Introduction and Background
In 1950, it was expected to take 50 years to double the body of medical knowledge. Today, it is estimated that knowledge doubles every 73 days.1 As a result, medical decisions are becoming more complex for patients and families and for the doctors, nurses and healthcare professionals at the center of these critical, often life-or-death decisions.
The American Medical Informatics Association (AMIA) has a workgroup dedicated to the topic of Clinical Decision Support and describes it as “a process for enhancing health-related decisions and actions. It empowers clinicians, patients, and other stakeholders by enhancing clinical decision-making and clinical processes and improving the quality of health care services and patient outcomes.”2
The U. S. Food and Drug Administration (FDA), offer a definition of Clinical Decision Support. In the FDA Safety and Innovation Act (FDASIA) Health IT Report of 2014, Clinical Decision Support is described by the FDA as a variety of tools including, but not limited to: computerized alerts and reminders for providers and patients; clinical guidelines; condition-specific order sets; focused patient data reports and summaries; documentation templates; diagnostic support; and contextually relevant reference information.3
Fortunately, we are in the cognitive computing era and can leverage technology to support human decision-making. Clinical Decision Support Systems (CDSS) have been around for quite some time, and, as the name implies, CDSS are intended to support this decision-making, not supplant it. While CDSS offer many benefits, they also come with significant risks that must be managed and mitigated to ensure that technology does not introduce any unintended consequences.
Case uses of CDSS can be traced back to 1960 and have evolved dramatically since then.4 Today, the Office of the National Coordinator (ONC) for Health Information Technology defines Decision Support Systems as:
Computer tools or applications to assist physicians in clinical decisions by providing evidence-based knowledge in the context of patient-specific data. Examples include drug interaction alerts at the time medication is prescribed and reminders for specific guideline-based interventions during the care of patients with chronic disease. Information should be presented in a patient-centric view of individual care and also in a population or aggregate view to support population management and quality improvement.5
The National Health Service (NHS) England defines CDSS as “electronic tools that apply knowledge systems which use two or more items of patient data to generate case-specific advice, to improve clinical decision-making.”6
The lack of consensus around a single definition of CDSS may be related to the continued evolution of the technology. Early CDSS focused on providing clinicians with information based on existing patient data in electronic health records, such as allergies, symptoms and medical history. Today, artificial intelligence and machine-learning capabilities have helped to advance and accelerate CDSS to accomplish much more, including real-time surveillance and predictive analytics for individuals and populations of patients.
Regardless of the specific CDSS definition, in general, modern Clinical Decision Support Systems should be designed to follow the “Five Rights” of CDSS:
- Getting the right information
- At the right time
- To the right person
- In the right format
- Via the right channel
In addition to the Five Rights, workflow and technical integration is essential. Poor integration with existing IT systems has been identified as a key barrier to CDSS adoption.7 CDSS logic, for example, that is intended to provide relevant information (Right Information) to a physician (Right Person) during the order entry workflow in the Electronic Health Record (Right Format and Channel) must appear at the appropriate moment (Right Time) and be seamless to the physician.
Well-designed and functional CDSS solutions that follow the Five Rights and are highly integrated will fall into one of four categories.
These include:
- Knowledge-Based Systems – Leverage expert knowledge and clinical guidelines to provide decision support based on the patient’s symptoms, medical history and other clinical data.
- Diagnostic Decision Support – Use patient data and algorithms to identify potential diagnoses based on the patient’s symptoms, medical history and other clinical data.
- Predictive Analytics – Use machine learning and other analytical techniques to identify patterns in patient data to predict potential clinical issues. Supports healthcare providers in taking proactive measures to prevent or mitigate patients’ risks of developing specific conditions or complications.
- Clinical Guidelines and Protocols – Use algorithms to compare patient data to established guidelines and provide decision support and alerts for potential clinical issues.
The CDSS Market
Globally, demand for all categories of CDSS is growing. The worldwide market size of CDSS is projected to reach USD 12.4 billion by 2030, with an estimated CAGR of 9.5% from 2022 to 2030. The United States dominates CDSS market share at 46%, followed by Europe (25%) and Asia Pacific (20%).
The CDSS market is competitive but highly fragmented, featuring a mix of global, regional and local suppliers, some of which hold dominant positions due to brand image and market regulations. CDSS solutions can be embedded within software products, including electronic health records (EHR). Examples of EHR vendors that offer CDSS capability are Oracle Cerner, Epic, Meditech and NextGen, along with many others.
Large conglomerates that may cross industries and offer software and other products and services also offer CDSS, including Siemens AG, Philips and McKesson.
Publishing companies like Elsevier B.V., EBSCO Dynamic Health and Wolters Kluwer focus more on the content side of CDSS.
Lastly, standalone CDSS suppliers have emerged and entered the market with niche products. These vendors include companies like Biofourmis (clinical surveillance), Cedar Gate Technologies (health benefits), IBEX (pathology) and INFINX (prior authorization), to name a few, with many others on the horizon.
Benefits and Limitations of CDSS
The Agency for Healthcare Research and Quality (AHRQ) states that clinical decision support can potentially lower costs, improve efficiency and reduce patient inconvenience. In fact, according to AHRQ, CDS can sometimes address all three of these benefits at the same time — for example, by alerting clinicians about possible duplicate tests a patient may be about to receive.8
While use cases and published studies of CDSS benefits continue to emerge, the major benefits can be summarized as follows:
Major Benefits of CDSS
Improve Patient Outcomes
CDSS improved patient outcomes by facilitating accurate and efficient diagnosis, standardizing appropriate care, enhancing cost savings, and engaging patients with personalized health information and education.
Additionally, CDSS has the ability to support patient-specific information, such as genetics, medical history and current conditions, in an effort to recommend personalized treatment plans. This approach enhances the efficacy of treatment and reduces adverse effects by tailoring interventions to individual patient profiles.
Reduce Medical Errors
CDSS can reduce medical errors by providing clinicians with timely, relevant and updated information at the point of care, promoting adherence to evidence-based guidelines and best practices, and alerting providers to potential errors or adverse events.
Increase Efficiency
CDSS increases efficiency by providing clinicians and decision-makers with real-time, data-driven insights that support more accurate, timely and informed decision-making.
Improve and Accelerate Telemedicine
As telemedicine and remote patient monitoring continue to become more prevalent, CDSS can aid healthcare providers in making accurate diagnoses and treatment decisions even when interacting with patients from a distance.
Reduce Cost
CDSS reduces cost by improving the accuracy and appropriateness of clinical decision-making, reducing unnecessary testing and treatments, and minimizing errors, adverse events and service denials.
Medical Research and Education
CDSS systems provide access to the latest medical research, guidelines and best practices, ensuring that healthcare professionals and researchers base their decisions on the most current reliable information available.
CDSS can also assist researchers in identifying suitable candidates for clinical trials based on specific criteria. This speeds up patient recruitment, trial enrollment and data collection, thereby expediting the research process.
Lastly, CDSS can serve as a valuable education tool, enabling healthcare professionals to stay updated on the latest medical research, advancements and treatment approaches. This fosters a culture of continuous learning and professional development and can assist with transition to practice for all clinicians.
The benefits of CDSS can also be broken down by market segment. Further detail on the benefits of CDSS by market segment are summarized below:
Benefits of CDSS by Market Segment
Hospitals and Health System
As a result of the HITECH Act and subsequent Meaningful Use regulations, by 2017 more than 90% of U. S. hospitals and 80% of clinics had implemented electronic health records with some form of clinical decision support.
A 2020 study by Sutton, et al., suggests that at least 41% of hospitals that have an EHR also had a CDSS, while 40.2% of U.S. hospitals had cutting-edge CDSS capability.9
Benefits:
- Improved patient care and outcomes
- Improved medication management
- Reduction in unnecessary imaging tests
- Real-time clinical alerts
- Management of chronic disease
- Following evidence-based clinical pathways
- Improved diagnostic accuracy
Ambulatory Practices
CDSS utilization in the U.S. varied from 68.5% to 100% in physician-owned or group-owned primary care practices with EMRs/EHRs, and from 44.7% to 96.1% regardless of EMR/EHR status.10
Benefits:
- Disease management
- Diagnostic support
- Support clinical workflow optimization
- Improved care delivery
- Improved patient safety
- Identify potential drug interactions or contraindications
Remote Patient Monitoring and Telehealth Providers
The global remote patient monitoring devices market is projected to grow from USD 30.05 billion in 2021 to USD 101.02 billion in 2028, at a CAGR of 18.9% during the forecast period.11
Benefits:
- Improved medication management
- Reduction in unnecessary imaging tests
- Real-time clinical alerts
- Management of chronic disease
- Following evidence-based clinical pathways
- Improved diagnostic accuracy
Long-Term Care Facilities
The demand for long-term care facilities is expected to increase due to the aging population, prevalence of chronic diseases, caregiver shortages and changing family structures.
CDSS offers the promise to improve quality in the long-term care setting and to assist with continued staff shortages.
Benefits:
- Recommending appropriate interventions and therapies
- Identifying potential adverse drug reactions
- Preventing hospital readmissions
- Medication management
- Coordinate personalized care
- Improve the quality of care
- Minimize risks
- Reduce costs
Pharmaceutical and Life Science Companies
In the pharmaceutical industry, CDSS can aid in drug discovery and development by analyzing vast amounts of molecular data to identify potential drug targets, predict drug interactions and optimize drug design. This can accelerate the drug development process and reduce costs.
Additionally, use of CDSS is expected to assist with clinical trials, where it is reported that globally more than 80% of clinical trials faced delays due to patient recruitment challenges.12
Benefits:
- Drug research and development
- Drug interaction predictions
- Clinical trial enrollment
- Drug and device surveillance
- Drug safety monitoring
- Patient safety
Government Agencies and Ministries of Health
Across the globe, government agencies and ministries of health are focused on improving healthcare access while managing costs.
Benefits:
- Quality improvement
- Disease outbreak surveillance and tracking; identifying potential public health threats in real time
- Health policy development
- Disaster response
- Improving health information exchanges
- Improving data security and efficiency
Payers
The demand for CDSS in the payer segment is expected to increase due to the growing complexity of healthcare delivery, complex disease management, increasing pressure to control costs, and the shift towards value-based care.
Benefits:
- Improving care management
- Support of value-based care programs
- Utilization management
- Fraud detection
- Population health management
- Risk assessments
Laboratories and Diagnostic Testing
The Centers for Disease Control and Prevention estimates that 70% of medical decisions depend on laboratory test results.13 CDSS can improve the accuracy of bladder tumor grading and estimated recurrence by 93%.14
Benefits:
- Enhanced laboratory and diagnostic testing and interpretation
- Diagnostic imaging interpretation
- Alerts and reminders for abnormal results
- Personalized reference ranges
- Automated interpretation (e.g., tumor grading, EKG and blood gas interpretation)
Limitations of CDSS
There are also well-documented limitations and challenges with CDSS. It is common practice for organizations to defer CDSS design, build, implementation and maintenance to the information technology function instead of to operations. CDSS are often initially managed by an application support team in the Information Technology department. This can seem to make sense, given that CDSS functionality is often included with and part of larger IT-focused initiatives such as EHR implementation. However, clinical and operational leadership, oversight and influence is essential both in the beginning phases and over the long term to ensure that CDSS is developed and configured in a way that supports clinical practices.
Furthermore, the content used to inform CDSS and the underlying data built into the logic needs to be vetted and approved by clinical and operational leaders both during and after implementation. In the absence of clinical and operational oversight, the CDSS may frustrate end-users or result in a lack of end-user confidence, leading to decreased adoption and use of CDSS.
Clinical content is another challenge. The rapid development of research and clinical evidence described earlier and the proliferation of new medical knowledge make it challenging to keep CDSS current. Further complicating matters, a CDSS solution often brings together multiple vendors. The EHR vendor, for example, may provide the technical infrastructure for CDSS but will often integrate with a third-party content provider to offer full CDSS functionality. Strong vendor management and a clinically led organizational oversight committee needs to be in place so that CDSS can be properly governed and maintained over the long term.
As mentioned, workflow and clinical practice integration is essential but is one of the challenges with the most risk. Poorly implemented CDSS can frustrate end-users and lead to decreased utilization of CDSS, for example, when clinical workflow is unnecessarily interrupted.
A physician, for example, ordering ibuprofen for a nine-year-old patient could become frustrated with a pop-up window that warns of the dangers of prescribing this medication to pregnant women. The pop-up may even require additional clicks to describe why it is being overridden, forcing the clinician to spend unnecessary time documenting reasons for ignoring the pop-up. The data exists within the record to make the CDSS “smarter” and to suppress this warning when it is not relevant. This example is one of many to consider in highly variable and complex care delivery environments, and points again to the need to understand the clinical practice environment within which the CDSS will operate.
Monitoring the ongoing use of CDSS and capturing the metadata that describe how CDSS logic is applied in practice is also an essential but often overlooked requirement. While many vendors of CDSS provide the reporting capability required to monitor CDSS utilization, it is up to operational leaders to review CDSS metadata, including end-user utilization, and to respond and react to those findings.
Lastly, there is an ongoing maintenance cost to any CDSS. While many organizations appreciate the upfront investments required for CDSS, few understand the long-term costs associated with maintaining CDSS. Ongoing fixed and variable costs need to be considered and included in annual budget reviews. According to the CDC the cost to develop the CDSS includes the cost of compiling evidence-based narrative guidelines and programming the guidelines and decisions into code to produce prompts for provider action. Resources are needed to then implement the system throughout the practice and for all providers. Likewise, the day-to-day use and maintenance of the CDSS requires staff time and other resources and are categorized under operating cost. In summary, the components of capital cost are development and implementation, and the components of operating cost are maintenance and operation.15
CDSS Risk-Mitigation Strategies
To reduce risk of harm, developers of CDSS solutions must ensure that products are designed with the patient at their core. Furthermore, those who adopt and use CDSS must likewise take a patient-centered approach to deployment of CDSS in the clinical environment in order to reduce the risk of patient harm.
In July 2023, the U. S. Agency for Healthcare Research and Quality (AHRQ) published “Trust and Patient-Centeredness Workgroup: Improving the Source Credibility of Patient-Centered Clinical Decision Support Tools.” For CDSS developers, the authors specifically outline design attributes that must be in place to ensure source credibility of the CDSS. These include accuracy, consistency, objectivity, reliability, currency, relevance, transparency, expertise, competence and usability.
The authors emphasize that end-users of CDSS must find the solution and the information to be authoritative, appealing, receptive and relatable.16
Collectively, these patient-centered characteristics of CDSS design and deployment can begin to help those who develop and use CDSS to set a standard for risk mitigation.
Additional CDSS risk-mitigation strategies across multiple different domains for all healthcare leaders to consider are presented below.
CDSS Risk-Mitigation Strategies Across Domains
Governance – Do you have strong oversight over CDSS?
Risk:
- Patient safety
- Legal
- Poor CDS utilization
Mitigation Strategy:
- Interprofessional clinical content oversight committee
- Change management focus with strong strategic communications plan
Rationale:
- Strong governance at enterprise level required to oversee CDSS content, use and effectiveness while considering implications for high-quality patient care
- End-user engagement through strong change management program with robust communication plan
Evidence – How do you incorporate new and emerging clinical evidence into existing and proposed CDSS?
Risk:
Mitigation Strategy:
- Interprofessional clinical content oversight committee
- Trusted clinical content provider
Rationale:
- Same as above
- Important to recognize that the tool to deliver CDSS (e.g., electronic health record) may not be the same partner for clinical content; know your vendors and know where their solution and products begin and end
CDSS Tool Design – Is the CDSS tool and content designed to reduce risk of patient harm?
Risk:
Mitigation Strategy:
- Healthcare organizations partner with CDSS vendors and suppliers who follow CDSS design risk-mitigation strategies
- Healthcare organizations implement CDSS end-user risk-mitigation strategies
Rationale:
- As CDSS solutions continue to evolve and mature, well defined and patient-centered CDSS design and use standards will help reduce risk of harm
- Read AHRQ’s “Trust and Patient-Centeredness Workgroup: Improving the Source Credibility of Patient-Centered Clinical Decision Support Tools”
Clinical Practice and Technical Integration – How is CDSS integrated into Clinical workflow?
Risk:
- Poor adoption of CDSS
- Failure to achieve CDSS benefits
- Alert fatigue
- Other unintended consequences of CDSS
Mitigation Strategy:
- Clinically led workflow design, testing, implementation and monitoring
Rationale:
- CDSS is only effective when it is introduced at the right time, in the right context and is applied; see “Five Rights” above
- Workflows and clinical practice change frequently and CDSS must evolve with practice.
CDSS Source Data – Do you have strong data governance in place for the underlying data that is part of CDSS?
Risk:
- Unintended consequences of CDSS
Mitigation Strategy:
- Data Governance Committee that promotes adoption of data standards where possible
Rationale:
- Garbage in; garbage out: poor quality of underlying data could have significant impact on CDSS output and reliability
CDSS Metadata – Is CDSS effective?
Risk:
- CDSS is ignored
- Patient safety
Mitigation Strategy:
Rationale:
- Formal audit of CDSS utilization by practitioner; review by Oversight Committee
Legal, Risk and Ethics – Does CDSS introduce bias? What are the legal and regulatory implications?
Risk:
- Noncompliance
- Litigation
- Reputation/Brand
Mitigation Strategy:
- Include Legal/Compliance/Risk in Oversight Committee
Rationale:
- 100% compliance with applicable laws and regulations
- Alignment with enterprise-level policy and standards
Funding
Risk:
- Obsolete CDSS that introduces risk
Mitigation Strategy:
- Adequately fund and resource CDSS
Rationale:
- CDSS requires continued investment beyond initial implementation
Conclusion
High-functioning Clinical Decision Support Systems hold the promise of improving healthcare quality. Healthcare leaders must recognize the critical role that evolving clinical content plays in the effective use and application of CDSS in clinical practice. Furthermore, CDSS must be fully integrated with clinician workflows and be presented to users at the right time, in the right context, and with valid, peer-reviewed information display. Enabling CDSS technology (the hardware and software) is arguably the easy part. The difficulties and risks with CDSS are related to the content and workflow integration and ongoing maintenance that is required to keep CDSS current with medical knowledge. Strong, collaborative clinical leadership and partnership with those implementing the CDSS are paramount to its adoption and success.
While there is great opportunity to leverage the potential of CDSS, there is also risk. As with any technology, the risk must be managed. Savvy healthcare leaders will understand the benefits and limitations of CDSS and will implement risk-mitigation strategies to keep patients safe and, above all else, to do no harm.
Footnotes:
1: Peter Densen, “Challenges and Opportunities Facing Medical Education,” National Library of Medicine (2011)
2: AMIA. (2023, July 7). Clinical Decision Support. Retrieved from AMIA
3: “FDASIA Health It Report – Onc.” Health IT, April 2014
4: Wasylewicz, Arthur T M, and A. M. J. W. Scheepers-Hoeks. 2018. “Clinical Decision Support Systems.” In Springer eBooks, 153–69
5: US Office of the National Coordinator for Health IT. (2017, September 14). Health IT Terms. Retrieved July 9, 2023, from Health IT.Gov
6: NHS England. (2023, August 16). Supporting clinical decisions with health information technology. Retrieved September 5, 2023, from NHS England
7: Sharma, Videha, Ibrahim Ali, Sabine N Van Der Veer, Glen P. Martin, John Ainsworth, and Titus Augustine. 2021. “Adoption of Clinical Risk Prediction Tools Is Limited by a Lack of Integration with Electronic Health Records.” BMJ Health & Care Informatics 28 (1): e100253
8: “Clinical Decision Support.” n.d. Agency for Healthcare Research and Quality
9: Sutton, Reed T., David Pincock, Daniel C. Baumgart, Daniel C. Sadowski, Richard N. Fedorak, and Karen I. Kroeker. 2020. “An Overview of Clinical Decision Support Systems: Benefits, Risks, and Strategies for Success.” Npj Digital Medicine 3 (1)
10: Jing X, Himawan L, Law T. Availability and usage of clinical decision support systems (CDSSs) in office- based primary care settings in the USA. BMJ Health Care Inform 2019;26:e100015. doi:10.1136/ bmjhci-2019-100015
11: Fortune Business Insights. (2023, August 12). Retrieved from link
12: Desai, M. (2020, June 5). Perspectives in Clinical Research. Retrieved from Challenges of recruitment and retention of participants in clinical studies
13: Center for Disease Control and Prevention. (2023, August 12). Division of Laboratory Services. Retrieved from CDC
14: Id
15: J Am Med Inform Assoc. 2017 May 01; 24(3): 669–676. doi:10.1093/jamia/ocw160
16: Hongsermeier, T., Dobes, A., Cope, E., Dullabh, P. M., Desai, P. J., Dungan, R., . . . Weinberg, S. (2023). Trust and Patient-Centeredness Workgroup: Improving the Source Credibility of Patient Centered Clinical Decision Support Tools. Agency for Healthcare Research and Quality. Rockville, MD: Agency for Healthcare Research and Quality. Retrieved Sepetember 5, 2023