The Covid-19 pandemic compelled hospitals worldwide to adopt innovative measures aimed at reducing the administrative burden on healthcare staff already battling burnout, according to Teray Johnson, Director of Data Automation and Transformation at Lifepoint Health. Applications ranged from hospital-at-home to within-EHR automation of clinical decision-making processes.
One of the notable innovations was a clinical decision-making AI algorithm crafted for the Cerner EHR at LifeBridge Health, where Johnson previously worked. This system used clinical criteria to identify potential palliative care candidates. The algorithm would generate a popup alert, which, if the patient’s healthcare provider agreed that the patient met eligibility criteria, would automatically result in the creation of a palliative care consult order.
To develop the algorithm, the palliative care team collaborated with data analysts and account managers at Cerner. The algorithm was customized to meet the unique needs of Baltimore and Carroll County’s patient communities. Through cross-team efforts from the data analysts, database administrators, and clinical informaticists at LifeBridge, metrics were designed to track the algorithm’s efficiency. The resultant system saw a sharp rise in identified palliative care patients while relieving administrative stress related to consult orders.
Existing conditions at Lifepoint Health are characterized by the intersection of operations and technology, and multidisciplinary teams are engaged in the creation of algorithms that fuel clinical decision-making. When a need is identified, Johnson asserts, the teams travel onsite to comprehend the prevailing conditions. This process involves understanding clinical workflows and collating suggestions for the best technology solutions.
One of Lifepoint’s successful projects concerned the streamlining of authorization processes in primary care practices. Through the identification of pain points, research into best practices, and the introduction of an automated solution offered by a vendor, the approval process was substantially simplified. This system’s efficiency was confirmed via several pilot tests, with feedback from physicians and administrative staff providing the necessary fine-tuning.
A common feature of algorithm creation and automation projects, Johnson argues, is sustained engagement from all team members. This ensures that adjustments are made to cater to changing patient and clinician needs. Constructive interpersonal communication features prominently in Johnson’s account, as a crucial tool for dissecting the scope, responsibilities, metrics, and potential impacts of projects. Early adopters play a significant part in confirming the feasibility of concepts, and the involvement of physicians early in the decision-making process can drive successful implementation.
In conclusion, Johnson asserts that multidisciplinary teams are critical when developing AI algorithms and technologies for automating clinical decision-making. Such an inclusive model of innovation enables clinicians to return to their essential role: patient care.
