When compared with other types of professionals, physicians have been relatively slow to adopt new technologies. This is now changing, however, with enactment of the Health Information Technology for Economic and Clinical Health Act (HITECH), passed in 2009. The legislation mandates that health care providers demonstrate “meaningful use” of health care IT, or face penalties in the form of reduced Medicare payments, beginning in 2014. One of the provisions of the meaningful use guidelines is that hospitals and doctors adopt the use of clinical decision support systems (CDSS).
As use of CDSS becomes more widespread, the health care industry will rely on computer science professionals with a health informatics degree to ensure these systems can be integrated smoothly into existing IT infrastructures and are compliant with privacy regulations like HIPAA.
The Basics of Clinical Decision Support
Doctors and other health professionals use CDSS to help prevent medical errors and to improve overall patient outcomes. These systems link physician observations and current patient data to a database of health knowledge, to provide customized alerts, diagnostic information, and treatment protocols that influence the decisions clinicians make about patient care. CDSS can also employ data mining to examine a patient’s medical history in the context of relevant clinical research, which can help physicians adhere to best practices in diagnosis and treatment.
The underlying idea is that when doctors combine their own knowledge with decision support, they can make a better analysis of patient data than either a human or an artificial intelligence system could make alone. Today’s medical knowledge base is too vast for any one person to remember – so it only makes sense for physicians to rely on smart systems to guide them in their decision making.
Most clinical decision support systems consist of three parts:
- The knowledge base, which contains medical data.
- The inference engine (a set of rules derived from established treatment protocols, evidence-based medicine, standardized drug dosages, etc.)
- The communications mechanism, which relays the results to the user and usually includes various tiers of alerts.
Often, CDSS relies on “if-then” rules to create alerts. For example, a common clinical use of decision support is to avoid adverse drug interactions – so a rule might be that IF Drug A is already being taken, AND Drug B is prescribed, THEN the system alerts the user of a potentially dangerous interaction. In this case, the inference engine is combining data from the knowledge base with data from the patient’s electronic medical record. In other instances, the system may consider a patient’s symptoms, lab results, medical history, family history, and genetics, in addition to epidemiological trends of disease occurrence and published clinical research on a condition or drug regimen. Physicians have the ability to over-ride an alert if they decide it does not apply to a specific case.
As an Informaticist, Where Do You Fit In?
Hospitals and private physician practices that use CDSS rely on the expertise of health IT professionals in the following areas:
- Workflow integration. Systems that require extra steps from a physician result in decreased efficiency. Decision support should not be a stand-alone application, but one that is seamlessly integrated with other systems, like the electronic medical record (EMR).
- Alert fatigue. When systems produce too many irrelevant warnings, physicians develop a tendency to override all alerts, and may miss a critical warning. Overall, the industry consensus is that the informaticians and programmers who create these systems must aim to improve the specificity of alerts to design warnings that have higher user-acceptance.
- Maintenance. In addition to routine system maintenance, CDSS requires constant data updates to ensure new drugs and treatment protocols are included.
- Evaluation. Hospitals need informaticians to devise metrics that can quantify the value of CDSS, in order to measure its effectiveness and improve its quality.
- Improvement. Informaticists also work with clinicians to design new alerts and modify existing alerts to ensure relevance to a specific patient population. For example, there have been reports of a “check pregnancy status” alert popping up for male patients. This constant need for improvements requires a clinical background and a degree in health care informatics