: Architecture of CompGuide system With the severity of recommendation interactions, it is pos- sible to determine the necessity of providing alternative rec- ommendations. The severity can assume values such as high 3.1 Representation of CPGs in CIGs if there is an adverse drug event resulting from the inter- The work described herein uses the CompGuide ontology action, and N/A, if there is no adverse effect. In case the to represent CPGs in the form of a task network. The interaction between drugs assumes as value high we call CompGuide ontology (Oliveira, Novais, and Neves 2013) RxNorm RxClass API. This service provides alternative rec- contains different types of clinical tasks such as Question, ommendations by offering ways to get similar classes of Action, Decision, End, Plan and Condition and constraints drug members. This service provides information such as similarity scoring (a score that determines the similarity be- tween drugs), the drug name, the source of the drug relations and the relationship of the drug class to its members. After processing the constraints of clinical tasks, deter- mining the interactions between drugs and their alternatives, the clinical recommendations are made available through the Personal Assistant Web App and the health care assistant Mobile App. The Personal Assistant Web Application access the data through the web services available in the CompGuide sys- tem (Silva et al. 2017). This component was developed as a web application following the Model-View-Control (MVC) paradigm using Java Server Faces (JSF). The Health care assistant Mobile Application is an android application de- veloped in Java, which also uses the same web services. 4 Execution Example This section describes how CompGuide processes the inter- Figure 2: Recommendation interactions between recom- actions between drugs given a case test example. For this mendations A and B in the CompGuide Personal Assistant purpose, we used two CIGs based on the NCCN Clinical Web Application. Practice Guideline for Prostate Cancer (Mohler et al. 2018) and the IDF Clinical Practice Recommendations for manag- ing Type 2 Diabetes (Aschner 2017). These guidelines were a comprehensive case study since it was possible to test sev- Later, the application tries to provide alternative drugs to eral aspects of the deployment of CIGs, namely those that address the identified conflicts, by calling RxNorm API as concern with CIG representation, acquisition and execution described in section 3.3. Through a mitigation function, the in the CompGuide system. So, it is possible to examine all system calculates which one will be applied. This function the stages of the deployment of CIGs, representing several has different mitigation principles, such as the similarity be- types of tasks, various temporal constraints and several con- tween drugs or user preferences. The objective is to deter- flicts among the guidelines. However, in this section, we mine which alternatives best fit the needs of users. One pos- only address the conflicts between recommendations from sible principle, which possibly will increase the effective- many guidelines. ness of this function, is a multiple criteria mechanism for For demonstrations purposes, we will consider two rec- supporting decision-making such as Multiple-criteria Deci- ommendations from the mentioned guidelines. The first one, sion Analysis (MCDA) (Thokala et al. 2016). This method named recommendation A belongs to the guideline for man- allow to evaluate possible solutions based on conflicting cri- aging Type 2 Diabetes: ”Apply insulin 0.2 units/kg and teria in decision problems. There are complex drug interac- titrate once weekly at one unit each time during six months tions that can impair the patient’s condition as well as several to achieve a target fasting blood glucose between 3.9 and solutions with conflicting objectives. Thus, it is essential to 7.2 mmol/L (70 and 130 mg/dL)”. The second recommen- evaluate the possible solutions according to criteria such as dation, named recommendation B belongs to the guideline user preferences in the best treatment alternative, benefit/risk for prostate cancer: ”Apply goserelin, leuprolide, histrelin assessment of different decision alternatives, the similarity 180 mg/m2 or Triptorelin 100mg/m2 as part of Androgen between different drugs, the severity of disease for which Deprivation Therapy”. recommendations are advised, among others. However, in The recommendation A has the action apply insulin, a pe- this case study we only use the similarity between drugs riodicity value of 1 with a temporal unit of week, a duration as the mitigation principle. This function is responsible for value of six, and the respective temporal unit of month. In finding the conflicts between drugs. For each conflict this this case, starting on the 18th of July of 2018 the system will function finds alternative drugs by calling the RxNorm API, create one event for each week with a duration of one day, according to section 3.3. Later, it calculates the high similar- during 6 months. The expected conclusion of this task will ity score provided by RxNorm API for the set of alternatives be on the 18th of January of 2019. As for recommendation B, drugs. For each alternative with the higher score, it tries to the action to apply goserelin, leuprolide, histrelin or triptore- find if a drug conflict exists. If there is a conflict, the algo- lin can be identified, with a duration value of 1 and temporal rithm finds the next alternative with the higher score, if there unit of day, starting and finishing on the 18th of July of 2018. is no conflict, it stores the alternative in the database and dis- In this case, the two recommendations are concurrently plays the alternative drug. Based on the given case example, being applied in 18th of July of 2018 and have drug con- the reproduced recommendation alternatives are shown in flicts, namely the drugs goserelin, leuprolide, histrelin and Figure 3. triptorelin have adverse effects on the therapeutic efficacy of In the work described herein, we provide a system that insulin. The Figure 2 shows the output regarding these con- automatically identifies conflicts and interactions between flicts. drugs for many guidelines. Comparing with OntoMorph, 5 Conclusions and Future Work There are several efforts in order to provide systems capable of determining drug-drug interactions and conflicts among guidelines. However, some of the studied systems are un- able to detect the conflicts for combinations of protocols automatically. Other approaches cannot propose alternative measures that would resolve the conflicts. Other CIG mod- els require all the possible conflicts and their solutions to be available in a knowledge base. Moreover, they cannot lead with cases where decision makers have conflicting solutions or cannot decide on the best treatment alternatives. Although we currently do not provide an MCDA ap- proach, it is our intention to implement a multiple criteria decision-making approach for not only assessing the benefit- risk of applying the recommendations but also getting pa- tient preferences on best treatment alternatives since some Figure 3: Recommendation alternatives for the given case treatment plans can have harmful effects on the patients example in the CompGuide Personal Assistant Web Appli- health . This allows to evaluate all possible solutions and to cation. specify different criteria to solve conflicts with medical rec- ommendations, beyond the simple comparison of drug inter- actions. Also, we intend to make a proper assessment of the fitness of the system for CIG deployment, by performing a study involving physicians interacting with the system. CLP and RBC, where conflicts are defined as constraints in the knowledge base having to be manually specified, CompGuide uses existing terminology services that aggre- 6 Acknowledgments gate different drug sources such as ONCHigh and Drug- This work has been supported by COMPETE: POCI-01-0145-FEDER-0070 43 and Bank. Thus, through the reuse and integration of existing ter- FCT Fundação para a Ciência e Tecnologia within the Project Scope UID/CEC/ 00319/2013. The work of Tiago Oliveira was supported by JSPS KAKENHI Grant minology services such as RxNorm, it is possible to identify Number JP18K18115. conflicts and interactions automatically, without the need to manually define them in the knowledge base. So, using existing terminology services is one of the possible solu- References tions for the limitation mentioned above. Other solution in Aschner, P. 2017. New idf clinical practice recommenda- regard to the usage of meta-rules, such used by the TM4I tions for managing type 2 diabetes in primary care. model. Meta-rules can be reused since they can be applied Balser, M.; Duelli, C.; and Reif, W. 2002. Formal semantics to many CIGs, and conflicts do not need to be manually of Asbru-an overview. In Proceedings of the International identified for each guideline, because they can be automat- Conference on Integrated Design and Process Technology. ically derived from the guideline representation. However, the bottleneck will be in converting guidelines to computer- Bottrighi, A.; Terenziani, P.; Montani, S.; Torchio, M.; and interpretable rules. Besides, these systems do not consider Molino, G. 2006. Clinical guidelines contextualization aspects such as decision-making. In most cases, there are in GLARE. In AMIA ... Annual Symposium proceedings several alternatives that can lead to conflicting objectives by / AMIA Symposium. AMIA Symposium, volume 2006, 860. the decision makers. In other cases, it is necessary to de- American Medical Informatics Association. cide which recommendation we want to choose, or which Boyd, C. M.; Darer, J.; Boult, C.; Fried, L. P.; Boult, L.; and recommendation, in the case at hand, is less adverse. For Wu, A. W. 2005. Clinical practice guidelines and quality this specific case, an MCDA approach can be a possible so- of care for older patients with multiple comorbid diseases: lution, since it allows the evaluation of possible solutions implications for pay for performance. Jama 294(6):716– based on conflicting criteria. In the given example we use a 724. mitigation function to determine which alternatives best fit De Clercq, P.; Kaiser, K.; and Hasman, A. 2008. Computer- the needs of users, by using as a mitigation principle, the interpretable guideline formalisms. Studies in health tech- similarity between drugs. One possible principle is a multi- nology and informatics 139:22. ple criteria mechanism for supporting decision making such as Multiple-criteria Decision Analysis (MCDA) (Thokala et Dumbreck, S.; Flynn, A.; Nairn, M.; Wilson, M.; Treweek, al. 2016). Since there may be complex interactions yield- S.; Mercer, S. W.; Alderson, P.; Thompson, A.; Payne, K.; ing multiple solutions with conflicting objectives, it is use- and Guthrie, B. 2015. Drug-disease and drug-drug interac- ful to score these solutions according to criteria spawning tions: systematic examination of recommendations in 12 UK from sources as diverse as patient preferences, the severity national clinical guidelines. Bmj 350(mar11 2):h949–h949. of disease for which recommendations are advised, bene- Gonçalves, F.; Oliveira, T.; Neves, J.; and Novais, P. fit/risk analysis, and so forth. 2017. Compguide: Acquisition and editing of computer- interpretable guidelines. In World Conference on Informa- Tinetti, M. E.; Bogardus Jr, S. T.; and Agostini, J. V. 2004. tion Systems and Technologies, 257–266. Springer. Potential pitfalls of disease-specific guidelines for patients Jafarpour, B., and Abidi, S. S. R. 2013. Merging disease- with multiple conditions. The New England journal of specific clinical guidelines to handle comorbidities in a clin- medicine 351(27):2870. ical decision support setting. In Conference on Artificial In- Vier, E.; Fox, J.; Johns, N.; Lyons, C.; Rahmanzadeh, A.; telligence in Medicine in Europe, 28–32. Springer. and Wilson, P. 1997. PROforma: systems. Computer Meth- ods and Programs in Biomedicine 2607(97). Jafarpour, B. 2013. Ontology merging using semantically- defined merge criteria and owl reasoning services: towards Wilk, S.; Michalowski, M.; Michalowski, W.; Hing, M. M.; execution-time merging of multiple clinical workflows to and Farion, K. 2011. Reconciling pairs of concurrently used handle comorbidity. Dalhousie University. clinical practice guidelines using constraint logic program- ming. In AMIA Annual Symposium Proceedings, volume Liu, S.; Ma, W.; Moore, R.; Ganesan, V.; and Nelson, S. 2011, 944. American Medical Informatics Association. 2005. Rxnorm: prescription for electronic drug information exchange. IT professional 7(5):17–23. Zamborlini, V.; Hoekstra, R.; Da Silveira, M.; Pruski, C.; Ten Teije, A.; and Van Harmelen, F. 2016. Inferring recom- López-Vallverdú, J. A.; Riaño, D.; and Collado, A. 2013. mendation interactions in clinical guidelines. Semantic Web Rule-based combination of comorbid treatments for chronic 7(4):421–446. diseases applied to hypertension, diabetes mellitus and heart failure. In Process Support and Knowledge Representation in Health Care. Springer. 30–41. Mohler, J. L.; Lee, R. T.; Antonarakis, E. S.; Armstrong, A. J.; D’Amico, A. V.; Davis, B. J.; Dorf, T.; Eastham, J. A.; Ellis, R.; Enke, C. A.; and Farrington, T. A. 2018. National Comprehensive Cancer Network - Prostate Cancer. Techni- cal report, National Comprehensive Cancer Network. Musen, M. A.; Tu, S. W.; Das, A. K.; and Shahar, Y. 1996. EON: A Component-Based Approach to Automation of Protocol-Directed Therapy. Emerging Infectious Diseases 3(6):367–388. Oliveira, T.; Silva, A.; Neves, J.; and Novais, P. 2017. De- cision support provided by a temporally oriented health care assistant. Journal of medical systems 41(1):13. Oliveira, T.; Novais, P.; and Neves, J. 2013. Representation of clinical practice guideline components in owl. In Trends in Practical Applications of Agents and Multiagent Systems. Springer. 77–85. Peleg, M.; Boxwala, a. a.; Ogunyemi, O.; Zeng, Q.; Tu, S.; Lacson, R.; Bernstam, E.; Ash, N.; Mork, P.; Ohno- Machado, L.; Shortliffe, E. H.; and Greenes, R. a. 2000. GLIF3: the evolution of a guideline representation format. In Proceedings / AMIA ... Annual Symposium. AMIA Sympo- sium, 645–649. American Medical Informatics Association. Samwald, M.; Fehre, K.; de Bruin, J.; and Adlassnig, K. P. 2012. The Arden Syntax standard for clinical decision sup- port: Experiences and directions. Journal of Biomedical In- formatics 45(4):711–718. Silva, A.; Oliveira, T.; Neves, J.; Satoh, K.; and Novais, P. 2017. A system for the management of clinical tasks throughout the clinical process with notification features. In Agents and Multi-Agent Systems for Health Care. Springer. 76–93. Thokala, P.; Devlin, N.; Marsh, K.; Baltussen, R.; Boy- sen, M.; Kalo, Z.; Longrenn, T.; Mussen, F.; Peacock, S.; Watkins, J.; et al. 2016. Multiple criteria decision analysis for health care decision makingan introduction: report 1 of the ispor mcda emerging good practices task force. Value in health 19(1):1–13.