Intervene Thoughtfully at the Exact Moment You Have the Ability to Change the Outcome
Opportunities to Add Care:
Medicare Advantage Star
Physician Quality Report System (PQRS)
Merit-Based Incentive Payment System (MIPS) Quality Measures
Opportunities to Improve Quality:
Opioid Management and De-prescribing
Stroke Prevention in Atrial Fibrillation
Unnecessary Care in the ED
The Million Hearts Cardiovascular Risk Reduction Program
CARA (Cardiac Bundles)
Opportunities to Reduce Excess Cost, Waste and Harm:
High Cost Medications
High Cost Labs
How Does PINC AI™ Stanson CDS Work?
The Stanson content selection process identifies evidence-based recommendations derived from highly reputable sources and relevant to population health management. Key clinical information, supporting references and implementation logic (how the information is gathered from the Electronic Medical Record (EMR) and where it fits best within typical workflows) are developed by clinicians and informaticists through a multi-step process before being sent to an EMR build team for programming and validation. Interventions are designed to use clinical and other data that are generally available, so that the CDS functions as intended across institutions.
A large body of published research shows that accurate CDS targeting and integration into routine clinical workflows is required for changing provider behavior and achieving intended goals. These workflows may vary by whether recommendations are intended to address omissions in care or potential overutilization. For example, many recommendations from the ABIM’s Choosing Wisely® initiative are intended to reduce low-value and unnecessary care. Clinical decision support derived from these recommendations is typically delivered as a “pop-up” message (alert) when an order is placed, in the appropriate clinical setting.
To determine the right circumstances in which to trigger an alert, a team of Stanson clinicians works to turn a recommendation into Boolean logical statements, using up to 30 different data elements in the EMR to create inclusion and exclusion criteria. These data elements include ICD-10 codes used in past medical history, current problem lists and encounter diagnoses as well as lab tests and results, vital signs, procedure codes, medications, provider type and specialty, department and venue, and an array of additional information relevant to accurately characterizing specific clinical scenarios.
The PINC AI™ Stanson CDS development process emphasizes accuracy to avoid the erroneous triggering of alerts fires, or “false positives,” to minimize alert fatigue and maintain credibility with providers. The alerts have a standardized, clear and efficient appearance, to facilitate recognition and promote adherence to the recommendation, as appropriate. For alerts based on Choosing Wisely®, a one sentence quote from the authoring medical society is presented, along with hyperlinks to the more detailed explanation supporting the relevant recommendation on the dedicated ABIM website. Additional hyperlinks are provided for references cited in the recommendation and companion patient education created by Consumer Reports®. The latter is included to promote efficient and effective informed discussions by physicians and patients, leading to better care decisions and improved patient satisfaction. Override reasons, derived from the detailed recommendation text and accompanying references, are made available as buttons, so that providers can address an alert with a single click of a mouse.
Sample Content from the American Board of Internal Medicine Foundation’s Choosing Wisely Initiative®
- Don’t screen for carotid artery stenosis (CAS) in asymptomatic adult patients.
- Don’t use benzodiazepines or other sedative-hypnotics in older adults as first choice for insomnia, agitation or delirium.
- Don’t do imaging for uncomplicated headache.
- Don’t administer erythropoiesis-stimulating agents (ESAs) to chronic kidney disease (CKD) patients with hemoglobin levels greater than or equal to 10 g/dL without symptoms of anemia.
- Don’t perform PET, CT and radionuclide bone scans in the staging of early prostate cancer at low risk for metastasis.
- Don’t prescribe testosterone therapy unless there is biochemical evidence of testosterone deficiency.
- Don’t use PET imaging in the evaluation of patients with dementia unless the patient has been assessed by a specialist in this field.
- Don’t recommend percutaneous feeding tubes in patients with advanced dementia; instead, offer oral assisted feeding.
- Don’t perform cardiac imaging for patients who are at low risk.
- Avoid transfusions of red blood cells for arbitrary hemoglobin or hematocrit thresholds and in the absence of symptoms of active coronary disease, heart failure or stroke.
- Don’t perform repetitive CBC and chemistry testing in the face of clinical and lab stability.
- Don’t place, or leave in place, peripherally inserted central catheters for patient or provider convenience.
- Don’t do work up for clotting disorder (order hypercoagulable testing) for patients who develop first episode of deep vein thrombosis (DVT) in the setting of a known cause.
The Stanson Health CDS Advantage
An essential element of Clinical Decision Support (CDS) is the appropriate identification of patients and clinical scenarios. The PINC AI™ Stanson content development process focuses on the needs of practicing clinicians, creating CDS that is triggered when appropriate and user-friendly, fits into typical workflows and provides clear and credible direction. As clinical logic is often highly complex and challenging to break down into logical statements that can be processed by an EMR, PINC AI™ Stanson deploys and validates interventions in a live patient care environment prior to their inclusion in the CDS library.
Each intervention is its own quality improvement project and therefore needs to be individually assessed. After deployment, the performance of new CDS is evaluated for accuracy, physician responses (was the recommendation presented in the alert followed, overridden or ignored) and specific override reasons given. The alert is then optimized in an iterative process until performance characteristics warrant inclusion in the PINC AI™ Stanson CDS library.
Winning Broad Physician Acceptance
Physician acceptance and avoidance of alert fatigue are key issues that challenge successful deployment of CDS. The PINC AI™ Stanson content selection process emphasizes topics that afford significant opportunities to improve quality and reduce costs, and that have face validity with the general community of physicians. The net result of this careful selection and extensive development process is the achievement of significant improvements in outcomes while generally avoiding alert fatigue. Accurate targeting of relevant clinical scenarios has resulted in the majority of physicians encountering few alerts (e.g., 0-5 per month), despite the deployment of hundreds of interventions. A smaller proportion of providers see up to one alert daily, while a very small subset may see alerts more frequently.