OpenClinical logo

Decision support systems

Clinical decision support systems (CDSSs) form a significant part of the field of clinical knowledge management technologies through their capacity to support the clinical process and use of knowledge, from diagnosis and investigation through treatment and long-term care.
spacer
contents
 bullet  Definition  bullet  Functions  bullet  Early AI/Decision Support Systems  bullet  Myths affecting the development of DSS  bullet  AI / Decision Support Systems in clinical practice  bullet  References

 bullet  AI systems in clinical practice  bullet  Public reports on decision support  bullet  DSS: Benefits, Drawbacks, Success factors  bullet  DSS: Evaluation  bullet  DSS: Evaluation studies  bullet  DSS: Evaluation reviews
spacer

Definition
Clinical Decision Support Systems are "active knowledge systems which use two or more items of patient data to generate case-specific advice"  [Wyatt J, Spiegelhalter D, 1991].

Clinical DSSs are typically designed to integrate a medical knowledge base, patient data and an inference engine to generate case specific advice.

Functions
Four key functions of electronic clinical decision support systems are outlined in [Perreault & Metzger, 1999]:
  1. "Administrative: Supporting clinical coding and documentation, authorization of procedures, and referrals.
  2. "Managing clinical complexity and details: Keeping patients on research and chemotherapy protocols; tracking orders, referrals follow-up, and preventive care.
  3. "Cost control: Monitoring medication orders; avoiding duplicate or unnecessary tests.
  4. "Decision support: Supporting clinical diagnosis and treatment plan processes; and promoting use of best practices, condition-specific guidelines, and population-based management. "
Early AI/Decision Support Systems.
Research into the use of artificial intelligence in medicine started in the early 1970's and produced a number of experimental systems.
AAPHelp: de Dombal's system for acute abdominal pain (1972) An early attempt to implement automated reasoning under uncertainty. De Dombal's system, developed at Leeds University, was designed to support the diagnosis of acute abdominal pain and, based on analysis, the need for surgery. The system's decision making was based on the naive Bayesian approach.

INTERNIST I (1974)

Pople and Myers begin work on INTERNIST, one of the first clinical decision support systems, designed to support diagnosis, in 1970.

INTERNIST-I was a rule-based expert system designed at the University of Pittsburgh in 1974 for the diagnosis of complex diagnosis of complex problems in general internal medicine. It uses patient observations to deduce a list of compatible disease states (based on a tree-structured database that links diseases with symptoms). By the early 1980s, it was recognized that the most valuable product of the system was its medical knowledge base. This was used as a basis for successor systems including CADUCEUS and Quick Medical Reference (QMR), a commercialised diagnostic DSS for internists.
MYCIN (1976) MYCIN was a rule-based expert system designed to diagnose and recommend treatment for certain blood infections (antimicrobial selection for patients with bacteremia or meningitis). It was later extended to handle other infectious diseases. Clinical knowledge in MYCIN is represented as a set of IF-THEN rules with certainty factors attached to diagnoses. It was a goal-directed system, using a basic backward chaining reasoning strategy (resulting in exhaustive depth-first search of the rules base for relevant rules though with additional heuristic support to control the search for a proposed solution). MYCIN was developed in the mid-1970s by Ted Shortliffe and colleagues at Stanford University. It is probably the most famous early expert system, described by Mark Musen as being "the first convincing demonstration of the power of the rule-based approach in the development of robust clinical decision-support systems" [Musen, 1999].

The EMYCIN (Essential MYCIN) expert system shell, employing MYCIN's control structures was developed at Stanford in 1980. This domain-independent framework was used to build diagnostic rule-based expert systems such as PUFF, a system designed to interpret pulmonary function tests for patients with lung disease.
CASNET/Glaucoma CASNET (Causal ASsociational NETworks), developed in the 1960s, was a general tool for building expert system for the diagnosis and treatment of diseases. The most significant Expert System application based on CASNET was CASNET/Glaucoma for the diagnosis and treatment of glaucoma.

Expert clinical knowledge was represented in a causal-associational network (CASNET) model for describing disease processes. CASNET/Glaucoma was developed at Rutgers University and implemented in FORTRAN.
PIP PIP, the Present Illness Program, was a system built by MIT and Tufts-New England Medical Center in the 1970s that gathered data and generated hypotheses about disease processes in patients with renal disease.
ABEL Acid-Base and ELectrolyte program. An expert system, employing causal reasoning, for the management of electrolyte and acid base derangements. Developed at the Laboratory for Computer Science, MIT, in the early 1980s.
ONCOCIN A rule-based medical expert system for oncology protocol management developed at Stanford University. Oncocin was designed to assist physicians with the treatment of cancer patients receiving chemotherapy. ONCOCIN was one of the first DSS which attempted to model decisions and sequencing actions over time, using a customised flowchart language. It extended the skeletal-planning technique to an application area where the history of past events and the duration of actions are important.
Some successful systems originating in the 1980s were commercialised.
DXplain logo
"DXplain is a decision support system which uses a set of clinical findings (signs, symptoms, laboratory data) to produce a ranked list of diagnoses which might explain (or be associated with) the clinical manifestations. DXplain provides justification for why each of these diseases might be considered, suggests what further clinical information would be useful to collect for each disease, and lists what clinical manifestations, if any, would be unusual or atypical for each of the specific diseases" [LCS MGH Harvard Medical School]. DXplain includes 2,200 diseases and 5,000 symptoms in its knowledge base.

Developed by Laboratory of Computer Science, Massachusetts General Hospital, Harvard Medical School. .

QMR

Quick Medical Reference

"A diagnostic decision-support system with a knowledge base of diseases, diagnoses, findings, disease associations and lab information. With information from the primary medical literature on almost 700 diseases and more than 5,000 symptoms, signs, and labs." QMR was designed for 3 types of use: "as an electronic textbook; as an intermediate level spreadsheet for the combination and exploration of simple diagnostic concepts; as an expert consultant program" [Miller RA, 1989].

Developed (1980) by the University of Pittsburgh and First Databank, California.

links
 bullet  Demonstration of DXplain [OC]
Myths affecting the development of DSS
Ted Shortliffe has challenged three basic assumptions which strongly influenced the development of decision support systems (particularly early systems) and which he now terms myths:
  1. "Diagnosis is the dominant decision-making issue in medicine"
  2. "Clinicians will use knowledge-based systems if the programs can be shown to function at the level of experts"
  3. "Clinicians will use stand-alone decision-support tools."

By implication, these myths, which are gradulally being overcome, partly contributed to the relative lack of success of DSS in clinical care.

[Ted Shortliffe, Medical Thinking Meeting, London, June 2006]
AI / Decision Support Systems in clinical practice
OpenClinical maintains an extensive archive of Artificial Intelligence systems in routine clinical use, previously administered by Enrico Coiera.

The archive contains summaries of Artificial Intelligence-based computer systems that are (or have been) in routine use in medical settings. Entries range from simple knowledge-based or expert systems to quite advanced systems capable of performing complex inferences. Individual systems are grouped by category: Acure Care Systems, Decision Support Systems, Educational Systems, Laboratory Systems, Medical Imaging Systems, Quality Assurance and Administration Systems.
References: general
E. Coiera. The Guide to Health Informatics (2nd Edition). Arnold, London, October 2003.

Free sample chapters include: [Chapter 25 - Clinical decision support systems]

Coiera, 2nd Edition, 2003 Chapter summary: " 1. Artificial intelligence (AI) systems are intended to support healthcare workers with tasks that rely on the manipulation of data and knowledge. 2. Expert systems are the commonest type of CDSS in routine clinical use. They contain medical knowledge about a very specifically defined task. Their uses include: ˇ alerts and reminders, ˇ diagnostic assistance, ˇ therapy critiquing and planning, ˇ prescribing decision support ˇ information retrieval, ˇ image recognition and interpretation. 3. Reasons for the failure of many expert systems to be used clinically include dependence on an electronic medical record system to supply their data, poor human interface design, failure to fit naturally into the routine process of care, and reluctance or computer illiteracy of some healthcare workers. 4. Many expert systems are now in routine use in acute care settings, clinical laboratories, educational institutions, and incorporated into electronic medical record systems. 5. Some CDSS systems have the capacity to learn, leading to the discovery of new phenomena and the creation of medical knowledge. These machine learning systems can be used to: ˇ develop the knowledge bases used by expert systems, ˇ assist in the design of new drugs, ˇ advance research in the development of pathophysiological models from experimental data. 6. Benefits from CDSS include improved patient safety, improved quality of care, and improved efficiency in health care delivery. "

Sintchenko V, Westbrook J, Tipper S, Mathie M, Coiera E. Electronic decision support activities in different healthcare settings in Australia. Commissioned study published as Appendix A of Electronic Decision Support for Australia's Health Sector, Report to Health Ministers by the National Electronic Decision Support Taskforce, January 2003.

[Australian Health Information Council - Full report]

" Objectives of this report are to:
  • detail the current status of electronic decision support implementation world-wide, identifying significant decision support systems in operation, the audience for those systems, information on implementation difficulties and rates of uptake
  • present an overview of the evidence of the effectiveness of electronic decision support systems in improving clinical outcomes
  • prepare an inventory of large-scale and/or significant electronic decision support activities in Australia
  • identify factors that are critical to ensuring successful development of electronic decision support systems on a national basis and barriers to the successful implementation of such systems that need to be addressed in the Australian context
  • describe the likely benefits of a more coordinated approach to the development of electronic decision support systems for clinicians and other key stakeholders. "
  • Berlin A, Sorani M, Sim I. A taxonomic description of computer-based clinical decision support systems. J Biomed Inform. 2006 Dec;39(6):656-67. Epub 2006 Jan 9.

    [PubMed]   []

    " OBJECTIVE: Computer-based clinical decision support systems (CDSSs) vary greatly in design and function. Using a taxonomy that we had previously developed, we describe the characteristics of CDSSs reported in the literature. METHODS: We searched PubMed and the Cochrane Library for randomized controlled trials (RCTs) published in English between 1998 and 2003 that evaluated CDSSs. We coded each CDSS using our taxonomy. RESULTS: 58 studies met our inclusion criteria. The 74 reported CDSSs varied greatly in context of use, knowledge and data sources, nature of decision support offered, information delivery, and workflow impact. Two distinct subsets of CDSSs were seen: patient-directed systems that provided decision support for preventive care or health-related behaviors via mail or phone (38% of systems), and inpatient systems targeting clinicians with online decision support and direct online execution of the recommendations (18%). 84% of the CDSSs required extra staffing for handling CDSS-related input or output. CONCLUSION: Reported CDSSs are heterogeneous along many dimensions. Caution should be taken in generalizing the results of CDSS RCTs to different clinical or workflow settings. "

    Perreault L, Metzger J. A pragmatic framework for understanding clinical decision support. Journal of Healthcare Information Management. 1999;13(2):5-21. Four key functions of DSSs are outlined: (i)Administrative: Supporting clinical coding and documentation, authorization of procedures, and referrals. (ii)Managing clinical complexity and details: Keeping patients on research and chemotherapy protocols; tracking orders, referrals follow-up, and preventive care. (iii)Cost control: Monitoring medication orders; avoiding duplicate or unnecessary tests. (iv)Decision support: Supporting clinical diagnosis and treatment plan processes; and promoting use of best practices, condition-specific guidelines, and population-based management.

    Trivedi MH, Kern JK, Marcee A, Grannemann B, Kleiber B, Bettinger T, Altshuler KZ, McClelland A. Development and implementation of computerized clinical guidelines: barriers and solutions. Methods Inf Med. 2002;41(5):435-42.

    [PubMed]   []

    " Research indicates that computerized decision support systems (CDSSs) can improve clinical performance and patient outcomes, and yet CDSSs are not in widespread use. Physician guidelines, in general, face barriers in implementation. Guidelines in a computerized format can overcome some of the barriers to conventional text-form guidelines; however, computerized programs have novel aspects that have to be considered, aspects such as technical problems/support and user interface issues that can act as barriers. Though the literature points out that human, organizational, and technical issues can act as barriers in the implementation of CDSSs, studies clearly indicate that there are methods that can overcome these barriers and improve CDSS acceptance and use. These methods come from lessons learned from a variety of CDSS implementation ventures. Notably, most of the methods that improve acceptance and use of a CDSS require feedback and involvement of end-users. Measuring and addressing physician or user attitudes toward the computerized support system has been shown to be important in the successful implementation of a CDSS. This article discusses: 1) the barriers of implementation of guidelines in general and of CDSSs; 2) the importance of the physician's role in development, implementation, and adherence; 3) methods that can improve CDSS acceptance and use; and 4) the types of tools needed to obtain end-user feedback. "

    Wong HJ, Legnini MW, Whitmore HH. The diffusion of decision support systems in healthcare: are we there yet? J Healthc Manag. 2000 Jul-Aug;45(4):240-9; discussion 249-53.

    [PubMed]

    " Clinical decision support (CDS) systems, with the potential to minimize practice variation and improve patient care, have begun to surface throughout the healthcare industry. This study reviews historic patterns of information technology (IT) in healthcare, analyzes barriers and enabling factors, and draws three lessons. First, the widespread adoption of clinical IT, including CDS systems, depends on having the right organization and individual financial incentives in place. Second, although CDS systems and clinical IT in general are powerful tools that can be used to support the practice of medicine, they alone cannot redefine the workflow or process within the profession, Healthcare managers counting on technology to restructure or monitor clinicians' work patterns are likely to encounter substantial resistance to CDS systems, even those that generate valuable information. Third, while the pace of implementing IT systems in healthcare have lagged behind that of other industries, many of he obstacles are gradually diminishing. However, several factors continue to inhibit their widespread diffusion, including the organizational turmoil created by large numbers of mergers and acquisitions, and the lack of uniform data standards. "

    Fieschi M, Dufour JC, Staccini P, Gouvernet J, Bouhaddou O. Medical decision support systems: old dilemmas and new paradigms? Methods Inf Med. 2003;42(3):190-8.

    [PubMed]   [schattauer.com]

    " The purpose of this paper is to examine past and present medical decision support systems and the environment in which they operate and to propose specific research tracks that improve integration and adoption of these systems in today's health care systems. METHODS: In preamble, we examine the objectives, decision models, and performances of past decision support systems. RESULTS: Medical decision support tools were essentially formulated from a technical capability perspective and this view has met limited adoption and slowed down new development as well as integration of these important systems into patient management work flows and clinical information systems. The science base of these systems needs to include evidence-based medicine and clinical practice guidelines and the paradigms need to be extended to include a collaborative provider model, the users and the organization perspectives. The availability of patient record and medical terminology standards is essential to the dissemination of decision support systems and so is their integration into the care process. CONCLUSION: To build new decision support systems based on practice guidelines and taking into account users preferences, we do not so much advocate new technological solutions but rather suggest that technology is not enough to ensure successful adoption by the users, the integration into practice workflow, and consequently, the realisation of improved health care outcomes. "
    Sim I, Sanders GD, McDonald KM. Evidence-based practice for mere mortals: the role of informatics and health services research. J Gen Intern Med. 2002 Apr;17(4):302-8.

    [PubMed]   [Blackwell Publishing]

    Desiderata for CDSS: (i)Computer-understandable clinical research databases; (ii)Electronic medical records (EMRs) and other clinical systems that use a standardized clinical vocabulary to ensure that systems are able to communicate with one another; (iii)Standardized interfaces among clinical and practice management systems that facilitate communication among multiple systems; (iv)New and higher-performance technologies (e.g., speech recognition and wireless computers) to make it easier for physicians, clinicians, and administrators to enter data and enable better workflow compatibility.
    R.A. Miller. Medical diagnostic decision support systems--past, present, and future: a threaded bibliography and brief commentary. J Am Med Inform Assoc 1994 Mar-Apr;1(2):160

    [PubMed]   [PubMedCentral

    " Articles about medical diagnostic decision support (MDDS) systems often begin with a disclaimer such as, "despite many years of research and millions of dollars of expenditures on medical diagnostic systems, none is in widespread use at the present time." While this statement remains true in the sense that no single diagnostic system is in widespread use, it is misleading with regard to the state of the art of these systems. Diagnostic systems, many simple and some complex, are now ubiquitous, and research on MDDS systems is growing. The nature of MDDS systems has diversified over time. The prospects for adoption of large-scale diagnostic systems are better now than ever before, due to enthusiasm for implementation of the electronic medical record in academic, commercial, and primary care settings. Diagnostic decision support systems have become an established component of medical technology. This paper provides a review and a threaded bibliography for some of the important work on MDDS systems over the years from 1954 to 1993. "
    D. C. Classen. Editorial: Clinical Decision Support Systems to Improve Clinical Practice and Quality of Care. J Amer Med Assoc. Vol. 280 No. 15, October 21, 1998.

    []

    " Glowing predictions about the all-encompassing and beneficial role of computers in medicine have appeared with increasing frequency in the scientific literature since the 1970s. However, this optimistic vision has not yet been realized almost 25 years later. A prescient commentary 15 years ago predicted the numerous obstacles that have prevented these rosy scenarios from coming true in clinical practice.1 Several factors continue to echo the challenges faced in this area, including lack of investment; lack of leadership from practicing physicians, medical schools, and professional societies; and continuing control of information services in most health care organizations by chief information officers and other administrators ... "

    Burke JP, Classen DC, Pestotnik SL, Evans RS, Stevens LE. The HELP system and its application to infection control. J Hosp Infect. 1991 Jun;18 Suppl A:424-31.

    [PubMed]   []

    " The HELP system is a comprehensive hospital information system that is linked to an allied financial data base. The clinical data base integrates information from areas such as admitting, pharmacy, radiology, surgery, pathology, nursing, respiratory therapy, and the clinical laboratories, including microbiology. This allows for the creation of an electronic medical record that contains all the clinical and financial data for each patient. The HELP system combines both communication and advice features through the use of data- and time-driven algorithms. We have used the HELP system to automate the surveillance and analysis of hospital-acquired infections and to identify patients at high risk for nosocomial infection. The expert system features have also been used to suggest alternatives for patients receiving inappropriate antimicrobial therapy, to improve the timing of antibiotic prophylaxis in surgery, and to curtail unnecessarily prolonged prophylaxis. Automated hospital information systems such as HELP can facilitate the investigation of a broad range of infection control, quality improvement, and cost-containment issues. "
    E.S. Berner, M. J. Ball (Eds.). Clinical Decision Support Systems to Improve Clinical Practice and Quality of Care. Springer Verlag, 1998.

    []

    " "
    Shortliffe E. Medical expert systems knowledge tools for physicians. West J Med 1986; 145: 830-839.

    [SMI]

    " Recent advances in the field of artificial intelligence have led to the emergence of expert systems, computational tools designed to capture and make available the knowledge of experts in a field. Although much of the underlying technology available today is derived from basic research on biomedical advice systems during the 1970's, medical application packages are thus far generally unavailable from the young artificial intelligence industry. Medical expert systems will begin to appear, however, as researchers in medical artificial intelligence continue to make progress in key areas such as knowledge acquisition, model-based reasoning, and system integration for clinical environments. It is accordingly important for physicians to understand the current state of such research and the theoretical and logistical barriers that remain before useful systems can be made available. One experimental system, ONCOCIN, provides a glimpse of the kinds of knowledge-based tools that will someday be available to physicians. "
    Sim I, Gorman P, Greenes RA et al. Clinical Decision Support Systems for the Practice of Evidence-based Medicine. J Am Med Inform Assoc 2001 Nov-Dec;8(6):527-34.

    [PubMed]   [PubMed Central]

    " BACKGROUND: The use of clinical decision support systems to facilitate the practice of evidence-based medicine promises to substantially improve health care quality. OBJECTIVE: To describe, on the basis of the proceedings of the Evidence and Decision Support track at the 2000 AMIA Spring Symposium, the research and policy challenges for capturing research and practice-based evidence in machine-interpretable repositories, and to present recommendations for accelerating the development and adoption of clinical decision support systems for evidence-based medicine. RESULTS: The recommendations fall into five broad areas--capture literature-based and practice-based evidence in machine--interpretable knowledge bases; develop maintainable technical and methodological foundations for computer-based decision support; evaluate the clinical effects and costs of clinical decision support systems and the ways clinical decision support systems affect and are affected by professional and organizational practices; identify and disseminate best practices for work flow-sensitive implementations of clinical decision support systems; and establish public policies that provide incentives for implementing clinical decision support systems to improve health care quality. CONCLUSIONS: Although the promise of clinical decision support system-facilitated evidence-based medicine is strong, substantial work remains to be done to realize the potential benefits. "

    Musen MA. Stanford Medical Informatics: uncommon research, common goals. MD Comput. 1999 Jan-Feb;16(1):47-8, 50.

    [PubMed]   [SMI]

    A summary of the research undertaken at Stanford Medical Informatics (SMI) since the 1960s written by the current head.

    Teich JM, Wrinn MM. Clinical Decision Support Systems Come of Age. MD Comput 2000 Jan-Feb;17(1):43-6.

    [PubMed]

    " "

    AI Systems and clinical practice

    E.Coiera. Question the Assumptions. In: P. Barahona, J.P. Christensen (eds), Knowledge and Decisions in Health Telematics - The Next Decade, IOS Press,Amsterdam, (1994), 61-66

    [coiera.com]

    " The lack of demonstrable success of computer based decision support technologies in health care requires a re-examination of the assumptions that support initiatives into these technologies. There are many ways in which one could explain away this lack of success, for example lack of technological maturity, lack of appropriate computer record infrastructures, or indeed professional resistance to novel technologies. However, one can also look at the fundamental assumptions that are taken as given when designing decision support systems. These assumptions about the problems that decision support systems should assist with, and the way that assistance should be provided are open to question. Much of the current work in decision support system research and development is based upon outdated views of the clinical workplace, and it is likely that the failure of the technology is in large part due to this. What is required is a principled re-examination of clinical practice, aimed at identifying the ways in which clinical workers should really be assisted. Until this is done, decision support technology will continue to be mismatched to the needs of health care workers. "

    References: individual systems

    de Dombal FT, Leaper DJ, Staniland JR, McCann AP, Horrocks JC. Computer-aided diagnosis of acute abdominal pain. Br Med J. 1972 Apr 1;2(5804):9-13.

    [PubMed]

    " This paper reports a controlled prospective unselected real-time comparison of human and computer-aided diagnosis in a series of 304 patients suffering from abdominal pain of acute onset. The computing system's overall diagnostic accuracy (91,8%) was significantly higher than that of the most senior member of the clinical team to see each case (79,6%). It is suggested as a result of these studies that the provision of such a system to aid the clinician is both feasible in a real-time clinical setting, and likely to be of practical value, albeit in a small percentage of cases. "
    E. H. Shortliffe. Computer-Based Medical Consultations: MYCIN. Elsevier/North Holland, New York NY, 1976.

    []

    " "
    E. H. Shortliffe. Artificial Intelligence in Management Decisions: ONCOCIN. Reprinted in: Frontiers of Medical Information Sciences (R.L. Kuhn, ed.), pp. 173-185. New York: Praeger Publishers, 1988. KSL-86-39.

    []

    " "
    Miller PL. Building an expert critiquing system: ESSENTIAL-ATTENDING. Meth Inf Med 1986: 25: 71-78.
    [Republished in Methods of Information in Medicine, Special Issue 1988: 297-304.]

    []

    " "
    Miller RA, Pople HE Jr, Myers JD. Internist-I, an experimental computer-based diagnostic consultant for general internal medicine. N Engl J Med. 1982 Aug 19;307(8):468-76.

    [PubMed]

    " Internist-I is an experimental computer program capable of making multiple and complete diagnoses in internal medicine. It differs from most other programs for computer-assisted diagnosis in the generality of its approach and the size and diversity of its knowledge base. To document the strengths and weaknesses of the program we performed a systematic evaluation of the capabilities of INTERNIST-I. Its performance on a series of 19 clinicopathological exercises (Case Records of the Massachusetts General Hospital) published in the Journal appeared qualitatively similar to that of the hospital clinicians but inferior to that of the case discussants. The evaluation demonstrated that the present form of the program is not sufficiently reliable for clinical applications. Specific deficiencies that must be overcome include the program's inability to reason anatomically or temporally, its inability to construct differential diagnoses spanning multiple areas, its occasional attribution of findings to improper causes, and its inability to explain its "thinking". "

    acknowledgements
     
    page history
    Entry on OpenClinical: 2001
    Last main update: 14 February 2005, 24 July 2006

    Search this site
     

    Privacy policy User agreement Copyright Feedback

    Last modified:
    © Copyright OpenClinical 2002-2011