January 19, 2021
Clinical decision support software: facilitating adoption
No matter how advanced the recent medical developments are, medical error-caused mortality is still there. Luckily, clinicians have a powerful assistant—clinical decision support software (CDSS), and this direction of healthcare app development seems to be gradually gaining momentum. By the end of 2025, the global CDSS market value is projected to reach $1.82bn. By 2027, it is to hit yet another milestone, $2.59bn, growing at a promising CAGR of 11.9%.
Though the future looks bright, the present situation is far from being that positive. Providers don’t rush to deploy CDSS fearing complications. So what are they and how to mitigate them? We’ll consider the matter below.
A clinical decision support system is a set of programmatic tools that extract data from patients’ EHRs and other hospital systems and deliver it to medical professionals for swifter clinical decision-making.
CDSS facilitates a wide range of tasks in a clinical setting. These tasks encompass diagnostics, disease management, drug control, and more. You can see its top functions in Statista’s chart below:
As a part of EHR implementation or a standalone tool, knowledge-based CDSS, which is the most popular type, uses if-then statements (rules) to extract the information that assesses the rule and then delivers an actionable recommendation to help clinicians prevent potential adverse health events. Such databases may provide a wealth of complementary materials to help clinicians save time on searching for valid information—guidelines, data reports, templates of clinical documents, and more. Clinicians may also set up alerts and reminders to keep track of patients and follow their medication plans without interruptions.
In terms of technologies, CDSS allows clinicians to use the benefits of big data analytics in healthcare for combing through massive amounts of clinical data, as well as machine learning (ML) and natural language processing (NLP). These AI technologies contribute to streamlining clinical processes through healthcare workflow automation.
There are many ways CDSS may assist clinicians. One of the top benefits is that it reduces human factor when it comes to drug selection. The cost of medication errors amounts to $40bn per year in the US, according to a 2020 research, and 75% of them are preventable. Such errors often happen because of clinicians’ lack of concentration due to fatigue. CDSS can help to automatically prevent medication errors that may lead to allergies, incorrect dosages, drug confrontation, and more.
CDSS may also contribute to clinicians’ adoption of in-house clinical practices, workflows, and protocols. Though they are essential to any provider’s services, clinicians don’t rush to study them. With CDSS at hand, it’s possible to build the guidelines and practices into the system. This helps clinicians interiorize the processes without immense time investments, just learning them on the go.
Other areas of CDSS application include:
Evidently, the advantages of CDSS are numerous. So why do providers hesitate when it comes to the deployment?
The thing is, CDSS presents a two-sided phenomenon: there’s a potential pitfall accompanying each benefit. So what are them?
Alert fatigue and burnout. While monitoring a patient in real time is a positive trend, excessive related notifications from CDSS can do more harm than good. Endless beeping alerts of varying priorities may lead to clinicians’ fatigue, frustration and burnout. According to a recent study by the Stanford University School of Medicine, the latter is the case for 35-60% of clinicians working with CDSS.
Logically, the majority of system alerts are of low to medium importance, so medical professionals gradually start to lose focus, disregarding all alerts without consideration. Eventually, they may ignore a critical alert, causing medication errors or health incidents.
Burnout has yet another equally harmful effect. When living through burnout, clinicians lose interest in their job and their patients, which hampers good rapport and the trust established before.
Over-reliance on the system. This has to do with diagnostic assistance. While at first it may arouse doctors’ suspicions, as they get used to validating, say, medical image analysis results, they will gradually start trusting the tool. As a result, they may stop being critical of the system’s recommendations. Here’s how an expert puts it:
“One of my fears is that physicians will blindly accept the outputs of these systems going forward, and that’s not what you want.”
Director of Research Strategy and Operations at the MGH & BWH Center for Clinical Data Science, PhD
This blind acceptance may turn dangerous not only for patients but also for clinicians, or, better say, their skills. This may lead to doctors becoming more comfortable with computers than their patients.
There’s yet another challenge—healthcare interoperability. CDSS are difficult to integrate with other hospitals or systems by default. This downscales the benefits of the technology: if a patient comes from another location or medical facility, their records have to be transferred to the system manually.
It looks like the potential drawbacks of CDSS implementation may put the need for this tool under serious consideration. Luckily, these pitfalls are preventable. It all boils down to well-grounded implementation and full-scale cooperation with clinicians.
CDSS is for medical professionals to use, so it’s only logical that they should participate in the creation of such a system in their familiar environment, as it’s them who know their work-related specifics and pain points. So, a good implementation strategy starts with providers.
CDSS deployment is an effort-intensive task, so a preliminary analysis of the clinical need is required. This means clinicians themselves should be interested in the tool instead of their organizations imposing it in a top-down fashion.
This is how it worked out at South Omaha Medical Associates (NE, US). Having analyzed their hospital needs before the CDSS implementation, the provider identified cardiovascular disease treatment as a weak link in care provision and centered CDSS on its improvement. As a result, they managed to improve the relevant workflows and gain a 25% increase in patient visits.
Sometimes IT experts work independently, prioritizing their expertise and not the final goal—delivering a handy assistant for doctors. As a result, the final product lacks user-centricity, which leads to the low adoption and underuse of the tool. At the same time, involving experienced medical professionals early in the development cycle may prevent a range of issues.
First of all, it’s about fatigue, frustration, and burnout. CDSS developers should discuss the alerts and their importance with the end users and set up corresponding priorities. Another worthy effort is setting up alert personalization, i.e. tailoring them to doctors’ specializations.
Consultations with medical experts may help resolve one more poignant issue—over-reliance on the system and subsequent professional degradation. It’s important to remember that CDSS is an assistant, not a decision-maker. It’s clinicians who should decide here, and CDSS only provides enough data for a timely decision. Therefore, developers should avoid a prescriptive tone when designing the system and allow doctors to make independent decisions, even if they don’t agree with the CDSS recommendation in full.
It may also be a good idea to appeal to a valid reference in this case. If the system refers to an expert to validate its recommendation, doctors may treat it with less distrust or hostility, as the recommendation comes from a fellow professional, not a machine.
A 2020 research out of the University of Adelaide provides one more valuable insight. Looking to single out some CDSS adoption facilitators, the researchers launched a simple survey to rank them:
As we can see, the ease of use makes the most powerful facilitator (75%), but who is responsible for ensuring it? At first glance, it’s CDSS developers. However, how would they know which features make using the solution easy? This is impossible without user testing run by clinicians who are supposed to employ the solution in their work.
The users also value technical support, training, and participation fairly high at all stages of the development process. Hence, we may conclude that seamless cooperation among all the teams provides for better adoption of CDSS in a clinical setting.
And yet, there is a point that often goes missing from providers’ agenda: a successful CDSS deployment is not a one-time effort.
To keep the system up and running continuously, providers and tech experts need to ensure CDSS maintenance. This is not only about upgrading the system or its components. Knowledge bases and decision-support algorithms require maintenance, too.
The problem is, providers need to keep their knowledge bases up to speed with relevant developments and rapidly changing clinical guidelines, and it’s only medical professionals who can do it efficiently. Consequently, this effort might put extra workload on practitioners’ tight schedules or require hiring an expert with a medical background. It’s a tough choice: the former may hurt care provision and the latter may put a strain on providers’ budgets. Nevertheless, without regular knowledge base updates, CDSS is likely to deliver erroneous and even harmful recommendations.
CDSS offers a great value for the healthcare industry, with the benefits ranging from improved personalized care and prevention to streamlined clinical workflows. However, CDSS acceptance has not yet become ubiquitous due to potential challenges. As a result, clinicians might prefer to stick to tried-and-true methods and approaches to be on the safe side.
Luckily, a clear implementation strategy may push adoption forward. It requires certain efforts and concessions from the two key groups of players—providers and health IT vendors.
To facilitate CDSS adoption, providers need to select a medical specialty that could be central to CDSS deployment. When the field is chosen, it’s time for tech specialists to step in. First of all, they should let clinicians participate in the CDSS development from the start. Knowing their professional area in and out, they can help prevent pitfalls timely. Besides, in the course of development, all participating teams should lend a hand to one another when their expertise may benefit the other party. It’s also important to remember that CDSS is dynamic, i.e. regular updates, including those for the knowledge base, are required.
When a CDSS project follows a well-tuned strategy and all teams know what they do and why, the adoption is only a matter of time.
Learn how healthcare automation with NLP, RPA, predictive analytics and more intelligent tools frees up medics to treat more patients, faster.
EHR implementation is a complex process loaded with challenges. We present some tips to smoothen your EHR deployment. Have a look.
We reflect on the most up-to-date applications of big data in medicine across the care cycle.
Itransition’s healthtech consultants look at how medical image analysis redefines diagnostics and supports clinical decision-making.
AR in healthcare is making headlines by transforming invasive procedures, ADT processes, medical training, and post-discharge support for patients.
Understand the importance and challenges of interoperability for your healthcare organization, and benefit from our implementation guide.
Itransition has developed 6 versions of the medical equipment system that helps blood centers and labs collect and manage blood data.
Learn how Itransition delivered a cloud-based healthcare business process management system for a UK care homes operator, driving 2x cost reduction.