Abstract
Dear Editor:
Palliative care is a relatively new field and was recognized as a subspecialty in 2006. Recently, it has seen substantial growth. Owing to the lack of uniformity in program structure compounded by the rapid growth of the field and shortage of manpower, wide variations in program structure exist.1,2 These variations can be challenging to overcome as it typically involves a shift from the status quo. The complexity of the variations can pose a challenge in determining the extent of their influence on outcomes. 3 As hospitals tend to consolidate into systems and systems into larger entities, there is increased focus on minimizing variation, improving the quality of care delivered and reducing cost. Traditionally, most programs have developed with varying clinical philosophies and operational structures.
Sentara Healthcare is a nonprofit organization that provides an array of services to communities in Virginia and North Carolina. This system, which is more than a century old, is a community-based health care system with 12 hospitals. In 2013, the system invested in a system-wide leadership position to assist with assimilating the practices of the various programs and ensure the delivery of consistent and high-quality palliative care throughout the hospitals. Most of the 12 hospitals had an inpatient palliative care consult program in urban locations, with some more remote than others. They were “home grown” in each location with different demands from the clinical services and hospital administrations along with varying staffing structures and philosophies. We started compiling data that were able to be extracted from the electronic medical record, Epic, and report to Center to Advance Palliative Care (CAPC) for national benchmarking. 4 In 2016, 7 of the 12 hospitals were on EPIC. We analyzed the data to study the utilization of palliative care services and variations between different facilities. We are sharing the data and our experience in this communication to hopefully assist other programs in similar stages of development.
A palliative care navigator was created on EPIC to assist with (1) data capture and (2) documentation.
The operational data with information that we feel relevant to the scope of this article are presented in Table 1.
Variations in Program Data within the Same System
ED, emergency department; FTE, full-time equivalent.
With the various models of delivery of palliative care, the consensus on the ideal model based on data and outcomes remains a subject of debate due to the young nature of our discipline. The information presented in the table is not unique to Sentara Healthcare and is commonly seen in programs within systems and between hospitals in the same and different regions. A point to make note of is that while the nuances of the variables are being deciphered based on the local and national data available, Sentara Healthcare has made a significant investment and improvements in staffing standardization as it stands in 2018. We have highlighted what we have experienced to be common data elements that are discussed within our field when comparisons are drawn and share food for thought if some variables are more influential than others.
Total number of consults is a great measure to assist with staffing. However, the volume does not reflect the intensity of involvement, as we will discuss later. The penetration rate is a measure of utilization of the services by the hospital. Since the volume of new consults includes patients that were on observation status, we include observation patients in the denominator when this measure was calculated. The full-time equivalent (FTE) per 100 consults is also a good measure when the team is looked at in total. It does not, however, convey the variations in disciplines within the team. Although these metrics are a point of reference, they do not account for the types of patients seen and daily caseload that the teams manage. Examples include how early are patients seen during the course of their disease, reason for consults such as symptom management versus discussions of goals (hospital 5 vs. hospital 7), and timing of consult within the hospitalization. Should a team that sees a mix of patients for symptom management versus a team that sees only prehospice patients have the same staffing ratio for seeing the same number of patients?
As a field the definition of “early palliative care ”is yet to be determined. What compounds this further are the differences in opinion of the time for a typical palliative care intervention such as a family meeting, the number of team members desired to be involved, and the need and depth of follow-up. Does the size of a hospital really matter? Does the fact that one of these hospitals is an academic hospital make a difference in workflow (hospital 5)? In addition, what are the ideal ratios of team members and do they vary when the aforementioned factors are taken into consideration? Higher utilization of the service from the emergency room (ER) (hospital 4) can certainly impact the workflow as typically the expectation is to attend to those consults right away. Depending on the nature of the situation, ED consults might demand more time to get the desired outcomes for the patients. In addition, does diagnosis (e.g., cancer vs. noncancer) matter? Clearly, when growth is planned, one metric is not enough to assist operationally with strategy. A mix of various factors such as some mentioned earlier need to be taken into account.
In sharing our experience, each team reports that the patients who are admitted to the hospital they work in are unique and different from others. Hence, to apply the same standard to different programs would seem futile. Interestingly some of the teams do not report the need for some disciplines, particularly physicians, to be part of the team. This is compounded by the fact that some hospitals are more rural than others and recruitment of fellowship trained or board-certified physicians has been a challenge. Does this compromise the quality of the program, and if the variation of not having a physician on the team is acceptable, then what is the threshold of necessitating one on the team? A counterpoint of view might be that if the model without a physician works in one location, why would it not work in others and perhaps be more cost-effective? Operationally, the staffing of teams can be a challenge when these factors are taken into account.
The most consistent consensus is for the program to cater to the needs of the local culture. With that said, it becomes a challenge to standardize programs within the same system. These will have to be tied in with outcomes and quality measures or the picture will be left incomplete. 3 The potential limitation of the national database is that it reflects what programs are doing and not the ideal state. We feel most programs are understaffed rather than overstaffed and the information in the registry needs to be interpreted with that in mind.
As we get more data available through the CAPC registry, we look forward to more clarity in answers to these questions. We are optimistic that over time the database will offer program-level data to better understand the variations and support their stability and growth. In the meantime, it is clear that these data elements are tied to each other and decisions made based on a single data element may be perilous.