Integrated Clinical Prediction Rules: Bringing Evidence to Diverse Primary Care Settings (iCPR2)

University of Utah, University of Wisconsin, Madison, North Shore University Hospital, National Institute of Allergy and Infectious Diseases (NIAID)

The study is a randomized controlled trial, with an Intervention Group and a Control Group at the University of Utah (U of U) and University of Wisconsin (UW). BU serves as the primary award and coordinating institution. The unit of randomization will be at the clinic level at each institution. UW will recruit all General Internal Medicine (GIM) Clinics and Department of Family Medicine (DFM) Clinics in Dane County as well as their East and West Urgent Care Clinics. U of U will recruit all affiliated primary care practices. The unit of randomization will be the clinic. The study biostatistician will receive a list of clinic sites that have agreed to participate in the study from the site PIs. Clinics will be randomized to either Intervention group or to a Control group stratified by clinic size. Both groups will receive a single 45 minute academic detailing session describing evidenced-based diagnosis and treatment for strep throat and pneumonia. The Intervention Group will also receive a demonstration of the iCPR tool during their academic detailing session. Providers and clinic staff will be invited to the academic detailing session. Any provider or staff that is unable to attend the session will receive written and electronic copies of the material. Individual providers will not be specifically recruited for participation and they will participate or not based on personal preferences as they would for any clinic quality improvement project. The iCPR tool will be "turned on" for providers in the Intervention group. This means that the best practice alerts will trigger for appropriate patients with suspected strep throat or pneumonia. We will collect and analyze data about the use of each element of the iCPR tool during patient visits, including which elements of the tool were used and how often. We will also collect data from the site EHRs about antibiotic and diagnostic test orders for strep throat and pneumonia from all clinics participating in the trial, both Intervention and Control groups. After one year of study implementation, we will run an Interim Primary Outcome Report comparing the antibiotic and diagnostic test orders between the Intervention and Control group clinics. This report will be in the aggregate and will not contain any personally-identifiable information. If there is a significant difference between the groups that meets our predetermined stopping end points, we will stop the randomized controlled trial.

As the nation continues its efforts to contain healthcare costs and improve quality, healthcare information technology provides some of our most potent yet underutilized tools. Clinical prediction rules are frontline decision aids that combine state-of-the-art evidence with real-time patient history, physical examination, and laboratory data. While often well-validated, clinical prediction rules have been underutilized in practice. Recently, our team developed the integrated clinical prediction rule (iCPR) system, embedding CPRs within the nation's largest commercial electronic health record (EHR) system. Using this novel system, we demonstrated high rates of provider utilization and a significant reduction in antibiotic prescribing and diagnostic test ordering among suspected cases of strep throat and pneumonia at a single healthcare facility. The objective of the proposed project is to generalize this platform across diverse settings and create a toolkit for further dissemination. Building on the success of the original iCPR project, the specific aims of this proposal are to (1) integrate our previously tested and refined iCPR tool into the same commercial EHR in three different clinical settings, adapting the innovation to provider preference, culture, and local workflow rather than imposing a rigidly standardized tool, (2) identify and measure rate and variability of iCPR uptake across different settings, (3) determine iCPR impact on antibiotic prescribing and diagnostic test-ordering patterns across diverse clinical settings with a randomized controlled trial, and (4) use a well-established theory-driven implementation framework to identify facilitators and barriers to integration in each setting, and develop a toolkit for adapting and implementing the tool in diverse settings. To achieve these aims, we propose a five-year study in which we first adapt, integrate and usability-test the original iCPR at three new diverse sites. We will then conduct a two-year randomized controlled trial with a one-year post-trial open-access observation period to determine the persistence of: 1) the tool's utilization and 2) its impact on antibiotic- and test-ordering in patients with suspected strep throat or pneumonia. In the final year, study findings will be compiled into a toolkit so that any healthcare facility using the Epic EHR can integrate iCPR into its ambulatory workflow. The study uses several innovative and significant approaches, including: 1) adapting the nation's most widespread commercial EHR system; 2) building the new tool with "off-the-shelf" technology included in every Epic EHR package, so the innovation can be easily ported to all Epic EHR users; 3) using highly specific, well-validated clinical prediction rules as its core content; 4) guiding the integration process with highly generalizable usability testing techniques; and 5) using a hybrid RE-AIM and normalization process theory implementation evaluation framework. Together, these innovative approaches make iCPR uniquely suited to overcome longstanding barriers and integrate and disseminate evidence-based tools into the primary care workflow at the point of care in real time.

EMR integrated clinical prediction rule system guiding antibiotic prescription choices for strep and pneumonia

clinical decision support guiding clinician through clinical prediction rule and associated evidence based orders for strep and pneumonia

Inclusion Criteria: seen for strep or pneumonia visit at participating site Exclusion Criteria: none

Mann D, Hess R, McGinn T, Richardson S, Jones S, Palmisano J, Chokshi SK, Mishuris R, McCullagh L, Park L, Dinh-Le C, Smith P, Feldstein D. Impact of Clinical Decision Support on Antibiotic Prescribing for Acute Respiratory Infections: a Cluster Randomized Implementation Trial. J Gen Intern Med. 2020 Nov;35(Suppl 2):788-795. doi: 10.1007/s11606-020-06096-3. Epub 2020 Sep 1.

Mishuris RG, Palmisano J, McCullagh L, Hess R, Feldstein DA, Smith PD, McGinn T, Mann DM. Using normalisation process theory to understand workflow implications of decision support implementation across diverse primary care settings. BMJ Health Care Inform. 2019 Oct;26(1):e100088. doi: 10.1136/bmjhci-2019-100088.

Feldstein DA, Hess R, McGinn T, Mishuris RG, McCullagh L, Smith PD, Flynn M, Palmisano J, Doros G, Mann D. Design and implementation of electronic health record integrated clinical prediction rules (iCPR): a randomized trial in diverse primary care settings. Implement Sci. 2017 Mar 14;12(1):37. doi: 10.1186/s13012-017-0567-y.