Clinical Decision Support – what does it mean?

Clinical decision support is a feature of Electronic Health Records (EHRs) that is felt to be central and crucial to contemporary, certified EHR technology.

In fact, both the Meaningful Use NPRM (notice of proposed rule-making) and the EHR Certification IFR (interim final rule) define a Qualified EHR as something that (A) includes patient demographic and clinical health information, such as medical history and problem lists; and (B) has the capacity to (i) provide clinical decision support, (ii) support physician order entry, (iii) capture and query information relevant to health care quality, and (iv) exchange electronic health information with, and integrate such information from other sources.

Unfortunately, the concept of “clinical decision support” is not precisely defined, and means different things to different people. Often, when engaged with other individuals and companies in the Health 2.0 space, conversations about “clinical decision support” are muddied by different assumption of what, precisely, that means.

Simple clinical decision support. The simplest form of clinical decision support (CDS) is alerting against drug-drug, and drug-allergy prescribing errors. Quite simply, alerting a practitioner against prescribing a medication to which the patient has a past declaration of allergy constitutes CDS. Helping the practitioner avoid medication errors like this is an important feature of EHRs, and is, technically, CDS.

However, Meaningful Use expects more than this. In fact, it specifically requests that a “meaningful user” implement 5 “clinical decision support rules (in addition to drug-drug and drug-allergy contraindication checking) according to specialty or clinical priorities that use demographic data, specific patient diagnoses, conditions, diagnostic test results and/or patient medication list.”

Compliance with quality metrics. The central concept referenced in the CDS intent for Meaningful Use is tied to working with clinical quality metrics. A “quality metric” would be something like “diabetic patients should have a glycohemoglobin blood test done within the past year (or more often).” An EHR technology should be able to present to the clinician an alert in the individual patient record stating “this patient is a diabetic, but has not had a glycohemoglobin done within the past year – he/she is due for one.” Alternatively, the EHR technology should be able to generate a report – a list – of all patients (using the same example) who are diabetics but have not had a glycohemoglobin done within the past year.

Quality metrics are widely accepted through a rigorous evidence-based process. There are two main sets of quality metrics in use: (1) PQRI (physicians quality reporting initiative), which is overseen by CMS (the center for Medicare and Medicaid services), and (2) HEDIS (healthcare effectiveness data and information set), which is developed by the NCQA (national committee for quality assurance).

PQRI measures are used for Medicare pay-for-performance, which has been in place prior to ARRA/HITECH Meaningful Use incentives. There are over 170 PQRI measures, grouped into 13 different categories. Their thrust tends to be more Medicare-focused – for example, in the Preventive Care Measures Group in PQRI includes adult/senior-oriented measures: (1) screening for osteoporosis, (2) screening for urinary incontinence, (3) influenza vaccination for patients >50, (4) pneumonia vaccination for patients >65, (5) mammography screening, (6) colorectal cancer screening, (7) screening about tobacco use, (8) advising tobacco users to quit, (9) body mass index screening, and (10) screening for unhealthy alcohol use.

By contrast, the HEDIS data set has been adopted by many HMO and other private health insurers in private-sector pay-for-performance activity. The HEDIS measures include measures more appropriate for children and adolescents (not generally addressed by the Medicare-centered PQRI measures). For example, the Effectiveness of Care domain in HEDIS includes measures such as (1) childhood immunization status, (2) immunization for adolescents, (3) lead screening in children, (4) Chlamydia screening in young women, (5) appropriate testing for children with pharyngitis, and so on – the HEDIS measures also include many of the PQRI measures, such as BMI testing, breast cancer screening, colorectal cancer screening, etc.

Because Meaningful Use incentives are being deployed through CMS, and paid out via Medicare or Medicaid channels, their assumptions on quality metrics lean towards PQRI. Not surprisingly. For EHR technology, it makes sense to start with adopting and implementing PQRI measures, and then later expanding to include HEDIS measure as well.

Diagnosis support and expert systems. Some individuals and companies, when they talk about Clinical Decision Support, are actually referring to Diagnosis Support – given a patient with “symptom set x” and with “lab test results y,” give me the likely diagnoses, with percent probabilities of each one, and recommended further testing to distinguish between them.

Companies have been trying to build these kinds of Expert Systems for many years, and have been challenged either by limited scope (“our expert system only is valid for patients with certain kinds of kidney disease”), or by limited adoption. Not being integrated into tools used day-to-day, they have been relegated, at best, to be side-bar applications. Some clinical disciplines make more use of such Expert Systems – oncology, for example, uses look-up of available newest recommended protocols for tumor x. Sometimes these systems are simply look-up portals (for example, Up To Date, which is a popular one) that allow physicians to review the current literature around a given situation.

From the standpoint of the current state of Qualified EHR Technology as defined by Meaningful Use, however, “clinical decision support” is not referring to using Expert Systems. It is referring to implementing compliance with Quality Metrics.

Systematic use of Quality Metrics (regardless of which measure set is used) will go a long way to improving general health care quality. One of the reasons for less-than-top-quartile quality seen in U
.S. healthcare (compared to that of other developed countries) is the lack of tools that allow physicians to measure their own quality performance at the point of care. Including these kinds of features in modern EHR technology is part of the vision of a transformed healthcare system.

Robert Rowley, MD
Chief Medical Officer, Practice Fusion, Inc.

Practice Fusion draws from a community of doctors, medical experts, and digital health influencers that contribute to blog posts. Read all posts from our guest writers

Read all posts by Contributing Writer