Clinical care is often described as being a blend of art and science. The rapid digitization of health data, led by widespread adoption of electronic health records (EHRs), has created tremendous opportunities for health IT to elevate the science part of the equation.
That elevation is especially evident in the rise of clinical decision support systems (CDSS), which support clinical diagnostic processes with analysis of information from beyond the exam room and the bedside. In the process, CDSS has emerged as one of the most concrete ways in which artificial intelligence (AI) is impacting healthcare and health IT.
Fulfilling a critical need
From initial patient exam through discharge, clinicians base their care decisions on a full view of the patient’s condition. In addition to information gathered in exams, care decisions are informed by the patient’s medical history, lab and test results, and evidence-based guidelines for best care practices.
In traditional care, accounting for all this information can vary from case to case and from clinician to clinician, and key information can be overlooked or unavailable when needed. When that happens, care outcomes and patient safety can be severely compromised, extending hospital stays, contributing to readmissions, and in some cases causing death.
CDSS systems minimize these dangers with real-time diagnostic analysis and timely alerts delivered to clinicians at the point of care. CDSS is by no means a substitute for the skills clinicians acquire through years of education and experience. It is instead a unique opportunity to support clinicians with intelligence that could never have been automated before the digitization of health data.
How CDSS works
Advanced CDSS continuously monitor the documentation of care in EHR systems while communicating with unrelated information systems, such as those in labs and radiology departments, and ambulatory settings. They are also typically programmed with rules informed by established clinical guidelines.
As clinicians record care steps in the EHR, the CDSS analyzes documentation entries in real-time relative to all other available patient information. When this analysis detects that an important clinical detail has been overlooked, CDSS issues an alert. When the system performs as intended, false alarms are minimal, and those that do appear can inform care decisions in a variety of important ways:
- Suggesting a course correction in care based on patient history. One of CDSS’s primary functions is to eliminate drug-drug interactions and track patient allergies. If a physician prescribes what is normally the best medication for a specific condition but is unaware of other medications the patient is taking, or of patient allergies to the drug in question, CDSS can suggest the most effective and safe alternative. This same principle applies to many other factors in patient history.
- Detecting an unfavorable change in patient condition. Patients are constantly monitored for changes in condition while hospitalized. In many cases, time is of the essence in detecting deterioration. Consider sepsis – the presence of harmful bacteria in tissues, typically through wound infection – which is a leading cause of hospital death. CDSS systems can detect when successive lab tests and changes in patient vital signs indicate the likelihood of sepsis onset, alerting staff to intervene at the earliest possible moment. This principle can guide intervention in a number of types of deterioration.
- Improving imaging safety. Alerting clinicians that an imaging order may duplicate an existing imaging study for the patient is a basic CDSS function. Advanced CDSS can further eliminate unnecessary imaging by evaluating the need for a study relative to the patient’s overall condition and history, with the potential for recommending a different diagnostic test that avoids exposure to radiation.
These are just a few examples of the contribution that AI is making today to care improvements through CDSS. As CDSS become increasingly sophisticated, we can expect them to assume a still greater role in improving care outcomes and patient safety.
Implications for health IT
For CDSS to deliver maximum value, they must serve as effective conduits of data flow between the EHR and multiple health information systems. For maximum positive impact on care, data collection and alert delivery must happen in real-time. To deliver their value throughout a hospital or health system, this process must be sustained for all episodes of care at all times.
This represents an entirely new network demand for capacity, often involving extremely data-intensive transmissions such as when relaying imaging. It also represents an entirely new need for 100% network uptime, as a down network can mean an instant return to greater risk to patient safety and care outcomes.
Fortunately, networking technology is advancing sufficiently to keep up with these demands. In healthcare IT – as in all other forms of IT supporting dramatically new, data-intensive initiatives – software-defined WAN (SD-WAN) solutions provide the affordable bandwidth and increased uptime CDS requires. As SD-WAN takes hold within healthcare systems, patients and clinicians alike will reap new benefits from the digital transformation sweeping healthcare.
The post Health IT Spotlight: AI in Clinical Decision Support & the Impact on Health IT appeared first on Windstream Enterprise.
About the AuthorFollow on Linkedin More Content by Molly True