|SYMPOSIUM: SIMULATION IN MEDICAL EDUCATION
|Year : 2017 | Volume
| Issue : 1 | Page : 78-83
Clinical decision support systems: From medical simulation to clinical practice
Scott M Pappada1, Thomas J Papadimos2
1 Department of Anesthesiology, University of Toledo College of Medicine and Life Sciences; Department of Bioengineering, University of Toledo College of Engineering; Department of Anesthesiology, Medical Director of the Lloyd Jacobs Simulation Center, University of Toledo College of Medicine and Life Sciences, Toledo, OH, USA
2 Department of Anesthesiology, University of Toledo College of Medicine and Life Sciences; Department of Bioengineering, University of Toledo College of Engineering, Toledo, OH, USA
|Date of Web Publication||7-Jul-2017|
Thomas J Papadimos
3000 Arlington Avenue, Toledo, OH 43614
Source of Support: None, Conflict of Interest: None
Medical simulation has become an integral part of the training of medical students, residents, and faculty. While our traditional sense of medical simulation involves scenarios, manikins, resuscitation, and procedures, we must now expand this vision. With the development of the electronic medical record and the resultant large data sets (“Big Data”) that encompass each patient's record, there is an opportunity to engage in the modeling and prediction of individual patient outcomes and trajectories of care. Here, medical simulation delves into the arena of clinical decision support systems (CDSSs) which leverage advanced algorithms, analytics, and machine learning approaches. The advent and continued adoption of health information technologies such as CDSSs into real-world health-care operations can lead to improved patient care, safety, and cost savings.
The following core competencies are addressed in this article: System-based practice, Practice-based learning, and Patient care.
Keywords: Clinical decision support systems, electronic health records, medical education, patient simulation
|How to cite this article:|
Pappada SM, Papadimos TJ. Clinical decision support systems: From medical simulation to clinical practice. Int J Acad Med 2017;3:78-83
|How to cite this URL:|
Pappada SM, Papadimos TJ. Clinical decision support systems: From medical simulation to clinical practice. Int J Acad Med [serial online] 2017 [cited 2019 Dec 14];3:78-83. Available from: http://www.ijam-web.org/text.asp?2017/3/1/78/209840
| Introduction|| |
Conventionally, medical simulation is regarded as taking a learner(s) through a series of procedures or scenarios to improve their skill set and/or determine deficiencies or strengths as individuals or teams. Clinical decision support systems (CDSSs), on the other hand, involve the processing of some form of electronic information or health record to make “simulations” or predictions regarding care and/or to recommend a specific treatment regimen. Specifically, a CDSS is “any computer program designed to help health professionals make a clinical decision.” This definition includes “sophisticated platforms to store and manage clinical data, and tools to alert clinicians of problematic situations, or decision-making tools to assist clinicians by providing patient-specific recommendations” through the use of electronic health records (EHRs). Even though some authors report that, currently, CDSSs integrated with EHRs might not affect mortality, they do demonstrate CDSSs' moderate morbidity outcomes, thereby providing a positive perspective regarding future research into the topic. Herein, we will discuss CDSSs as a form or implementation of medical simulation and the importance of their development to alleviating the overall workload (i.e., both cognitive and physical demands) placed on health-care providers (HCPs) in the Intensive Care Unit (ICU).
| Discussion|| |
The importance of electronic health records and health information technology systems in informing health-care providers and guiding health-care delivery
The management of information related to health-care delivery is critical to optimizing the quality of care. Quality of care can be defined as a multidimensional construct that describes the overall value of health-care delivery in terms of three factors and dimensions: (1) quality, (2) efficiency, and (3) cost-effectiveness. The federal government has invested billions of dollars in incentives to promote the “meaningful use” of EHRs. The intent is for electronic EHRs to be utilized not merely for retrospective analysis but leveraged in real time to improve health-care delivery and quality of patient care. While the adoption of EHRs and other health information technology (HIT) systems by health-care institutions and organizations across the United States provides the groundwork for this lofty objective to be reached, research and efforts in this area still have significant potential for growth. Previous efforts have ranged from defining somewhat simplistic quality of care metrics  to implementing models and algorithms designed to predict key patient outcomes such as hospital readmissions. Leveraging EHRs on a real-time basis through CDSSs is becoming viewed as increasingly valuable in improving health-care delivery and overall patient safety, especially in high-risk settings such as the ICU.
The growth of health information technology systems and the resultant cognitive overload in the Intensive Care Unit
There continues to be a considerable growth in the implementation and use of HIT and HIT systems across in- and out-patient health-care settings. Examples of specific types of HIT systems which have been become commonplace include EHRs, computerized provider order entry, electronic prescribing, and CDSSs. Due to the massive amount of information collected and generated by HIT systems and presented to HCPs, cognitive and information overload are rapidly becoming pervasive problems within today's health-care settings. Cognitive overload and information overload can lead to suboptimal and at times inappropriate treatment decisions which can mean the difference between life and death in certain high-risk clinical settings such as the ICU. In the ICU, health-care providers are required to rigorously monitor patient data that are distributed across a number of information systems. Frequently changing patient conditions require health-care professionals to closely monitor and assess their patients and reprioritize their treatment approaches. To identify, track, and interpret all relevant patient data sources is both time consuming and labor intensive. As such, cognitive and information overload resulting from the aforementioned operational conditions can result in significant reductions in patient safety through an increased potential for medical errors, operational inefficiencies, and the increased incidence of adverse events and patient outcomes. While order entry and prescribing (through EHRs) is important, it is the role of CDSSs in assisting in clinical decision-making that can perhaps best alleviate the substantial cognitive load placed on HCPs in the ICU, thereby contributing to better outcomes.
Clinical decision support systems, simulation, and improved Intensive Care Unit care
Today's health-care operations will continue to evolve due to the continued adoption and utilization of technology and HIT systems [Figure 1]. Technology is advancing, not only real-world health-care operations, but the training of health-care professionals as well. One of the more recent trends in health care is the increased adoption and utilization of medical simulation to train knowledge and skills among health-care providers. In fact, the use of medical simulation is becoming increasingly routine and many educators within the health-care community deem it as necessary. Medical simulation offers a unique “life-like,” realistic, yet risk-free environment, in which learners can be pushed to their limits without any harm coming to the patient. We postulate that CDSSs, in fact, can be viewed as a form of medical simulation. CDSSs can provide a “simulation-like” approach that health-care providers can leverage to improve upon their real-world clinical decision-making and health-care delivery as shown in [Figure 1]. CDSSs are designed with a combination of algorithms, analytics, models, and/or logic which process the various data sources [nonlimiting list shown in [Figure 1] stored within an EHR to generate treatment recommendations and guide care. Overall, the intended functionality of CDSSs is to “simulate” the thinking, reasoning, and decision-making of an expert HCP given changes in patient data and status that occur over time. Most CDSSs are specifically designed to guide treatment that is routinely managed by a specialized HCP. For example, CDSSs have been developed and applied to guiding antibiotic therapy that is routinely managed by an infectious disease physician and/or pharmacist., In addition. CDSSs, as a form of medical simulation, can be used as educational tools to instill better decision-making and situational awareness in HCPs. As a testament to this trend, a recent study leveraged a point-of-order CDSS during a radiology undergraduate medical education curriculum that involved medical simulation.
|Figure 1: Clinical decision support systems leveraging various data sources from the electronic medical record to simulate and guide decision-making of health-care providers|
Click here to view
CDSSs can be grouped into two broad categories: knowledge-based and nonknowledge-based systems. Knowledge-based CDSSs have resulted from prior research surrounding the development of “expert systems.” In this context, “expert systems” are defined as a computer program that is intended to reproduce, replicate, or simulate human thinking and reasoning. Knowledge-based CDSSs have since evolved from the initial “expert system” use case to assisting the clinical decision-making of health-care providers. The design of knowledge-based CDSSs usually consists of a knowledge base and an inference or reasoning module. The knowledge base is often, but not always, compiled with information that is organized in the form of logic functions such as IF-THEN or IF-ELSE statements. The inference or reasoning module usually comprises equations or formulae that suggest an intervention or treatment (e.g., dosage of drug to deliver) given patient data entered or received by the CDSS from the EHR in real time or near real time [Figure 2].
|Figure 2: Big Data in health care and the need for clinical decision support systems with machine learning and computational approaches to identify subtle patterns and trends in relevant data to guide and optimize health-care delivery|
Click here to view
Contrary to knowledge-based CDSSs, nonknowledge-based CDSSs implement a form of artificial intelligence known as machine learning that provides computers the ability to learn without being explicitly programed. Machine learning surrounds the development of computer applications that can learn even the most subtle patterns and relationships in datasets as they are exposed to new data. In the past, HCPs have been reluctant to adopt and utilize nonknowledge-based CDSSs as there is little transparency to their recommendations due to the “black box” nature of machine learning models. However, this stance is now significantly changing, and the development, adoption, and use of nonknowledge-based CDSSs among health-care professionals continue to grow due to evolution of the present day health-care operations. As previously discussed, there is a tremendous amount of electronic health-care data being collected and generated by HIT technologies, and this predicates the need for more intelligent CDSSs that can process extremely large data repositories. The era of “Big Data” in health care , is upon us, and there will be a significant need to develop CDSSs that implement machine learning approaches which are capable of identifying patterns and trends in data that are subtle and undetectable by one or more humans. A number of information systems are distributed across health-care institutions, and for a single human to sift through and process all the relevant data sources and identify relevant patterns and trends is both unfeasible and unrealistic. As such, and as shown in [Figure 2], machine learning, advanced analytics, and algorithms, and other computational approaches will play a significant role in the future health-care technologies (including CDSSs) in optimizing HCP situational awareness, decision-making, and consequently clinical performance. Although CDSSs have been developed and used to improve health-care delivery and quality of care across a number of clinical settings (as referenced above), the remainder of this article will focus on a more narrow discussion of CDSSs and focus on those explicitly developed for the critical care setting (i.e., ICU).
Clinical decision support system development, adoption, and use in the critical care setting
The growth in ICU admissions and critical care service utilization across the aging population of the United States has resulted in an unfavorable ratio of critical care physicians (intensivists) and ICU beds across the United States. This is very problematic as prior work has alluded to the need for both an experienced and well-staffed ICU. Unfavorable patient-to-caregiver ratios and staff workload contribute to an increased risk of suboptimal health-care delivery and consequentially poorer patient outcomes. Maintenance of full-time care by an intensivist has been associated with improved ICU patient outcomes including significant reductions in mortality and ICU length of stay. Shortages of nursing staff are also a growing issue and concern in acute and critical care settings. This shortage, specifically in the ICU, has been associated with increases in hospital and ICU lengths of stay. Suboptimal nursing staff workload should be avoided at all costs as it can lead to ICU operational inefficiencies and contribute to reductions in patient safety and quality of care.
Efforts are currently ongoing by members of the critical care research community that are aimed at the transformation of ICU health-care delivery. Prior efforts have involved the development of technology to improve the visualization of data from the electronic medical record (EMR), through the development of intelligent CDSSs and patient monitoring technologies. Specific functionalities include alerting staff, guiding treatment, improving adherence to treatment guidelines/procedures, identifying at-risk patients, and predicting patient outcomes.,, In addition, there are efforts to develop open database architectures that provide data visualizations and CDSSs real-time/near real-time access to EMR data. As research in this area continues, we inch closer to the ultimate goal of developing a vision of the “ICU of the future.” In this vision, CDSSs and other forms of technology will serve a key role in optimizing the ability of health-care professionals to deliver safe and effective care for/to their patients.
In today's growing data-driven society, and as HIT systems continue to arise and expand across all health-care settings, not just the ICU, but also CDSSs can serve as extremely useful tools for health-care professionals. Specifically, they will serve to mitigate the incidence of cognitive and information overload across HCPs by eliminating and simplifying some of the information processing required by health-care professionals. To this end, CDSSs will serve to reduce the overall workload (both physical and mental) required by HCPs during their routine day-to-day operations. Furthermore, CDSSs will provide a substitute for expertise when it remains unavailable in resource- and personnel-constrained health-care settings that are growing in their prevalence across health-care institutions in the United States (and the world). Today, there is an increased utilization of health-care services that is a result of an aging U.S. population that will require well-resourced health-care institutions and an increased availability of health-care staff. Unfortunately, the current state of health-care operations in the United States does not provide an ideal environment that optimizes treatment and patient care practices in a straightforward fashion. In 1999, the Institute of Medicine issued a report entitled, To Err is Human: Building A Safer Health System, which has spurred a growing interest in the understanding of clinical operations and the workload experienced by HCPs during everyday patient treatment and care. Factors which are related to and directly affect workload in the health-care settings include: the number of patients being cared for at a given time, the patients' acuity, and communication among team members. Such factors are correlated with the quality of care and overall patient outcomes. Appropriate institutional resources and staffing levels are extremely important in high-risk settings such as the ICU. In the ICU, health-care professionals routinely experience a variety of highly stressful and emotional situations. CDSSs will serve as a useful tool and resource for health-care providers to guide treatment and care, given the absence of expertise poor staff and resource availability.
| Conclusion|| |
While the implementation of standardized protocols and guidelines has been shown to be effective in guiding treatment and care for certain patient populations, modern medicine and clinical practice are rapidly changing.,,, Many health-care providers today are coming to the realization that following standardized protocols and clinical practice guidelines is not always the optimal solution. In today's medicine, one size definitely does not fit all, and each patient can and will likely respond differently to treatments and have their own unique set of requirements for care. As such, medicine is evolving from implementation of standardized care to a more “personalized” medicine approach. Under this paradigm, patient care practices will be targeted to an individual patient rather than a cohort or group of patients meeting some common criteria. Given the increased availability of data generated by HIT systems as described above and demonstrated in [Figure 1] and [Figure 2], there is considerable potential for numerous sources of data to be processed and presented to an individual HCP to influence and improve their clinical decision-making and consequently, the overall quality of care they deliver. It is in today's current health-care environment that technology such as CDSSs will play a key role in ensuring that our patients achieve the highest quality of health care possible.
At the same time, modeling large health-care data sets (“Big Data”) can provide improved organizational outcomes that can go hand in hand with improved clinical outcomes. These efforts now range from extracorporeal membrane oxygenation to oncology and from the clinical environment to the basic sciences., We encourage multi-institutional efforts to further the research into predictive modeling of EHR information, and other information data sets, through the development of novel CDSSs. These CDSSs will contribute to the enhancement of patient safety and provide HCPs valuable tools to improve the quality of patient care.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| References|| |
Abrahamson S, Denson JS, Wolf RM. Effectiveness of a simulator in training anesthesiology residents. J Med Educ 1969;44:515-9.
Marcos M, Maldonado JA, Martínez-Salvador B, Boscá D, Robles M. Interoperability of clinical decision-support systems and electronic health records using archetypes: A case study in clinical trial eligibility. J Biomed Inform 2013;46:676-89.
Cimino JJ, Shortliffe EH. Biomedical Informatics: Computer Applications in Health Care and Biomedicine (Health Informatics). Washington, D.C.: Springer-Verlag New York, Inc.; 2006.
Moja L, Kwag KH, Lytras T, Bertizzolo L, Brandt L, Pecoraro V, et al.
Effectiveness of computerized decision support systems linked to electronic health records: A systematic review and meta-analysis. Am J Public Health 2014;104:e12-22.
Chassin MR, Galvin RW. The urgent need to improve health care quality. Institute of Medicine National Roundtable on Health Care Quality. JAMA 1998;280:1000-5.
Blumenthal D, Tavenner M. The “meaningful use” regulation for electronic health records. N Engl J Med 2010;363:501-4.
Marcotte L, Seidman J, Trudel K, Berwick DM, Blumenthal D, Mostashari F, et al.
Achieving meaningful use of health information technology: A guide for physicians to the EHR incentive programs. Arch Intern Med 2012;172:731-6.
Romano MJ, Stafford RS. Electronic health records and clinical decision support systems: Impact on national ambulatory care quality. Arch Intern Med 2011;171:897-903.
Kansagara D, Englander H, Salanitro A, Kagen D, Theobald C, Freeman M, et al.
Risk prediction models for hospital readmission: A systematic review. JAMA 2011;306:1688-98.
Carayon P, Alvarado CJ; Systems Engineering Initiative for Patient Safety. Workload and patient safety among critical care nurses. Crit Care Nurs Clin North Am 2007;19:121-9.
Beaudoin M, Kabanza F, Nault V, Valiquette L. Evaluation of a machine learning capability for a clinical decision support system to enhance antimicrobial stewardship programs. Artif Intell Med 2016;68:29-36.
Shortliffe EH, Axline SG, Buchanan BG, Merigan TC, Cohen SN. An artificial intelligence program to advise physicians regarding antimicrobial therapy. Comput Biomed Res 1973;6:544-60.
Willis MH, Frigini LA, Lin J, Wynne DM, Sepulveda KA. Clinical decision support at the point-of-order entry: An education simulation pilot with medical students. Acad Radiol 2016;23:1309-18.
Holst H, Aström K, Järund A, Palmer J, Heyden A, Kahl F, et al.
Automated interpretation of ventilation-perfusion lung scintigrams for the diagnosis of pulmonary embolism using artificial neural networks. Eur J Nucl Med 2000;27:400-6.
Murdoch TB, Detsky AS. The inevitable application of big data to health care. JAMA 2013;309:1351-2.
Krumholz HM. Big data and new knowledge in medicine: The thinking, training, and tools needed for a learning health system. Health Aff (Millwood) 2014;33:1163-70.
Dara SI, Afessa B. Intensivist-to-bed ratio: Association with outcomes in the medical ICU. Chest 2005;128:567-72.
Rollins G. ICU care management by intensivists reduces mortality and length of stay. Rep Med Guidel Outcomes Res 2002;13:5-7.
Stechmiller JK. The nursing shortage in acute and critical care settings. AACN Clin Issues 2002;13:577-84.
Pronovost PJ, Jenckes MW, Dorman T, Garrett E, Breslow MJ, Rosenfeld BA, et al.
Organizational characteristics of Intensive Care Units related to outcomes of abdominal aortic surgery. JAMA 1999;281:1310-7.
Carayon P, Gurses A. Nursing workload and patient safety in Intensive Care Units: A human factors engineering evaluation of the literature. Intensive Crit Care Nurs 2005;21:284-301.
Ahmed A, Chandra S, Herasevich V, Gajic O, Pickering BW. The effect of two different electronic health record user interfaces on intensive care provider task load, errors of cognition, and performance. Crit Care Med 2011;39:1626-34.
Pickering BW, Dong Y, Ahmed A, Giri J, Kilickaya O, Gupta A, et al.
The implementation of clinician designed, human-centered electronic medical record viewer in the Intensive Care Unit: A pilot step-wedge cluster randomized trial. Int J Med Inform 2015;84:299-307.
Kumar KA, Singh Y, Sanyal S. Hybrid approach using case-based reasoning and rule-based reasoning for domain independent clinical decision support in ICU. Expert Syst Appl 2009;36:65-71.
Frize M, Ennett CM, Stevenson M, Trigg HC. Clinical decision support systems for Intensive Care Units: Using artificial neural networks. Med Eng Phys 2001;23:217-25.
Sintchenko V, Iredell JR, Gilbert GL, Coiera E. Handheld computer-based decision support reduces patient length of stay and antibiotic prescribing in critical care. J Am Med Inform Assoc 2005;12:398-402.
Herasevich V, Pickering BW, Dong Y, Peters SG, Gajic O. Informatics infrastructure for syndrome surveillance, decision support, reporting, and modeling of critical illness. Mayo Clin Proc 2010;85:247-54.
Kohn LT, Corrigan JM, Donaldson MS. To Err is Human: Building a Safer Health System. Washington, D.C: National Academies Press; 2000.
Goldberg PA, Siegel MD, Sherwin RS, Halickman JI, Lee M, Bailey VA, et al.
Implementation of a safe and effective insulin infusion protocol in a medical Intensive Care Unit. Diabetes Care 2004;27:461-7.
Barr J, Hecht M, Flavin KE, Khorana A, Gould MK. Outcomes in critically ill patients before and after the implementation of an evidence-based nutritional management protocol. Chest 2004;125:1446-57.
Marshall J, Finn CA, Theodore AC. Impact of a clinical pharmacist-enforced Intensive Care Unit sedation protocol on duration of mechanical ventilation and hospital stay. Crit Care Med 2008;36:427-33.
Arabi YM, Haddad S, Tamim HM, Al-Dawood A, Al-Qahtani S, Ferayan A, et al.
Mortality reduction after implementing a clinical practice guidelines-based management protocol for severe traumatic brain injury. J Crit Care 2010;25:190-5.
[Figure 1], [Figure 2]