The risks and benefits of many diagnostic approaches and treatments differ for older adults compared to middle aged adults. When diagnostic and therapeutic strategies are misapplied to older adults this can lead to increased morbidity and mortality. Well established examples where clinicians do not often follow best practices in the care of older adults include those identified by the American Geriatrics Society for the Choosing Wisely initiative: 1) testing and treatment for asymptomatic bacteriuria, 2) prostate specific antigen testing in older men without prostate cancer, and 3) overuse of insulin or oral hypoglycemics for type 2 diabetes.

Clinical decision support nudges, informed by social psychology and delivered via electronic health records (EHRs), are promising strategies to reduce the misuse of services in cases where optimal utilization may not be zero but should be well below current practice. These interventions seek to influence conscious and unconscious drivers of clinical decision making, are low cost to implement and disseminate, and can be incorporated into existing delivery systems. In the R21 phase of this Behavioral Economics Applications to Geriatrics Leveraging EHRs (BEAGLE) study, we will: select EHR delivered nudges to address 3 topics of potential misuse in older adults based on the main psychological drivers of overuse identified in interviews with high-using clinicians; develop and pilot test decision support tools within a health systems' EHR to understand technical feasibility, work flow fit, preliminary impact on clinical outcomes, and clinician acceptability; and develop and validate electronic clinical quality measures of potential overuse/misuse related to the care of older adults.

The risks and benefits of many diagnostic approaches and treatments differ for older adults compared to middle aged adults. When diagnostic and therapeutic strategies are misapplied to older adults this can lead to increased morbidity and mortality. Well-established examples where clinicians do not often follow best practices in the care of older adults include those identified by the American Geriatrics Society for the Choosing Wisely initiative: 1) testing and treatment for asymptomatic bacteriuria, 2) prostate specific antigen testing in older men without prostate cancer, and 3) overuse of insulin or oral hypoglycemics for type 2 diabetes. There are several hypotheses as to why clinicians fail to incorporate best evidence into geriatric clinical care. First, they may underestimate downstream harms of testing which seems easy to do (e.g., a urinalysis for a non-specific symptom) or treatment that may be appropriate for younger patients (e.g. intensifying insulin to achieve tight control). Second, clinicians may overweigh the risks of not performing the action (e.g., missing cancer diagnosis, failing to diagnose urinary tract infection (UTI) in a patient presenting without urinary tract symptoms). Third, clinicians may respond to real or perceived social norms (from patients and their families, other clinicians or both) that set expectations to behave in specific ways. Fourth, force of habit may lead clinicians to act in a way similar to how they have done in the past even if current evidence doesn't support it. And fifth, clinicians may overuse a test or treatment to avoid feeling they are expressing an ageist bias toward their patients. Clinical decision support nudges, informed by social psychology and delivered via electronic health records (EHRs), are promising strategies to reduce the misuse of services in cases where optimal utilization may not be zero but should be well below current practice. These interventions seek to influence conscious and unconscious drivers of clinical decision making, are low cost to implement and disseminate, and can be incorporated into existing delivery systems. In the R21 phase of this Behavioral Economics Applications to Geriatrics Leveraging EHRs (BEAGLE) study, we will: select EHR delivered nudges to address 3 topics of potential misuse in older adults based on the main psychological drivers of overuse identified in interviews with high-using clinicians; develop and pilot test decision support tools within a health systems' EHR to understand technical feasibility, work flow fit, preliminary impact on clinical outcomes, and clinician acceptability; and develop and validate electronic clinical quality measures of potential overuse/misuse related to the care of older adults.

Inclusion Criteria:

• Northwestern Medicine primary care clinician caring for adults
• Provides informed consent

We report age and sex for PSA and UA/UC outcome measures. We have not collected sex for patients eligible for diabetes outcome measure.

6 month intervention period

1) female patients with an emergency department visit or hospitalization with diagnosis of UTI or sepsis following visit with Clinical Decision Support (CDS) exposure and 2) diabetes patients with an emergency department visit or hospitalization for hyperglycemia following a primary care visit with CDS exposure. Number of patients at risk is not equal to all enrolled subjects. Only subgroups of patients receiving CDS are in the denominator for adverse event reporting.