Clinical Decision Support Systems(CDSS)
Introduction#
Last time, we talked about data sources and coding system. Electronic Health Record system is where the patient’s records being stored. Different coding system(ICD-9, ICD-10, SNOWMED-CT, etc) are being used to ensure consistency and communication. This post will talk about Clinical Decision Support System(CDSS), a computer-based programs that analyzes data from EHR.
Clinical Decision Support System(CDSS)#
Clinical Decision Support System is used for assisting patient-related decision making process. It can effectively determine the optimal treatment, estimating clinical and financial outcomes of different treatment. Therefore, CDSS has largely improved the pharmacy and billing system. Pharmacies use CDSS to check for negative drug interactions. Billing department uses CDSS to devise the treatment plan and provide the optimal financial expenses.
| CDSS | Language | Usage | pros/cons |
|---|---|---|---|
| VisualDX | JAVA-based CDSS | Visual aid in assisting surface-level illness. Use machine process to matches images of patient’s abnormalities with pre-existing images within database. | Written summary is brief, and thus, unclear. |
| DXplain | Web-based | Simple diagnosis by entering medical terms, and return potential diagnoses from knowledge base. | Simple, cost reduction, accurate. |
| Isabel | Web-based | Contains two sub-systems: diagnostic checklist utility and knowledge mobilizing utility. Returned recommended diagnoses and research purposes. | Exceptionally accurate |
Knowledge-Based CDSS#
Knowledge-based CDSS contains built-in reference table with data of various diseases and treatments. Usually, the patient’s data is entered through a computer-based system, and clinician manipulates the system to get results. The output is in the form of ranked list of solutions in ASCII text or graphical format.
Nonknowledge-Based CDSS#
Unlike knowledge-based in which it possess some form of medical knowledge, Nonknowledge-based system used artificial intelligence to learn from previous experiences. One of the types of Nonknowledge-based system is called Artificial Neural Network(ANN), in which it stimulates human thinking and studies disease pattern. The ANN contains input layer, hidden layer and output layer, where the input layer receive data, output layer finalized results. Because ANN doesn’t have predefined knowledge base, it cover much smaller range of disease compared to Knowledged-based system.
Challenges#
Despite all the advantages of CDSS, some physicians choose not to use them. One of the concerns is called machine adaptability, which describes the ability of machine to learn new medical knowledge from outdates information. The other concern such as medication errors, which may lead to harmful outcomes to patients and unnecessary procedures. Moreover, it contribute to clinician frustration and burnout because of poor design interface and excessive alarms received.
Conclusion#
CDSS is an important aspect of EHR to support decision making in healthcare industry. However, the implementation of CDS may causes frustration of clinicians. There are several ways to improve the design of CDS as suggested by researchers. For example, CDS should only save problems that clinician needs and incorporate patient-specific information. Customization is needed to minimize the alert burdens(Jankovic and Chen, 2020)
References
Jankovic I, Chen J H. Clinical Decision Support and Implications for the Clinician Burnout Crisis. Yearb Med Inform. 2020:145–54.