Difference between revisions of "CDS"

Revision as of 22:13, 4 September 2014

Clinical decision support (CDS) refers broadly to providing clinicians or patients with clinical knowledge, intelligently filtered and presented at appropriate times. Clinical knowledge of interest could range from simple facts and relationships (such as an individual patient's vital signs, allergies and lab data) to relevant medical knowledge (such as best practices for managing patients with specific disease states, new clinical research, professional organizations' practice guidelines, expert opinion, systematic reviews, and other types of information.

History

Clinical decision support tools existed prior to development of electronic medical records (EMRs). They include expert consultation, practice guidelines carried in clinicians' pockets, patient cards used by nurses to track a patient's treatments, tables of important medical knowledge, and ICU patient flow sheets. Many of these CDS tools are still relevant, but integration of CDS with current EMRs presents an opportunity for the various types of decision support to be immediately available at the time of the decision-making. CDS can be more relevant, more accurate, and can facilitate and be integrated with clinical workflow.

For more on the history of CDS, see here and here.

CDS components

There are several key components of a good clinical decision support system.

• Documentation tools
• Clinician Checklists
• Calculators
• Reference Links

Documentation forms/templates

As mentioned above, these existed prior to EMRs in the form of structured documentation forms for conducting clinician assessments. Many of these have been supplanted by digital reproductions in EMR of the original paper documentation form.

Examples include:

• nursing intake forms
• physician "History & Physicals"
• ER templates

Other tools that were artifacts of clinician workflow and existed prior to EMR implementation, now have the potential for added functionality when computerized, web-based, or automated. Added functionality includes dispersed access to the tool's information (ability for multiple users from multiple disciplines and geographic locations to share a single set of information), auto-population of accurate and current data from the clinical information system, linkages between tool task lists and CPOE, and improved order fulfillment efficiency.

Examples of these tools include:

• Handoff tools (lists of patients with summations of clinical data used at time of a shift handoff between clinicians)
• Rounding tools (summaries of data on a single patient, clinical task lists
• ICU flowsheets for documenting and charting vital signs and hemodynamic data.

Alerts and reminders

Examples include:

• Alert that appropriate cancer screening is due.
• Drug allergy alert
• Drug interaction alert
• Underdose/overdose alerts based on renal or liver function, age, drug level

Relevant data presentation

Examples of this include:

a) Patient specific data such as:

• Display of relevant labs during medication CPOE such as patient's renal and liver function.
• Display of other relevant patient data during CPOE such as patient's age (which may affect side affects and dosing) or conditions.

b) Population-specific data such as:

• Retrospective filtering and aggregate reporting: disease registries and clinic population dashboards.
• Microbiograms: tables of local bacterial flora and their sensitivity and susceptibility to various antibiotics

Order creation facilitators

Examples include: order sets, order menus, tools for complex ordering, and "single-order completers including consequent order."

Order Sets

An order set is a group of related orders which a physician can place with a few keystrokes or mouse clicks. An order set allows users to issue prepackaged groups of orders that apply to a specified diagnosis or a particular period of time. Using order sets reduces both time spent entering orders and terminal usage. An order set may or may not contain medication orders as part of the set.

An example order set for Cardiac MRI order would include:

• MRI order
• Prescription to dispense IV contrast
• Prescription for sedative during MRI
• Order for renal function lab if none in EMR in last week
• Order for transportation from hospital ward to radiology at time of MRI

Order Menus

An order menu is a group of related orders which are depicted onscreen together via an EMR's GUI so that an ordering clinician visualizes the breath and organization of the orders. An order menu allows CPOE/EMR developers to direct clinicians towards the most common or appropriate orders for a particular topic. Using order sets reduces time spent searching for the desired orders and provides a rudimentary level of knowledge and education. Order sets are commonly made up of medication orders, but non-medication orders may be included.

Examples of order menu content include:

• anti-hypertensive medications arranged by class, by preference, by cost, or other means.
• common pulmonary medications to treat COPD, asthma, embolisms, and chronic cough.

Single-order completers including consequent order

These may be broken down in to Medication Safety Rules and Non-medication Safety Rules.

Medication safety rules and decision support

• Adverse drug reactions
• Basic Dosing Guidance for medications in CPOE
• Formulary decision support
• Duplicate Therapy Checking
• Advanced Dosing Guidance in CPOE
• Patient Characteristic dosing support
• Medications to be avoided in the elderly
• Medications requiring dosage adjustments in renal insufficiency
• Medications requiring dosage adjustments in hepatic disease
• Medications to be avoided during pregnancy
• Medications to be avoided while breastfeeding
• Vaccination contraindications
• Common Corollary orders
• Detection of Adverse Mediation-Related Events

Non-medication safety rules

• Diagnosis-Order RulesDrug and level. Postop order sets, disease specific order sets. Suggested dose. Suggested alternate medication for shortage or formulary. Guided dose algorithims for complex orders sucha s those required with insulin and heparin infusions in which nurses are given parameters with which to adjust dose on a regular basis.

Interaction models

An interaction model is a set of rules for making clinical decisions. The rules are based on a large collection of medical knowledge and an accurate computer representation scheme.

Artificial intelligence

Artificial intelligence is a system that was developed by a team of system engineers and clinicians. The system would take some of the workload from medical teams by assisting the physicians with tasks like diagnosis & Therapy recommendations.

Business Intelligence and Data Warehousing

• Business intelligence
• Data warehouse

Validation and Verification of Clinical Decision Support

• On Validation and Verification Of Decision Support Protocol Subsystems During Implementation-Optimization: Encapsulating P(X)

Sample Decision Support Content

• Diabetes CDS Content
• Drug-Drug Interaction Rules
• Clinical Reminders from Beth Israel/Deaconess Medical Center in Boston
• Symptom Triage Decision Support for Consumers (example: "Chest Pain") [1]
• Weight-based Heparin Dosing Guidelines
• Flowchart-based decision support sample content
• Preventive care reminders
• Mental health clinical decision support

CDS Implementation

CDS should be designed to provide the right information to the right person in the right format through the right channel at the right time.

At the stage of planning for implementation of any new health IT system or their components, there are some considerations and steps that should be followed to maximize CDS system success:

1. Needs Assessment: ensuring that identified clinical needs and functional requirements
2. Assessing Organizational Readiness

        i)   Understanding prior physician and organizational experience with CDS
       ii)   Assessment of level of physician knowledge, perception, engagement, and willingness to change
      iii)   Aligned leadership with clear objectives

1. CDS related factors

        i)    Deciding whether to purchasing a commercial system or build the system
       ii)    CDS usability: Will CDS increase physician workload? Can the level of intrusiveness of alerts be customized?
       iii)   Adequate planning for encouraging physicians to use CDS
       iv)    Appropriate training on using CDS
        v)    Mechanisms in  place to evaluate usage and effectiveness of the CDS

Clinical Decision Support overview

• National Roadmap for Clinical Decision Support
• General system features associated with improvements in clinical practice
• Support Decisions with Diagnostic Aids
• Clinical Decision Support Liability

CDS success measures

To estimate the success of the system we should look at the following points[3]:

1. System quality.
2. Information quality
3. Usage
4. User satisfaction (Process Outcome)
5. Individual impact (Clinical Outcome)
6. Organizational impact (Financial outcome).

CDS considerations

• Liability of physicians, hospitals, and EHRs
• Workflow integration
• Improved Usability
• Evidence based content / Clinical content accuracy
• Changing behavior (limited interaction by users, adherence to protocol)
• Training and communication
• System design limitations
• Alert fatigue
• Choosing the right metrics for reporting (Process / Clinical)
• Potential breaks due to system upgrades

Information Resources

• The HIMSS Clinical Decision Support (CDS) Task Force wiki
• Alert placement in clinical workflow
• Initial Selection of What to Alert on...
• Alerts versus on-demand CDS
• Sources of clinical decision support content
• Here is a video of CDS in action within the free EHR drchrono [2].

CDS benefits

Results indicate the potential of CDS to improve the quality of care. These are good reasons for institutions to adopt CDS, but they should do so at their own pace and volition.

Promote use of evidence based recommendations

A stand-alone, disease-specific CDSS can improve concordance with established prescribing guidelines for a period measured in months.

Better clinical decision-making

• Reduced medication errors
• Improved cost-effectiveness
• Increased patient convenience
• Improved quality of healthcare delivery
• Improved healthcare outcomes for patients and patient populations.

References

1. Slater, B. Osheroff, JA. Clinical Decision Support, in Electronic Health Records: A Guide for Clinicians and Administrators. American College of Physicians. 2008.
2. Franklin, MJ, et al, Modifiable Templates Facilitate Customization of Physician Order Entry, [3]
3. Sittig, DF, and Stead, WW, Computer-based Order Entry: The State of the Art, J Am Med Informatics Assoc., 1994;1:108-123. [4]
4. Anderson, JG, et al, Physician Utilization of a hospital information system: a computer simulation model. Pric Annu Symp Compu Appl Med Care, IEEE, 1988;12:858-861. [5]
5. Southern Ohio Medical Center, [6]
6. Clinical Decision Support Systems :State of the Art AHRQ Publication No.09* 0069* EF June 2009
7. Grand challenges in Clinical Decision Support Journal of Biomedical Informatics 41(2008) 387* 392
8. Determinants of Success of Inpatient Clinical Information Systems: A Literature Review. M J van der Meijden, H J Tange, J Troost, et al. JAMIA 2003 10: 235* 243
9. Improving Outcomes with Clinical Decision Support: An Implementer's Guide [Paperback]: Jerry Osheroff, Jonathan Teich, Donald Levick, Luis Saldana, Ferdinand Velasco, Dean Sittig, Kendall Rogers and Robert Jenders

Updated by (Edward A W Dyer)