ePneumonia: Development of an Electronic Clinical Decision Support System for Community-Onset Pneumonia

Since the launch of a paper-based pneumonia care process model in 1994, decision support for pneumonia care has been under continuous development at Intermountain. Studies published in 2001 and 2006 demonstrated decreased mortality using paper-based methods. An electronic pneumonia Clinical Decision Support System was later developed in the original Intermountain computing environment and implemented in 4 regional emergency departments (ED) in 2011. This tool featured a novel mortality predictor and real-time synthesis of clinical data to guide diagnosis, risk stratification, admission triage and guideline-concordant treatment. An outcome study published in 2015 demonstrated reduction in mortality with tool use compared to usual care. Most recently, Intermountain researchers led by study co-Investigator, Dr. Brandon Webb, developed an innovative tool to predict risk of drug-resistant bacteria and demonstrated its potential to improve antibiotic use and outcomes.

The investigators have entered a robust phase of additional development and adaptation of ePneumonia into the Cerner Electronic Health Record (EHR) system. The objective of this study is to advance development of an evidence-based, electronic CDSS for pneumonia care with interoperability across EHRs in order to improve clinical outcomes and reduce healthcare resource utilization. The specific aim of this study is to evaluate the usability of ePneumonia and its associated impact on clinical, patient-centered and healthcare resource utilization outcomes in a stepped-wedge implementation study in 16 hospital EDs in the Intermountain Healthcare integrated health system.

• Hypothesis #1: Healthcare providers will affirm ePneumonia usability, lack of interference with clinical workflow and only minor unintended consequences of use.
• Hypothesis #2: In patients with community-onset pneumonia, ePneumonia use will improve clinical and patient-centered outcomes and decrease healthcare resource utilization.

One year of baseline clinical outcome data will be gathered for all 16 emergency departments. The first of 6 clusters of ED's will begin prospective data collection in January 2018, with the remaining coming on at 2 month intervals until ePneumonia has been deployed at all sites. An additional 1 year of data collection will be continued through 2019.

Inclusion criteria

• All patients ≥ 18 years who are identified by either (2a) ICD-10 codes for pneumonia; or acute respiratory failure or sepsis with secondary pneumonia codes or (2b) clinician completion of ePneumonia for Cerner.
• Intermountain Healthcare physicians working in the 16 ED's

Exclusion criteria

• Patients without radiographic confirmation of pneumonia
• subsequent episodes of pneumonia within the study period, so as not to over-represent patients with recurrent pneumonia caused by recurrent aspiration or structural lung disease, and
• immunosuppressed patients, such as those with AIDS.