Successful Implementation of Workflow-Embedded Clinical Pathways During the COVID 19 Pandemic

Background and Objectives:

Clinical pathways have been found effective for improving adherence to evidence-based guidelines, thus providing better patient outcomes. As coronavirus disease-2019 (COVID-19) clinical guidance changed rapidly and evolved, a large hospital system in Colorado established clinical pathways within the electronic health record to guide clinical practice and provide the most up-to-date information to frontline providers.

Methods:

On March 12, 2020, a system-wide multidisciplinary committee of specialists in emergency medicine, hospital medicine, surgery, intensive care, infectious disease, pharmacy, care management, virtual health, informatics, and primary care was recruited to develop clinical guidelines for COVID-19 patient care based on the limited available evidence and consensus. These guidelines were organized into novel noninterruptive digitally embedded pathways in the electronic health record (Epic Systems, Verona, Wisconsin) and made available to nurses and providers at all sites of care. Pathway utilization data were analyzed from March 14 to December 31, 2020. Retrospective pathway utilization was stratified by each care setting and compared with Colorado hospitalization rates. This project was designated as a quality improvement initiative.

Results:

Nine unique pathways were developed, including emergency medicine, ambulatory, inpatient, and surgical care guidelines. Pathway data were analyzed from March 14 to December 31, 2020, and showed that COVID-19 clinical pathways were used 21 099 times. Eighty-one percent of pathway utilization occurred in the emergency department setting, and 92.4% applied embedded testing recommendations. A total of 3474 distinct providers employed these pathways for patient care.

Conclusions:

Noninterruptive digitally embedded clinical care pathways were broadly utilized during the early part of the COVID-19 pandemic in Colorado and influenced care across many care settings. This clinical guidance was most highly utilized in the emergency department setting. This shows an opportunity to leverage noninterruptive technology at the point of care to guide clinical decision-making and practice.

Clinical pathways have been found effective for improving adherence to evidence-based guidelines and, therefore, providing better patient outcomes.1,2 Clinical care pathways are defined as clinical decision-making tools at the point of delivery of health care with the goal of providing evidence-based patient care to all.3–5 In May 2001, the Centers for Medicare & Medicaid Services placed an emphasis on evidence-based performance with the development of the first core measures.6 The use of clinical pathways has grown, as hospitals and providers focus on providing care that is high-quality, timely, and reliable.7–9

Currently, clinical pathways are applied in many health systems across the country and the world, with care recommendations founded on research-based literature.5,10,11 However, low adoption of and poor adherence to clinical pathways have been notable issues. The difficulties arise from trying to implement this type of clinical decision support into complex behavioral systems as well as trying to manage local contextual issues.12,13 Additionally, clinical pathways are not always embedded into clinical workflows, translated into local hospital governance, or specifically available at the level of the patient encounter in the electronic medical record platform.2,14

The coronavirus disease-2019 (COVID-19) pandemic response introduced a new framework for clinical pathways, as validated evidence-based literature was not available. A single COVID-19 clinical care pathway for emergency department (ED) care was developed in New York City, illustrating that clinical pathways could be useful in situations in which demand for care exceeds capacity.15 Additionally, this care pathway assisted providers in optimizing resource allocation within the institution.15

Globally, clinical pathways were utilized in similar contexts. The Italian Hospitals and Local Health Authorities (ASL) implemented clinical pathways in multiple care settings during the first 2 waves of the COVID-19 pandemic to assist in how care was delivered.16 Additionally, clinical pathways were utilized in the UK hospital system for decision support associated with the diagnosis and management of COVID-19 patients.17 This assisted in resource prioritization, specifically, rapid testing and single-occupancy rooms.17

Our project centers on a large, 12-hospital health system in Colorado. As COVID-19 clinical guidance changed and evolved rapidly, this hospital system established clinical pathways within the electronic health record (EHR) to guide clinical practice, provide the most up-to-date information to frontline providers, and optimize resource allocations across the system. This clinical pathway system delivered standardized COVID-19 diagnosis and treatment recommendations via workflow-embedded clinical algorithms to include all aspects of care, including emergency, ambulatory, inpatient, and surgical care. To evaluate provider utilization, we retrospectively examined pathway utilization over time and evaluated adoption rates compared with hospitalization rates.

METHODS

Our large, 12-hospital health system cared for more than 25% of Colorado's COVID-19-associated hospitalized patients. This hospital system manages patients in a variety of care settings, including a large quaternary care academic hospital, multiple community hospitals ranging from large to small, as well as multiple outpatient ambulatory care spaces, and a critical access hospital. This project was designated as a quality improvement initiative, and, therefore, exempt from review by the institutional review board.

On March 12, 2020, a system-wide multidisciplinary committee of specialists in emergency medicine, hospital medicine, surgery, intensive care, infectious disease, pharmacy, care management, virtual health, informatics, and primary care was recruited to develop clinical guidelines for COVID-19 patient care based on the limited available evidence and consensus. These pathways provided comprehensive care recommendations including, but not limited to, ambulatory symptom testing criteria, ED care and disposition, and in-hospital treatment and discharge criteria. These recommendations included guidance on medication administration, such as anticoagulation recommendations and supplemental oxygen delivery, mechanical ventilation settings, transitions of care, and end-of-life treatment incorporating compassionate extubation (Figure 1). They were organized into novel noninterruptive digitally embedded pathways in the EHR (Epic Systems, Madison, Wisconsin), using a software platform (AgileMD, San Francisco, California). This noninterruptive platform did not interrupt or pause providers' workflow but, rather, was available within every patient's chart to be accessed when convenient.

Figure 1.

Example of COVID-19 initial ED evaluation and management pathway, revised and published on March 31, 2020. ED indicates emergency department.

To access pathways within the EHR, providers located the “Pathways” tab in the patient's chart. The COVID-19 pathways were listed under the “Top Suggested” list on the homepage, so providers could quickly find relevant content. Pathways were characterized by easy-to-follow flowchart diagrams and were keyword searchable in the EHR (Figure 1). The recommended orders were hyperlinked and could be ordered directly off the pathways. Additional facility-specific resources were available via the pathway and embedded references were also included (Figure 1). These pathways were made available to nurses, providers, and ancillary teams at all sites of care across the system (eg, ambulatory clinics, virtual health, and all acute care settings) 2 days after the initial committee meeting. The medical staff was notified of these resources by email from each hospital's chief medical officer.

These pathways were rapidly updated, sometimes daily, based on evolving evidence. To ensure providers and staff were aware of when edits and optimizations were made to the pathways, the “date of last update” was included in the pathway subtitle. The evidence utilized was included as hyperlink references and the provider or service approval was included at the top of the pathway. Email communication was sent to provider groups through the system-wide COVID-19 Pathway Committee, which was a multidisciplinary committee of specialists. Feedback from the frontline staff was received through a feedback submission link embedded in each pathway, which allowed all staff to send immediate feedback and suggestions for optimizations to the committee. Every COVID-19 pathway went through at least one iteration based on provider feedback, changes made to treatment recommendations and/or availability of resources. Lastly, each facility was able to customize based on its needs from a workflow perspective.

Provider utilization data were collected from March 14, 2020, to December 31, 2020, from all care locations as the pathways went live. The start date selected was the first date COVID-19 pathways were available to frontline providers in the system. Of note, the first confirmed case of COVID-19 in Colorado was on March 5, 2020. Pathway utilization was enumerated by collecting each time a provider or staff entered a unique pathway with a unique Epic session and a unique patient encounter. These uses could be on the same day, as long as one of the listed components was different. Pathway utilization data were collected for all provider types (eg, attending, resident, fellow, registered nurse, and respiratory therapist), although most clinical pathways were focused on the provider (physician or advanced practice provider) decision-making.

RESULTS

Nine unique pathways were developed for emergency medicine, ambulatory, inpatient, and surgical care. Because of differing facility needs and available resources (eg, pharmacotherapy, care levels, and testing availability), the original 9 unique pathways developed were customized for each facility in the health system. Therefore, a total of 69 facility-specific pathways were created throughout the health system.

The COVID-19 clinical pathways were opened and viewed 21 099 times from March 14 to December 31, 2020. The majority of pathway utilization, 16 991 (81%), occurred in the ED setting across the health system (Table 1), and 19 485 (92.4%) viewed the up-to-date embedded testing recommendations. When comparing facility-specific utilization, pathway utilization was closely correlated to the size of facility (Table 2). Of note, this facility analysis did not include all clinical pathway utilization data, as there were shared service-line pathways introduced that were not facility specific.

Table 1. COVID-19-Specific Pathway Utilization Stratified by Care Space, Between March 14 and December 31, 2020 (N = 21 099).

Care Space	COVID-19-Specific Pathway Utilization (Percentage)
Emergency department	16 991 (81%)
Inpatient care	3 226 (15%)
Surgical care	373 (2%)
Ambulatory	509 (2%)
Total	21 099 (100%)

Table 2. COVID-19 Specific Pathway Utilization Stratified by Facility Between March 14 and December 31, 2020 (N = 17 322).

Facility	COVID-19-Specific Pathway Utilization
University of Colorado Hospital	13 941
Broomfield Hospital	940
Grandview Hospital	34
Highlands Ranch Hospital	479
Greely Hospital	165
Longs Peak Hospital	389
Medical Center of the Rockies	502
Memorial Hospital Central	298
Memorial Hospital North	103
Poudre Valley Hospital	422
Yampa Valley Medical Center	49
Total	17 322

A total of 3474 distinct providers employed these pathways for patient care across the health system. As compared with COVID-19 hospitalizations in Colorado, pathway utilization paralleled the trend over time (Figure 2) with peaks in both rates during the same time.

Figure 2.

Daily pathway utilization and hospitalizations attributed to COVID-19 in Colorado from March 14 to December 31, 2020.

DISCUSSION

Although operationalizing evidence-based guidelines has been shown to be challenging, clinical pathways have been shown to improve patient outcomes by decreasing unnecessary variation in provider care.10,18 Clinical pathways reduce in-hospital complications and readmission rates as well as result in shorter hospital length of stays and a reduction in overall health care costs.5,19,20 As hospital systems across the country became overwhelmed by the sheer number of resources needed due to the COVID-19 pandemic, our clinical pathways provided standardized care and resources across many care settings. These novel noninterruptive digitally embedded pathways in the EHR, using a third-party vendor software platform, were unique as providers could access the most up-to-date evidence-based guidance, apply easy-to-follow flowchart diagrams pathway to place orders, and provide patient care.

It appears that providers quickly and broadly viewed these noninterruptive digitally embedded clinical care pathways with no formal training. Additionally, because formal training is not needed, clinical pathways in the form used here can be easily disseminated to additional providers and care settings.

During the study period of March 14 to December 31, 2020, national data regarding both diagnostic evaluation and treatment were rapidly evolving. Initially, in early March 2020, national guidance essentially did not exist.21 Then through the summer of 2020, recommendations from federal and state agencies began to rapidly grow based on limited data (case reports) and consensus.22,23 By the winter of 2020, most states had experienced what was to become their first peak in the number of COVID cases.24 As COVID-19 care rapidly changed and more information was gathered and analyzed, these local pathways gave providers the most up-to-date information from a system-wide multidisciplinary committee of specialists. Pathway utilization mirrored COVID-19 hospitalization data, which could reflect the perceived benefits of the use of the pathways for this patient population.

Our embedded clinical pathways were most highly utilized in the ED setting and were specific to testing recommendations. This is likely since the ED was the first place patients with COVID-19 symptoms sought care. Additionally, many of the ED care settings in the system routinely used clinical pathways prior to the pandemic for other complaints; therefore, providers were already familiar and comfortable with this form of clinical guidance. Lastly, the ED care setting had rapidly changing evidence, specifically, testing, management, and discharge criteria, early in the pandemic. Therefore, the high pathway utilization suggests changes in clinical pathways in the context of rapidly emerging evidence positively influence pathway utilization. Conversely, lower utilization in other service lines could be explained by relatively lower rates of change in those clinical pathways (eg, ambulatory and surgery) and no previous exposure to clinical pathways in routine care.

There were several limitations to this project. We are unable to determine how many patients would qualify for each pathway; therefore, determining a proportion of patients who were eligible for the pathway use was difficult. Additionally, we did not collect data on outcomes of patients who had the clinical pathways applied to their care. Future work will need to focus on the effect of pathway usage on clinical outcomes as well as examining the barriers to clinical pathway utilization. Data were not collected on how many specific orders were placed through each pathway using the embedded order linking functionality; therefore, stratification by provider type is another area of ongoing work. Although end users have the ability to provide feedback on all pathways, we did not collect clinician user perspective on the pathway in a formal way. There is ongoing research focusing on the end users to determine whether additional steps should be taken to create and sustain a culture of acceptance and usage of clinical pathways.

The findings of high adoption illustrate that it is possible to rapidly implement novel pathways and frequently update them to support provider clinical decision-making to drive standardized care in a pandemic setting. The high pathway utilization rates represent the success of leveraging noninterruptive technology at the point of care to provide clinical decision support and guide provider practice.

CONCLUSIONS

Noninterruptive, digitally embedded clinical care pathways were quickly implemented and broadly used during the early part of the COVID-19 pandemic, influencing practice across many care settings. This clinical guidance was most highly utilized in the ED setting, likely due to the nature of the point of entry of patients to the health care setting during this unprecedented time. There is an opportunity to leverage noninterruptive technology at the point of care to guide clinical decision-making and practice, so providers can continue to foster a culture of evidence-based clinical care even in the setting of a novel disease outbreak.