Abstract
Dear Editor:
For the past few decades, advance care planning (ACP) conversations have evolved. Once focused on documents (i.e., advance directives or health care proxy completion), ACP now focuses on the process of supporting people to understand and share values, goals, and preferences regarding future medical care. 1 Implicit in a focus on process is the understanding that ACP is usually a series of conversations, often with different providers, and that people's goals and values change to reflect changing health. 2 However, with an increased focus on a longitudinal understanding of patients' changing goals and values comes an increased need for discoverable documentation.3,4
The electronic medical record (EMR) is a key component of effective ACP because it offers the potential for both centralized and context appropriate access to the most up-to-date ACP information. However, too much or disorganized data can cause cognitive overload for clinicians and impair our ability to use ACP to guide patient care. 5 Most interfaces require clicking through notes. Even if all ACP notes have consistent formatting, relevant data may be buried within other events or never seen at all, as clinicians fatigued from the search move on to other tasks. To help with the problem of too much data, some interfaces allow review of only the latest data. Although this approach decreases searching, it does not allow for easy review of historical data, which may be more relevant or of higher quality.
Ideally, EMR documentation would allow for two levels of ACP review. The first level provides basic information for time-sensitive decision making such as resuscitation preferences. This is a place where the emergency department clinician can quickly look to see the POLST or code status. 6 The second level provides nuanced information to help guide longitudinal decision making such as details about patients' hopes and worries for their health, information about their prognosis, and key goals and values. 7 To document this second level of information, we report on an EMR template that captures the dynamic process of ACP yet minimizes cognitive overload, enabling clinicians to find information.
Our template was developed in Epic Systems EMR (Epic, Verona, WI) using a flowsheet structure (Fig. 1), which has three important advantages. First, the flowsheet enables the segmented display, called a printgroup in EPIC, of ACP conversations (Fig. 2). Each column in Figure 2 represents a single conversation. Each row shows a segment of that conversation. The segmented display helps clinicians quickly scan for relevant details, provides a teaching scaffold for how to document (and conduct) effective conversations, and enables more efficient data collection for research or quality improvement purposes. Second, the flowsheet enables the longitudinal display of multiple conversations. This enables clinicians to more easily build on each other's work and highlights the patient's internal psychological process of adaptation and acceptance, which may not be otherwise visible to clinicians new to the care team. In addition, clinicians can use an EPIC smartphrase to place the documented ACP conversation into any clinically relevant EPIC documentation area (e.g., handoffs, notes, or assessments and plans). Because of this uniformity of documentation, using the search feature built into EPIC, clinicians can locate any historical documentation of ACP conversations. Finally, the flowsheet allows us to report data for the purposes of clinician feedback, to assess the efficacy of our interventions and to follow these patients for outcome measures.
Serious Illness Conversation flowsheet for documentation. Text boxes allow for free text while check boxes encourage efficient documentation and data collection.
Serious Illness Conversation printgroup allows for longitudinal review of conversations that are segmented into areas that allow clinicians to scan for relevant information, to appreciate the dynamic process involved in advance care planning, as well as providing a scaffold for teaching.
In July 2017, we activated this template in our health system. We report on its use throughout our system, which includes two 800 plus bed tertiary care academic hospitals, community hospitals and health centers, a postacute care network, several physician networks, a home health agency, and other health-related entities. Starting in 2015, our health system launched a major implementation of the Serious Illness Care Program, 8 and a key element of the program is documentation. We have trained 1800 interprofessional clinicians. Importantly, the template supports interprofessional collaboration. For example, in 2019 (1/1/19–6/30/19), at one site (Massachusetts General Hospital) 1890 conversations were documented; 47% by nonphysician interprofessional clinicians. At Brigham and Women's Hospital, where the program was founded and piloted in 2014 9 among all high-risk patient who died, 70% have had conversations; the program has been collecting and reporting metrics for six years, demonstrating sustainability.
From our experience, we note that the flowsheet and printgroup structure has two limitations. First, although conversations can be reviewed longitudinally, the number of conversations displayed by the printgroup is functionally limited by the display area. At our institution, up to five conversations are displayed. Second, the printgroup will only display the last conversation had during an encounter. So if several important conversations were documented during a long hospitalization, the printgroup would only display the final one. It is possible to review all conversations by using the EPIC search function. However, this reveals all instances of documentation and requires clicking through multiple displays to locate the conversation of interest. It would be ideal to be able to search for a particular conversation such as the one documented by a patient's specialist or before a particular event. Structural improvements require redesign of the flowsheet architecture or the design of a new more flexible form. Such improvements cannot be made locally and instead require collaboration with EPIC leadership at a national level.
We believe that the structure of our ACP documentation template promotes having, sharing, and documenting serious illness conversations. This synergy happens because of the longitudinal segmented nature of the documentation template. First, this structure makes it easy for clinicians to access and share useful information. Clinicians then document because the impact on patient care is worthwhile when weighed against the documentation burden. Second, the documentation structure reminds clinicians of the key elements of successful serious illness communication, 8 thereby reinforcing skills learned in training and giving confidence to clinicians to again engage in these conversations. Third, the structure allows for shared responsibility of ACP between clinicians, which is key in an environment where multiple clinicians (primary care providers [PCPs], hospitalists, and nurse care managers) care for patients in a variety of settings. If the flowsheet limitations were addressed by EPIC, this could be an even more powerful tool for documenting and reviewing ACP. Future study is needed to confirm that high levels of clinician documentation can be sustained and become the standard of care.