Clinical decision support system

From Citizendium, the Citizens' Compendium
Jump to: navigation, search
This article is developing and not approved.
Main Article
Talk
Related Articles  [?]
Bibliography  [?]
External Links  [?]
Citable Version  [?]
 
This editable Main Article is under development and not meant to be cited; by editing it you can help to improve it towards a future approved, citable version. These unapproved articles are subject to a disclaimer.

Clinical (or Diagnostic) Decision Support Systems (CDSS) are interactive computer programs that directly assist physicians and other health professionals with decision making tasks. These fall under the class of Decision support systems within medical informatics. CDSS offer a powerful medical informatics tool that can promote evidence based medicine.

Why do we need such systems at all? If diagnosis becomes the job of the computer program, what will the doctors do? Such questions overlook the key term support - i.e., they are intended to support the clinician, rather than replace them. Computers are not error or fatigue prone like human beings; however, computers may cause harm through other faults.[1][2][3][4]

In medical diagnosis, there is scope for ambiguity in the inputs, such as the history (patient’s description of the diseased condition), physical examinations (especially in cases of uncooperative or less intelligent patients), and laboratory tests (faulty methods or equipment). Moreover, in treatment, there are the chance of drug reactions and specific allergies, and also of patients' non-compliance of the therapy due to cost or time or adverse reactions. All these factors may not be taken into account by a busy clinician attending hundreds of patients daily.

In all these areas, computers can help the clinician to reach an accurate diagnosis faster. Another new branch of medicine pharmacogenomics is the product of breeding between information technology and biology, leading to individualized treatment.

The basic components of a CDSS include a dynamic (medical) knowledge base and an inference mechanism (usually a set of rules derived from the experts and from the data provided by evidence-based medicine). It may be based on Expert systems or artificial neural networks or both (Connectionist expert systems) - all of which can be loosely termed as Artificial intelligence - AI - techniques.

CDSS may be linked to Electronic health records, for decision making and can also be used for assisting the practice of Evidence-based medicine (EBM) in a more structured way.

The role of even an apparently simple search by Google in making diagnoses has been quantified.[5]

Methods of the underlying decision making process

Various methods are used to created the automated decisions.[6]

Knowledge based systems / expert systems

Knowledge-based expert systems are created by having experts use the biomedical literature to identify relationships between independent variables (such as signs and symptoms) and dependent variables (such as likely underlying diseases). Sometimes, local hospital information, such as rates of surgical complications, may be incorporate.[7] These relationships become predefined rules in the form of {if-else if-then} to guide the decision making. Rules may also be obtained by various forms of decision trees (e.g., Iterative Dichotomizer or ID variants) or Bayesian networks.

Rule-based CDSS are the ones that are mostly found in the commercially available clinical informatics applications. Alerts for allergies and possible drug interactions, prompts for drug doses corrected for weight, height, sex, age and underlying clinical condition are the ones that are most commonly touted as CDSS. Arden Syntax was designed and developed as a product-independent syntax for codifying these alerts. Examples of rule-based systems are:

  1. Logical or deductive: branching logic (e.g., ID3 or iterative dichotomizer 3); e.g., MYCIN, NEOMYCIN[8]
  2. Probabilistic: Bayesian systems weight probabilities rather than use binary options as done be logical systems.[9] An example is DXplain.[10]
  3. Hybrid: heuristic reasoning e.g., QMR (Quick Medical Reference), ILIAD[11], ISABEL[12][13]. Hueristic methods may perform better for uncommon diseases.[14]

The critical difficulty in constructing rule-based expert systems for medical decision support has been in recruiting experts with domain knowledge to create the knowledge base and train the system. Training expert systems is time-consuming, and has only produced usable results in narrowly-scoped projects. However, due to issues related to its acceptance (a doctor may need to 20-40 minutes to enter a case[15]) and standardization, rule-based CDSS has not delivered its immense promise.

Evidence-adaptive systems

Evidence-adaptive systems are proposed that are based on current evidence and have a mechanism for routine updating of recommendations with new research findings.[7]

Machine learning

In these systems, the relationships between independent variables (such as signs and symptoms) and dependent variables (such as likely underlying diseases) is created by having the system be trained on a "large collection of previously classified examples during a period of supervised learning"[16][17]. Machine learning may also be called case-based due to the system being trained on a collection of previously classified cases.[18] A classic example is automated electrocardiogram interpretation.[19]

Artificial Neural Networks

Artificial Neural Networks ANNs (also referred to as connectionist architectures, parallel distributed processing, and neuromorphic systems), are information-processing paradigm inspired by the way the densely interconnected, massively parallel structure of the mammalian brain processes information. ANNs are collections of mathematical models that emulate some of the observed properties of biological nervous systems and draw on the analogies of adaptive biological learning. ANN can perform supervised or unsupervised machine learning, depending on the categorical information maps available at the time of rule inferencing.

Bayesian Belief Network

A Bayesian Belief Network (BBN) is a mathematical model for using AI (probabilistic) networks for predictive purposes.[17] It is used in the CDSS module as an AI engine to help make probabilistic predictions based upon the observations made by the clinician. These could be used for diagnostics, treatment planning, and even predicting outcomes. An example is Pathfinder for surgical pathology.[20][21]

Connectionist expert systems

Connectionist expert system is a type of Artificial Neural Networks (ANNs) in which humans can help the system revise weights. Although humans help revise the system based on the empiric data in the system, literature based knowledge is not directly used to modify weight. Connectionist expert system are the "inferencing" methods of the ANNs can be backtracked and "rules generation" is possible. This might actually lead to the enhancement and enrichment of the medical knowledge base itself. This system architecture corresponds to the concept of "unsupervised" machine learning, where the algorithm looks for patterns in data without being instructed about the actual categories of information. This is also known as data mining. An example of a connectionist system is HYCONES.[22]

Hybrid systems

An example of a hybrid system has been developed for diagnosing congenital heart disease.[18] This example combines case-based reasoning with a neural network.

Methods of giving the advice

Various methods exist for displaying the advice for the clinicians.[23] The systems must be very user friendly.[24][1][25] The advice may be actively displayed or may be passive in which case the clinician must request the advice to be displayed. Active display was more effective in a cluster randomized controlled trial.[26]

If a system offers advice, either actively or passively, it should be possible to request display of the rule chains or other path the system used in reaching its recommendation. Such displays do not necessarily help in the direct clinical process, but are quite valuable in validating the knowledge and inference base. Ideally, these will be examined by people who are comfortable both in the medical and information system domains.

When presenting the advice in the form of narrative rather than graphics or recommended decision, for the system to be "physician-friendly", the designer must be aware that direct communication among clinicians, as well as progress and consultant notes, exist in an extremely high context of domain knowledge. One of the better demonstrations of the need for context is to listen to medical students and junior residents to present cases, with a great deal of true but irrelevant information.

Part of the art of medical diagnosis is to know what is important in a particular context. Computer scientists new to the field would do well to read some of the better autobiographies of physicians, especially in training. To a programmer not sensitive to the context, they may assume that a review of systems must always be presented in the same order. An object lesson, in this case from a medical student, was recounted by William Nolen, when he was a senior resident in obstetrics. [27]

He had asked one of his students to assess a woman in labor. A few minutes later, he came into the delivery suite, to find it dark, save for a tiny light. Mystified, he turned on the room lights, to find his student with flashlight in hand. Nolen inquired what the student was doing, and was told that he was "assessing" as he had been taught, starting with the scalp and working downward. When he was interrupted, he was testing pupillary response to light. His mentor, more or less gently, suggested that there were more important things to assess in active labor, such as the fetal position and the mother's cervical dilation.

If a informaticist does not understand why Nolen was frustrated, that person is not ready to design clinically useful systems.

Medical order entry system

For more information, see: Medical order entry system.


Effectiveness

Systematic reviews of studies conclude that reminders help.[28][29][30][31][32]

A more recent trial found that a clinical decision support system could improve antimicrobial prescribing.[33]

Point of care computer reminders may improve care according to a systematic review by the Cochrane Collaboration.[34]

In internal medicine, a study of four diagnostic support systems diagnosed a collection of tests cases with accuracy ranging from 0.52 to 0.71.[35] Another study of ISABEL found it included the correct diagnosis in its list of 30 proposed diagnoses for a collection of tests cases in 96% of the cases.[13] A study of QMR found that it could correctly list the final diagnosis among the first five diagnoses it suggested for 36% to 40% of 1144 consecutive inpatients.[36]

Iliad has helped detect diagnostic errors missed by peer review for health care quality assurance.[37][38]

In pediatrics, a systematic review by the Cochrane Collaboration concluded "there are very limited data from randomised trials on which to assess the effects of clinical decision support systems in neonatal care."[39]

Open source initiatives

OpenCLIPS is an open source release of the CLIPS expert system shell.[40] It has been based on the inference engine of EGADSS.[41]. Their goal is to build an active community focused around the support and development of [42].

One of the earlier evaluations on the efficacy of Internet-based CDSS [43] has been favorable.

A CDSS using open source software and delivered through wireless hand-held device has been found to be effective in stopping smoking in primary care [44].

PubMed Central [45] has approximately 200 full-text articles on CDSS built with the help of open source software.

Market surveys [46] have predicted that development of more robust CDSS is likely to increase its adaptation.

Problems

Sim et all have identified five challenges to CDDS:[7]

  1. "Capture of both literature-based and practice-based research evidence into machine-interpretable formats suitable for CDSS use"
  2. "Establishment of a technical and methodological foundation for applying research evidence to individual patients at the point of care"
  3. "Evaluation of the clinical effects and costs of CDSSs, as well as how CDSSs affect and are affected by professional and organizational practices"
  4. "Promotion of the effective implementation and use of CDSSs that have been shown to improve clinical performance or outcomes"
  5. "Establishment of public policies that provide incentives for implementing CDSSs to improve health care quality"

Alert fatigue

May alerts are ignored.[47]

Increased time required by providers

Physicians have asked an institution to removed reminders that improved care at the expense of physician time.[48]

Regarding expert diagnostic systems, a doctor may need to 20-40 minutes to enter a case[15]) and standardization, rule-based CDSS has not delivered its immense promise.

References

  1. 1.0 1.1 Koppel R, Metlay JP, Cohen A, et al (2005). "Role of computerized physician order entry systems in facilitating medication errors". JAMA 293 (10): 1197–203. DOI:10.1001/jama.293.10.1197. PMID 15755942. Research Blogging.
  2. Nielsen, Jakob (April 11, 2005). Medical Usability: How to Kill Patients Through Bad Design (Jakob Nielsen's Alertbox). Retrieved on 2007-10-23.
  3. Ash JS, Sittig DF, Poon EG, Guappone K, Campbell E, Dykstra RH (2007). "The extent and importance of unintended consequences related to computerized provider order entry". Journal of the American Medical Informatics Association : JAMIA 14 (4): 415–23. DOI:10.1197/jamia.M2373. PMID 17460127. Research Blogging.
  4. Han YY, Carcillo JA, Venkataraman ST, et al (2005). "Unexpected increased mortality after implementation of a commercially sold computerized physician order entry system". Pediatrics 116 (6): 1506–12. DOI:10.1542/peds.2005-1287. PMID 16322178. Research Blogging.
  5. Tang H, Ng JH (2006). "Googling for a diagnosis--use of Google as a diagnostic aid: internet based study". BMJ 333 (7579): 1143–5. DOI:10.1136/bmj.39003.640567.AE. PMID 17098763. Research Blogging.
  6. Fagan, Lawrence Marvin; Shortliffe, Edward Hance; Perreault, Leslie E.; Wiederhold, Gio (2001). “Clinical Decision-Support Systems”, Medical informatics: Computer Applications in Health Care and Biomedicine, 2nd. Berlin: Springer, 585. ISBN 0-387-98472-0. 
  7. 7.0 7.1 7.2 Sim I, Gorman P, Greenes RA, et al (2001). "Clinical decision support systems for the practice of evidence-based medicine". J Am Med Inform Assoc 8 (6): 527–34. PMID 11687560[e]
  8. Sotos JG (1990). "MYCIN and NEOMYCIN: two approaches to generating explanations in rule-based expert systems". Aviat Space Environ Med 61 (10): 950–4. PMID 2241738[e]
  9. de Dombal FT (1995). "Computer-aided diagnosis and medical decision support are not synonymous". Methods Inf Med 34 (4): 369–70. PMID 7476468[e]
  10. MGH Laboratory of Computer Science - projects - dxplain. Retrieved on 2008-01-22.
  11. Friedman CP, Elstein AS, Wolf FM, et al (1999). "Enhancement of clinicians' diagnostic reasoning by computer-based consultation: a multisite study of 2 systems". JAMA 282 (19): 1851–6. PMID 10573277[e]
  12. Isabel Healthcare. Retrieved on 2008-01-22.
  13. 13.0 13.1 Graber ML, Mathew A (2008). "Performance of a web-based clinical diagnosis support system for internists". J Gen Intern Med 23 Suppl 1: 37–40. DOI:10.1007/s11606-007-0271-8. PMID 18095042. Research Blogging.
  14. Puppe B, Ohmann C, Goos K, Puppe F, Mootz O (1995). "Evaluating four diagnostic methods with acute abdominal pain cases". Methods Inf Med 34 (4): 361–8. PMID 7476467[e]
  15. 15.0 15.1 Graber MA, VanScoy D (2003). "How well does decision support software perform in the emergency department?". Emerg Med J 20 (5): 426–8. PMID 12954680[e]
  16. Fagan, Lawrence Marvin; Shortliffe, Edward Hance; Perreault, Leslie E.; Wiederhold, Gio (2001). “Clinical Decision-Support Systems”, Medical informatics: Computer Applications in Health Care and Biomedicine, 2nd. Berlin: Springer, 586. ISBN 0-387-98472-0. 
  17. 17.0 17.1 Kononenko I (August 2001). "Machine learning for medical diagnosis: history, state of the art and perspective". Artif Intell Med 23 (1): 89–109. PMID 11470218[e]
  18. 18.0 18.1 Reategui EB, Campbell JA, Leao BF (1997). "Combining a neural network with case-based reasoning in a diagnostic system". Artif Intell Med 9 (1): 5–27. DOI:10.1016/S0933-3657(96)00359-4. PMID 9021057. Research Blogging.
  19. Willems JL, Abreu-Lima C, Arnaud P, et al (1991). "The diagnostic performance of computer programs for the interpretation of electrocardiograms". N. Engl. J. Med. 325 (25): 1767–73. PMID 1834940[e]
  20. Heckerman DE, Nathwani BN (1992). "Toward normative expert systems: Part II. Probability-based representations for efficient knowledge acquisition and inference". Methods Inf Med 31 (2): 106–16. PMID 1635462[e]
  21. Heckerman DE, Horvitz EJ, Nathwani BN (1992). "Toward normative expert systems: Part I. The Pathfinder project". Methods Inf Med 31 (2): 90–105. PMID 1635470[e]
  22. Leão Bde F, Guazzelli A, Mendonça EA (1994). "HYCONES II: a tool to build hybrid connectionist expert systems". Proc Annu Symp Comput Appl Med Care: 747–51. PMID 7950024[e]
  23. Fagan, Lawrence Marvin; Shortliffe, Edward Hance; Perreault, Leslie E.; Wiederhold, Gio (2001). “Clinical Decision-Support Systems”, Medical informatics: Computer Applications in Health Care and Biomedicine, 2nd. Berlin: Springer, 584. ISBN 0-387-98472-0. 
  24. Bates DW, Kuperman GJ, Wang S, et al. (2003). "Ten commandments for effective clinical decision support: making the practice of evidence-based medicine a reality". J Am Med Inform Assoc 10 (6): 523–30. DOI:10.1197/jamia.M1370. PMID 12925543. PMC 264429. Research Blogging.
  25. Nielsen, Jakob (April 11, 2005). Medical Usability: How to Kill Patients Through Bad Design.
  26. van Wyk JT, van Wijk MA, Sturkenboom MC, Mosseveld M, Moorman PW, van der Lei J (2008). "Electronic alerts versus on-demand decision support to improve dyslipidemia treatment: a cluster randomized controlled trial". Circulation 117 (3): 371–8. DOI:10.1161/CIRCULATIONAHA.107.697201. PMID 18172036. Research Blogging.
  27. Nolen, William A. (1972), The Making of a Surgeon, Pocket
  28. Bright TJ, Wong A, Dhurjati R, Bristow E, Bastian L, Coeytaux RR et al. (2012). "Effect of Clinical Decision-Support Systems: A Systematic Review.". Ann Intern Med. DOI:10.1059/0003-4819-157-1-201207030-00450. PMID 22529043. Research Blogging.
  29. Holt TA, Thorogood M, Griffiths F (2012). "Changing Clinical Practice Through Patient Specific Reminders Available at the Time of the Clinical Encounter: Systematic Review and Meta-Analysis.". J Gen Intern Med. DOI:10.1007/s11606-012-2025-5. PMID 22407585. Research Blogging.
  30. Roshanov PS, Misra S, Gerstein HC, Garg AX, Sebaldt RJ, Mackay JA et al. (2011). "Computerized clinical decision support systems for chronic disease management: A decision-maker-researcher partnership systematic review.". Implement Sci 6 (1): 92. DOI:10.1186/1748-5908-6-92. PMID 21824386. PMC PMC3170626. Research Blogging.
  31. Shojania KG, Jennings A, Mayhew A, Ramsay C, Eccles M, Grimshaw J (2010). "Effect of point-of-care computer reminders on physician behaviour: a systematic review.". CMAJ 182 (5): E216-25. DOI:10.1503/cmaj.090578. PMID 20212028. PMC PMC2842864. Research Blogging.
  32. Garg AX, Adhikari NK, McDonald H, et al (2005). "Effects of computerized clinical decision support systems on practitioner performance and patient outcomes: a systematic review". JAMA 293 (10): 1223–38. DOI:10.1001/jama.293.10.1223. PMID 15755945. Research Blogging.
  33. McGregor JC, Weekes E, Forrest GN, et al (2006). "Impact of a computerized clinical decision support system on reducing inappropriate antimicrobial use: a randomized controlled trial". J Am Med Inform Assoc 13 (4): 378–84. DOI:10.1197/jamia.M2049. PMID 16622162. Research Blogging.
  34. Shojania KG, Jennings A, Mayhew A, Ramsay CR, Eccles MP, Grimshaw J (2009). "The effects of on-screen, point of care computer reminders on processes and outcomes of care". Cochrane Database Syst Rev (3): CD001096. DOI:10.1002/14651858.CD001096.pub2. PMID 19588323. Research Blogging.
  35. Berner ES, Webster GD, Shugerman AA, et al (1994). "Performance of four computer-based diagnostic systems". N. Engl. J. Med. 330 (25): 1792–6. PMID 8190157[e]
  36. Lemaire JB, Schaefer JP, Martin LA, Faris P, Ainslie MD, Hull RD (1999). "Effectiveness of the Quick Medical Reference as a diagnostic tool". CMAJ 161 (6): 725–8. PMID 10513280[e]
  37. Lau LM, Warner HR, Poulsen AL (1995). "Research review: a computer-based diagnostic model for individual case review.". Top Health Inf Manage 15 (3): 67-79. PMID 10140306[e]
  38. Lau LM, Warner HR (1992). "Performance of a diagnostic system (Iliad) as a tool for quality assurance.". Comput Biomed Res 25 (4): 314-23. PMID 1511593[e]
  39. Tan K, Dear PR, Newell SJ (2005). "Clinical decision support systems for neonatal care". Cochrane Database Syst Rev (2): CD004211. DOI:10.1002/14651858.CD004211.pub2. PMID 15846701. Research Blogging.
  40. OpenCLIPS. Retrieved on 2008-01-24.
  41. EGADSS. Retrieved on 2008-01-28.
  42. CLIPS Web Site Moving - CLIPSESG. Retrieved on 2008-01-28.
  43. Thomas KW, Dayton CS, Peterson MW (1999). "Evaluation of internet-based clinical decision support systems". J. Med. Internet Res. 1 (2): E6. DOI:10.2196/jmir.1.2.e6. PMID 11720915. Research Blogging.
  44. Michel G, Marcy T, Shiffman RN (2005). "A wireless, handheld decision support system to promote smoking cessation in primary care". AMIA Annu Symp Proc: 530–4. PMID 16779096[e]
  45. open source clinical decision support system - PMC Results. Retrieved on 2008-01-28.
  46. Development of Robust Clinical Decision Support Systems Key to Increased Adoption Across Europe. Retrieved on 2008-01-28.
  47. Baysari MT, Westbrook JI, Richardson KL, Day RO (2011). "The influence of computerized decision support on prescribing during ward-rounds: are the decision-makers targeted?". J Am Med Inform Assoc. DOI:10.1136/amiajnl-2011-000135. PMID 21676939. Research Blogging.
  48. Drescher FS, Chandrika S, Weir ID, Weintraub JT, Berman L, Lee R et al. (2011). "Effectiveness and acceptability of a computerized decision support system using modified wells criteria for evaluation of suspected pulmonary embolism.". Ann Emerg Med 57 (6): 613-21. DOI:10.1016/j.annemergmed.2010.09.018. PMID 21050624. Research Blogging.