Continuing with the theme of clinical informatics, today we will touch on the topic of Clinical Decision Support Systems (CDSS), or sometimes just called Clinical Decision Support (CDS). The basic premise of clinical decision support is that it delivers timely and useful patient/disease/ or treatment-related information in an organized manner, with the goal of improving healthcare delivery/outcomes and avoiding adverse events. Although initially quite primitive in it’s functionality, CDS has blossomed into a nuanced, diverse, and rich area of study.
So what does CDSS actually look like in the hospital. Well, you interact with CDSS daily, and it rarely rises to the level of conscious thought. Sometimes this is bad, sometimes it’s great. Think computer generated EKG reads (it analyzes waveforms for patterns, in real-time, and provides you an assessment when you need it–which is as you are evaluating the patient). In it’s more sophisticated forms CDSS can help with, “adverse drug event monitoring, drug and parenteral nutrition dosing, antibiotic prescribing, ventilator management, report formatting, laboratory result alerting, blood product ordering, infusion pump monitoring, quality benchmarking, isolation bed management, clinical documentation, diagnostic and therapeutic consultation services (midddleton et. all)”. There is virtually no aspect of modern medicine that has not been touched by CDSS.
While CDSS has it’s uses, it can also contribute to alarm fatigue. Anyone using a modern EMR and treating a well appearing 35 year old with a URI, would likely agree. No, this patient is not septic. No, I don’t want fluids. No, they do not need antibiotics. When executed poorly, CDSS can feel overwhelming and self-defeating.
Where CDSS may seem annoying to a seasoned physician, it tends to shine when a provider is under-experienced, or a disease process is rare and complex. CDS can supply the inexperienced provider with relevant knowledge on the illness. Subsequently, the CDSS can formulate treatment suggestions based upon treatment guidelines. In this way, when executed well, CDS is a vital part of creating an educational cycle that is generated from analyzed and aggregated patient data, which is used to guide management, hopefully to optimal outcomes, thus creating more data for analysis and aggregation.
CDSS is still in it’s infancy, and yet it has had a tremendous effect on medicine in the 21st century. Advancements in AI and processing ability are only likely to make our reliance on CDSS more complete. Given that, it is imperative that physicians remain cognizant of the influence CDSS have, and constantly make judgments on the utility of the information that they are suppliying. In the end, it may be that clinical decision support is only as good as the clinician it is supporting.
Source: Middleton, B., et al. (2016). “Clinical Decision Support: a 25 Year Retrospective and a 25 Year Vision.” Year Med Inform 25(S 01): S103-S116.