Clinical Decision Support Systems: Key to Making EHR an Improvement

Zusman, Edie E.; Vinokur, Yuriy

doi: 10.1227/01.neu.0000419707.79663.e7
Science Times

Many hospitals and neurosurgery practices are quickly working to convert from paper charts with years of successfully established patient workflows to electronic health records (EHR) to take advantage of financial incentives under the federal government's “meaningful use” regulations.1 Paperless patient records have been touted as an improvement in health care efficiency, safety, accuracy, consistency and accountability. But while EHRs can improve accessibility and legibility of information, healthcare experts agree that no significant improvement in the quality and cost of health care is realized from this expensive conversion alone. Benefits for patient care quality measures and cost savings can only result when EHR is used with rigorous, well designed Clinical Decision Support Systems (CDSS) implementation2 and accompanying workflow changes.

Computerized systems that assist with diagnostic interpretation, treatment planning and therapy recommendations,3 Clinical Decision Support Systems, combine clinical knowledge with patient-specific information. CDSS are critical to helping the movement in health information technology reach its potential. The article, “Clinical Decision Support Systems: Potential with Pitfalls,” published in the Journal of Surgical Oncology, describes both the shortcomings and the potential benefits of the systems integrated with the EHR that are designed to improve quality of care and patient outcomes. The article illustrates that for CDSS to be effective, it is essential that the planning and development of systems employ multidisciplinary teams of people who understand hospital workflow and operations.

The AHRQ report offers this simple scenario to illustrate the importance of CDSS to improve the value of EHR: While his doctor is out-of-town, an elderly asthma patient who has developed severe knee pain sees another physician in his doctor's office. An EHR provided documentation of the last visit, including recent laboratory results and a list of the patient's medications. This information easily brought the doctor up to date on the patient's condition. The doctor entered an order for medicine for the knee pain into the system, printed out a legible prescription for the patient, and sent him on his way. Unfortunately, within 2 months, the patient wound up in the emergency room with a bleeding ulcer caused by interaction of the pain medicine with the patient's asthma medicine.

With neuroscience patients requiring specialized care outside what is typically rendered by the majority of physicians, these patients are not clearly benefitted from existing EHR and CDSS capabilities designed to facilitate primary care physician's management of common patient problems. It is therefore important that neurosurgeons and our neurology peers participate in developing EHR order sets for patients and consider CDSS process improvements. If poorly executed, CDSS can make using EHR a time consuming process which fails to yield the desired outcomes, or potentially drives worse outcomes for our patients.

Although many clinicians are reluctant to embrace or remain unaware of the implications of CDSS, policymakers have made them a priority.3 The Institute of Medicine, for example, has made it clear in its position statements that “patients should receive care based on the best available scientific knowledge. Care should not vary illogically from clinician or from place to place.”4 In addition, the AHRQ has funded and launched programs to study CDSS, and Centers for Medicare and Medicaid Services (CMS) has explored initiatives to link reimbursement to “meaningful use” of health information technology.1,3

In the current healthcare environment, physicians, including specialists like neurosurgeons, will need to learn CDSS to optimize use of the EHR. Converting to EHR in a meaningful way to incorporate best neurosurgery practices will require awareness of both the potential benefits and drawbacks of the technology.

For neurosurgeons, the first steps toward inpatient neurosurgery/neuroscience EHR conversion with CDSS are collection of all neuroscience order sets from the organization with benchmark examples from the EHR vendor and/or other institutions. Neurosurgery colleagues can share their order sets and templates, and then form a collaborative group of the entire patient care team that is charged with creating an inclusive list of optional orders for neuroscience patients in the EHR format. The list can then be revised by the team as evidence-based best practices and other information become available. Existing workflows must then be evaluated, keeping in mind how they will change with EHR. Workflows should be optimized and improved as needed, and CDSS prompts and warnings should be incorporated to increase performance and prevent errors. Once the new plan is in place for a trial period, feedback and evaluation should be performed with a set schedule for review and revision. Conversion from paper charts to EHR should be implemented with planned tests of change as any business would roll out a major restructuring.

The term “clinical decision support systems” has been defined in many ways and conjures many meanings. It may suggest systems that provide references or referrals to clinical trials or the computerized version of the Physician's Desk Reference and its warnings of drug interactions. According to AHRQ “common features of CDS systems that are designed to provide patient-specific guidance include: the knowledge base (compiled clinical information on diagnoses, drug interactions and guidelines), a program for combining that knowledge with patient-specific information, and communication mechanisms (a way of entering patient data or importing it from the EHR into the CDS application), and providing relevant information (lists of possible diagnoses, preventive care reminders) back to the clinician.”3

If a physician enters an EHR order for morphine for post-operative pain, and the patient is allergic to morphine, for example, EHR alone does not prevent the order from being entered or executed—it is the addition of CDSS which prevents the EHR from accepting the order for the drug and notifies the clinician that the patient is allergic to morphine. The system would then ask if the physician wants to prescribe the drug anyway or suggest alternatives that may work for the patient.

CDSS alerts also can be used to remind nurses and other providers to take actions to prevent complications in post-surgical care. For example, the Centers for Disease Control and Prevention (CDC) recommends removal of foley catheters as soon as possible postoperatively, preferably within 24 hours, unless there are appropriate indications for continued use.5 Prompt removal of a catheter has been associated with a reduction in urinary tract infections, a common post-surgery complication. The CDSS can alert the nurse to the need to remove the catheter, and prompt the nurse to override the alert or provide justification for why the catheter is not being removed as recommended by best-practice guidelines. Well designed CDSS can also prompt shared expertise across specialties, such as creating a consultation with urology to support weaning a patient from the urinary catheter in more complex cases which meet predetermined criteria.

As any physician whose EHR is set with CDSS alerts and prompts knows, when they are poorly developed for clinical use they can be a hindrance to efficient care. An example from a Northern California health system is a CDSS prompt which asks whether a creatinine test has been ordered at the time of EHR order entry for a non-contrast MRI, even if there is no plan or order entered for using a contrast agent. None the less, the physician has to click several EHR boxes to circumvent the hard-stop on the MRI order which prevents the radiology department from scheduling the requested study unless a creatinine blood test is ordered, or until the physicians enters the rationale for why it is not ordered. Erroneous or poorly developed CDS systems, then, consume clinical time and add unnecessary frustration to the process.

Physicians in general and neurosurgeons specifically need to be involved in the time consuming but necessary EHR and CDSS development process to prevent wasted time and unnecessary prompts which physicians will otherwise encounter for each patient they see after EHR implementation. Working with the EHR development team to fix problems will take significant effort unless neuroscience representatives are already part of the build team, and a methodology for feedback and process improvement has been well established from the initial technology roll out.

CDSS is taking place in an era of significant expansion of medical knowledge, with statistics suggesting doubling of available health related data every 3 years.6 Combined with the increasing expectations for personalized health potentially based on individual genetics, John Eberhardt and his coauthors in their Journal of Surgical Oncology article, acknowledge the increase in the mental workload for clinicians: “As the amount of diagnostic, biomarker, and pathological data on a given individual patient case continues to expand, the result is an overwhelming amount of information that is almost impossible to keep pace with.”3 CDSS are envisioned as a way to help “clinicians to develop personalized therapy plans based on statistics and rule sets that are oriented toward patient- rather than population-specific estimates of risks and outcome.3 CDSS then allow the clinician to maintain awareness of an individual patient's history, treatment plans, and someday genetics, while shifting much of the information redundancy to a computer.

But while these kinds of systems are heralded as a technological means of improving care and patient outcomes, Eberhardt et al cite varied degrees of success and failure. One analysis of 97 CDSS studies, for example, showed that in 62 of the studies (64%), practitioner performance improved, particularly with the utilization of reminder systems. But the paper also reports only 7 out of 52 (13%) of the studies reported improvements in outcomes.7 Other studies have reported degradation in outcomes with use of CDSS. One study by Tsai et al8 demonstrated a system designed to help interpret electrocardiograms improved diagnostic accuracy, but when the system provided incorrect results, clinicians were more likely to agree with the incorrect results. The study raised concerns about clinician “overdependence” on CDS algorithms. Another study, by Han et al,9 observed mortality of pediatric critical care patients increased from 2.80% to 6.75% after the implementation of a computerized physician order entry system was implemented.

The discrepancies in outcomes indicate problems in implementation, the authors suggest, specifically: a lack of focus on a specific clinical problem; selection of an inappropriate system or method; poor consideration of and integration with clinical workflow; and, inadequate testing and user training.3

Indeed, moving from a paper record to an electronic record is not as simple as switching from pen and paper to computer keyboard. Conversion from paper to the EHR first requires an examination of processes for the continuum of care from when an imaging test is ordered for a stroke patient to the selection, timing and dosing of tPA using accepted best practices and evidence-based medicine. The electronic health record combined with clinical decision support systems should enable every patient to get the best care the organization can provide in the safest and most efficient manner.

As the authors in the Journal of Surgical Oncology write, the value of the CDSS is only as good as its design.3 “If efforts have not been focused on an appropriate clinical problem, with the appropriate approach selected specific to that problem, and a corresponding solution with workflow integration developed, one may find, despite one's best efforts, that the CDSS only has a limited impact on clinical practice and patient outcomes.”3

The authors describe developing systems that balance the need for clinician autonomy with the responsibility for sorting and interpreting knowledge using predictive algorithms to provide rapid quantitative interpretation, which ensures speed and immediate applicability.3 They further argue that use of algorithms must be validated, both for accuracy and utility in the clinical setting.3

Ensuring the most effective CDSS, therefore, requires meetings at the outset with multidisciplinary teams of care providers that include all of the stakeholder staff, clinicians and physicians. These efforts should be facilitated both by the IT technicians who can build the EHR order sets with CDSS capabilities, but also health system managers who know how to optimize hospital operations and workflow.

Back to Top | Article Outline


1. Zusman EE. Meeting meaningful use objectives for Electronic Health Record implementation. Neurosurgery. 2011;69(2):N24–N26.
2. Berner ES. Clinical Decision Support Systems: State of the Art. AHRQ Publication No. 09-0069-EF. Rockville, MD: Agency for Healthcare Research and Quality. June 2009. Available at:
3. Eberhardt J, Bilchik A, Stojadinovic A. Clinical decision support systems: potential with pitfalls. J Surg Oncol. 2012;105(5):502–510.
4. Osheroff JA. Improving Medication Use and Outcomes with Clinical Decision Support: a Step-by-Step Guide. Chicago, IL: The Healthcare Information and Management Systems Society; 2009.
5. Gould CV, Umscheid CA, Agarwal RK, Kuntz G, Pegues DA the Healthcare Infection Control Practices Advisory Committee (HICPAC). Guideline for prevention of catheter-associated urinary tract infections 2009. Available at
6. Stroetmann B, Aisenbrey A. Medical knowledge management in healthcare. World Acad Sci Eng Technol. 2012;64:558–562.
7. Garg AX, Adhikari NK, McDonald H, et al.. Effects of computerized clinical decision support systems on practitioner performance and patient outcomes: a systematic review. JAMA. 2005;293(10):1223–1238.
8. Tsai TL, Fridsma DB, Gatti G. Computer decision support as a source of interpretation error: the case of electrocardiograms. J Am Med Inform Assoc. 2003;10(5):478–483.
9. Han YY, Carcillo JA, Venkataraman ST, et al.. Unexpected increased mortality after implementation of a commercially sold computerized physician order entry system. Pediatrics. 2005;116(6):1506–1512.
Copyright © by the Congress of Neurological Surgeons