Introduction
The Decision-making process in the field of medicine especially about patient diagnosis, treatment, and recommendation has been boggled by errors in the past. In an attempt to minimize these errors, there has been the growth of clinical decision support systems (CDSS), which generally constitute computer systems developed to improve clinician decision making about individual patients at the point in time that these decisions are generated (Berner, 2007). At the same time, CDSS functioning together with computer-based physician order entry (CPOE) have been evaluated and found to possess the ability to improve patient safety, reduce overall operating cost and at the same time impact positively on the medical care process (Berner, 2007).
ATHENA CDSS constitutes one of the various types of clinical decision-support tools that have become useful in the provision of recommendations for managing hypertension, specifically in primary care (Martins, 2006). In most cases, it is known as the ATHENA-HTN, whereby, it is described as the computer system that enables the clinical decision process by delivering guideline-based patient-specific recommendations about hypertension management (Martins, 2006). Moreover, ATHENA-HTN knowledge in most cases is stored in a knowledge-base-KB (Martins, 2006). This unique computer system is built on a component-based architecture known as EON. EON is a knowledge-based system architecture that provides physicians with decision support in guideline-based care (Martins, 2006). As a result, ATHENA CDSS can be seen to possess the capability of explaining, reasoning, and justifying its recommendations that in turn become useful in decision-making process (Martins, 2006).
Designing ATHENA-HTN CDSS
Why technology is necessary
Hypertension has been described as the most frequent treatable cardiovascular risk factor, where it remains a challenge in the medical field (Dossel, Dossel and Schlegel, 2009). The disease leads to cardiovascular disease or stroke, which in turn leads to premature morbidity and mortality. The prevalence of hypertension in the global world is uneven, but its impact on the majority of western countries is huge (Dossel, Dossel, and Schlegel, 2009). At the same time, treatment for the disease remains a complex undertaking, and success in treatment still puts many medical experts in dilemma. In recent times, the approach of dealing with the disease has shifted to concentration on clinical decision support system (CDSS), which possesses the potential and ability to increase health care quality when applied in treatment of hypertension (Dossel, Dossel and Schlegel, 2009). As a result, there have emerged some knowledge-based systems for automated guideline-based decision-support such as the ATHENA, which is a system for hypertension care (Dossel, Dossel, and Schlegel, 2009).
ATHENA Assessment and Treatment of Hypertension (Evidence-Based Automation) constitute a type of decision support system that is in clinical use for the treatment of hypertension and has been in use since 2002 (Goldstein, 2004). At the same time, the ATHENA CDSS implements guidelines from Stanford Medical Informatics EON architecture for hypertension (Goldstein, 2000). ATHENA CDSS operates similarly to other CDSS, where it helps in direct decision-making. It uses two or more patients’ data and helps in getting a diagnosis for the patient data. The patient data is put into CDSS and then result is obtained, out of which the clinician acts on the correct output by discarding the errors. This helps in obtaining a better analysis and diagnosis of the patient data (Goldstein, 2004). The development of ATHENA CDSS is a milestone that revolutionalized and is still likely to change the medical treatment and management of hypertension. The technology became a relief to many medical providers including clinicians, physicians, nurses, staff, and other health care professionals involved with treatment of hypertensive patients. ATHENA became specifically preferred due to its ability to provide proper diagnosis of the patient data (Goldstein, Hoffman, and Colem, 2000).
Usefulness/benefits of ATHENA CDSS
With the emergence of ATHENA-CDSS, it has become possible to address barriers medical professionals experienced in the treatment and management of hypertension. Such barriers include lack of provider education, confusion in dosing calculation and titration schedules, access to specific and relevant patient health information, discontinuity, documentation and access to validated assessment tools (Goldstein, 2004). At the same time, ATHENA-CDSS has resulted in elimination of some of the barriers that existed with convectional CDSS systems (Goldstein, 2004). Such barriers include constraints clinicians face with regard to time coupled with competing priorities in primary care, discomfort experienced when communicating with patients during care provision, and absence of evidence to guide hypertension prescribing decisions where all combined to make treatment and management of hypertension difficult process (Goldstein, 2004).
Overall, ATHENA has evolved as beneficial technology in the treatment and management of hypertension, which at the same time has eliminated former problems that existed in the field. The benefits can be associated with the ability of ATHENA CDSS to change its knowledge base. In most cases, the knowledge base generally assigns the eligibility criteria, risk stratification, blood pressure targets, relevant comorbid disease, guideline-recommended drug classes for patients with the comorbid disease, preferred drugs with each drug class, and the clinical message (Goldstein, 2004). Other notable benefits of the ATHENA include clinicians and medical professionals using the technology are able to get information quickly, and further, the cost of overall procedure is reduced since emphasis is put on the most relevant and necessary task, while costs associated with unnecessary care are avoided (Goldstein, Hoffman, and Colem, 2000). More so, ATHENA CDSS provides information about the latent along with the obvious need to the clinicians at the time of need (Goldstein, Hoffman, and Colem, 2000). As a result, a combination of the above benefits has increased the quality and efficiency of health care, which in turn has increased the safety of the patients (Coiera, 2003).
Characteristics of ATHENA CDSS
Numerous characteristics or elements have been identified that make ATHENA technology the most appropriate functional technology in the health care process. First, ATHENA CDSS is a guideline-based system, which operates from guidelines that have international recognition and application in the treatment and management of hypertension. This aspect manifests in two components: hypertension knowledge base modeled in Protégé and possession of guideline interpreter (Coiera, Jack-Li, and Fieschi, 2004). At the same time, ATHENA CDSS has been constructed in such a way that its basic concerns fulcrum at making recommendations that are used in delivering a specific list of drugs recommendations, and this is meant to achieve the intention of the guideline. To reinforce the above statement, ATHENA CDSS can also be seen to have the capability of recommending the need to change the dose of the medications, specifically in three distinct scenarios. These include when the blood pressure control is inadequate, choice of therapy, and the presence of compelling comorbid conditions whereby there is additional anti-hypertensive medication without considering the blood pressure control (Coiera, Jack-Li, and Fieschi, 2004). Other characteristics of ATHENA CDSS include presence of a recommended medication, substitution of a recommended medication, increase in the dose of the recommended medication, and presentation of recommendations in the pop-up advisory window superimposed on EMR. In addition, ATHENA CDSS monitors the order of medication and makes sure it avoids unnecessary tests, aiming at reducing the cost of treatment (Coiera, Jack-Li, and Fieschi, 2004).
ATHENA CDSS as Knowledge-Based System
The development of ATHENA will require the use of a knowledge-based (human-intensive) system (Coiera, 2003). It will be automated in nature and the CDSS technology will basically be for guideline-based care. At the same time, ATHENA will be developed with capability to treat and manage hypertension, where the guideline knowledge will originate from national (JNC 6) and VA guidelines (Shankar, et al., N.d; Goldstein, Hoffman and Colem, 2000). EON architecture will be used in developing the ATHENA CDSS, where it was seen earlier that this architecture constitutes knowledge-based system that provides clinicians with decision support-based care. The EON will have two components: protégé and guideline interpreter, which will operate together and will apply information in the knowledge base to create patient-specific treatment recommendations. EON has the capability of making the review process and update of knowledge-based guidelines easy (Goldstein, Hoffman, and Colem, 2000).
Methodology to be Used in ATHENA CDSS
Different methodologies have been identified, which can be used in CDSS development and application, but in this case, generation and synthesis of information will take form of IF-THEN rules, probabilities, and other related rules. IF-THEN rules posit that when a particular condition has the ability to reach certain envisioned degree of validity or it is construed to be true, then the clinician becomes alert and appropriate conclusion is generated. Accordingly, application of IF-THEN rule in this particular manner helps to prevent the duplicate test ordering. Such process can be captured in the following example; if a blood sample is introduced for testing and it changes slowly then it sends an alert to the physician when that blood test is ordered in the previous 48 hours. Therefore, other knowledge bases like probabilistic association of signs and symptoms with diagnosis or known drug-drug or drug-food interactions can be included.
How ATHENA CDSS Promotes Evidence-Based Medicine
There is always a call for professionals to practice evidence-based medicine, which incorporates and integrates holistic aspects of the medical conditions involving the patient. As a result, evidence-based medicine has evolved as a systematic approach that is normally adopted in dealing with any particular medical condition, while at same time ensuring furtherance of appraisal with an aim of applying and adopting only the must-know recommendations. In this way, evidence-based medicine practice provides the necessary help to the practitioner to tackle a patient’s situation by incorporating evidence from real patients, real challenges, and real solutions. Treating hypertension is a complex process that requires wide knowledge and information. Nevertheless, the incorporation of ATHENA CDSS has become of great help where achieving evidence-based practice in treatment of hypertension relies on identifying multiple sources where ATHENA can extract information.
Evidence-based practice will be achieved through initial development and in the progress of the practice of treatment and management of hypertension. In this way, it will be necessary to identify some of the sources that are specific in nature and possess the ability to enable ATHENA CDSS to obtain and promote the evidence-based medical practice. Such sources may include identified databases such as Cochrane Database of System reviews; Dynamed, PubMed Clinical Queries, TRIP database, and many more.
Logic rules to be used in ATHENA CDSS
Clinical decision support system operates based on logic rules, which make the functioning of the specific system more productive and easy to operate. For instance, there exist logic rules such as the Set Theory, Boolean Logic, Probability, Bayesian Logic, and more. In the case of hypertension treatment using ATHENA CDSS, rules of Bayesian probabilistic logic will be utilized. In this way, perception is on the Bayesian network, which is seen to be appropriate and functional within ATHENA CDSS. Bayesian network identifies the probabilistic relations between different patient variables such as symptoms and causes, or diseases. As a result, it is always envisioned that, it is possible to reconstruct the most probable cause when the effects are known (Greenes, 2007). Therefore, Bayesian probabilistic logic will be the primary core engine of ATHENA CDSS and in this way, it is perceived that it will provide necessary help to clinicians in determining whether hypertensive patients suffer from other related diseases. Through the process of analyzing symptoms, it will be possible for the inference engine to identify the disease, which is most probable. Furthermore, ATHENA CDSS to be developed will reflect characteristics of Boolean logic, especially with regard to IF-THEN rules. In this way, the system will be able to determine the set of actions based on specific symptoms. Therefore, it can be stated that ATHENA CDSS to be developed will combine and utilize elements of Bayesian and Boolean logic, and through this, the clinical decision support system will be more efficient and productive in generating useful information.
Use of clinical guidelines in ATHENA CDSS
Any meaningful system should be based on identified and appropriate guidelines. In the available case of treating hypertension, there is a need to develop and describe the most appropriate practices that can be incorporated in the treatment of hypertensive patients. As a result, the system to be developed will be premised on the recommendations developed and provided by the Institute for Clinical Systems Improvement established in 2010. Preference for these guidelines has to do with their ability to provide specific and clear-cut instructions for risk assessment and treatment of hypertensive patients. As a result, the overall perception is that adoption of these guidelines provides clinicians with opportunity to determine the stages involved in treatment of hypertension and at same time, provide drug therapy process (Institute for Clinical Systems Improvement, 2010). Furthermore, the guidelines provided can be translated into the IF-THEN construct. In this way, the practitioner is presented with an opportunity to carry out holistic assessment of the patient in terms of identifying risks and choosing the appropriate treatment mode.
Use of Medical Vocabularies and Ontologies in ATHENA CDSS
Appropriate and accurate functioning of any medical system requires adoption and utilization of vocabularies and ontologies that are easy to understand and comprehend. As a result, the use of the Systematized Nomenclature of Clinical or Medical Terms has been adopted as the best way of ensuring medical systems achieve their purpose. The essence of this is to make sure that medical terms used do not result in system ambiguity and that the computer has the capability to analyze and process them. In the case of hypertension treatment, medical terms will be categorized into several groups that include disease and their symptoms, different drugs and their side effects, various forms of therapy, and medical tests (Jones, 2004). The emphasis will primarily center on hypertension and related diseases. Moreover, adopting ontological approach allows one to appropriately link various terms and concepts so that the users are able to quickly retrieve the information that they need. The main advantage of this approach is that various medical terms will be related like nodes in a graph.
ATHENA-HTN CDSS Model
The above model depicts how ATHENA CDSS can be developed and operated, where the aim is to generate appropriate medical recommendations for the treatment of hypertension and related diseases. As the model indicates, the process begins with the clinician putting query in the computer system and the information is interpreted in the execution engine. Therefore, the engine becomes the pivotal point for information absorption, processing, and execution. The execution engine relies on three critical components: ATHENA HTN Guideline Knowledge Base; Electronic Medical Record System Patient Data; and SQL Server relational database that operates in front and back simulation hence feeding the execution engine with necessary information. Information relayed in these three databases can be categorized into three classes, represented by the three models: Strategy model, Communication model, and Clients model. When a query is made, the system, through the execution engine, consults the strategy model with the aim to identify the information necessary to generate elements of recommendations. To achieve this, the strategy model has to further consult client’s model, which in essence identifies the clients who can provide the information required.
The three models operate at a level known as strategy layer where information to generate recommendations is developed. All the information generated is then integrated into the computer system constituting the presentation layer. Here, visual clients are established, playing the role of displaying information, specifically relating to how recommendations should be utilized as well as the entire process of achieving it. The content layer thus becomes the area where different sources of information are generated and subsequently, they become the source and basis of medical recommendations and supporting information.
Conclusion
The development of ATHENA CDSS has revolutionized treatment and management processes for hypertension. Earlier medical errors that were common in the field have been eliminated and this has been good news to hypertension patients and medical professionals who have found the treatment process to be smooth and easy. Nevertheless, the current ATHENA model cannot be said to be perfect, rather, there is a need for future efforts to improve the system so that recommendations generated have wide array of applications, and their application is more achievable. This requires more research work to establish the appropriate features that can be included in the model, which of course originates from the medical needs of hypertension patients and the available technology. For now, it can be said that ATHENA CDSS technology has improved overall operation, treatment, and management of hypertension and related diseases.
References
Berner, E. T. (2007). Clinical decision support systems: theory and practice. NY: Springer.
Coiera, E. (2003). The Guide to Health Informatics. London: Arnold Publishers. Web.
Coiera, E., Jack-Li, Y. C., & Fieschi, M. (2004). Medinfo 2004: proceedings of the 11th World Conference on Medical. Amsterdam: IOS Press.
Dossel, O., Dossel, O., & Schlegel, W. C. (2009). World Congress on Medical Physics and Biomedical Engineering, September 7-12, 2009 Munich, Germany: Information and Communication in Medicine, Telemedicine and E-Health. NY: Springer.
Goldstein, M. (2000). ATHENA :Assessment and Treatment of Hypertension: Evidence-Based Automation. Open Clinical knowledge management for medical care. Web.
Goldstein, M. (2004). Assessment and treatment of Hypertension: Evidence-Based Automation (ATHENA). FSI Health Policy-Center for Health Policy/ Center for Primary Care And Outcomes Research. Web.
Goldstein, M., Hoffman, B., & Colem, R. (2000). Implementing clinical practice guidelines while taking account of changing evidence: ATHENA DSS, an easily modifiable decision-support system for managing hypertension in primary Care. Web.
Goldstein, M. K. (2008). Evaluating Clinical Decision Support Systems. VA HSR&D Cyber Seminar. VA Palo Alto Health Care System and Stanford University. Web.
Greenes, R. (2007). Clinical decision support: the road ahead. NY: Elsevier. Institute for Clinical Systems Improvement. (2010). Health Care Guideline: Hypertension Diagnosis and Treatment. Web.
Jones R. (2004). Oxford textbook of primary medical care. Oxford: Oxford University Press.
Shankar, R.D., Martins, S. B., Tu, S. W., Goldstein, M. K., & Musen, M. A. (N.d). Building an explanation function for a Hypertension Decision-Support System. Stanford Medical Informatics. Web.