Barriers to and facilitators of successful implementation of a palliative approach to care in primary care practices: a mixed methods study

Abstract

Objective

Integrating a palliative approach to care into primary care is an emerging evidence-based practice. Despite the evidence, this type of care has not been widely adopted into primary care settings. The objective of this study was to examine the barriers to and facilitators of successful implementation of a palliative approach to care in primary care practices by applying an implementation science framework.

Design

This convergent mixed methods study analysed semistructured interviews and expression of interest forms to evaluate the implementation of a protocol, linked to implementation strategies, for a palliative approach to care called Early Palliation through Integrated Care (EPIC) in three primary care practices. This study assessed barriers to and facilitators of implementation of EPIC and was guided by the Consolidated Framework for Implementation Research (CFIR). A framework analysis approach was used during the study to determine the applicability of CFIR constructs and domains.

Setting

Primary care practices in Canada. Interviews were conducted between September 2020 and November 2021.

Participants

10 individuals were interviewed, who were involved in implementing EPIC. Three individuals from each practice were reinterviewed to clarify emerging themes.

Results

Overall, there were implementation barriers at multiple levels that caused some practices to struggle. However, barriers were mitigated when practices had the following facilitators: (1) a high level of intra-practice collaboration, (2) established practices with organisational structures that enhanced communications, (3) effective leveraging of EPIC project supports to transition care, (4) perceptions that EPIC was an opportunity to make a long-term change in their approach to care as opposed to a limited term project and (5) strong practice champions.

Conclusions

Future implementation work should consider assessing facilitators identified in our results to better gauge primary care pre-implementation readiness. In addition, providing primary care practices with support to help offset the additional work of implementing innovations and networking opportunities where they can share strategies may improve implementation success.

Strengths and limitations of this study.

• Use of the Consolidated Framework for Implementation Research implementation framework to guide the development of interview guides and analysis.

• Triangulation and synthesis of expression of interest (EOI) and interview data to compare facilitators of and barriers to implementation between primary care practices.

• EOI forms submitted by practices to be part of the Early Palliation through Integrated Care quality improvement project may have been positively biased.

• The study was conducted in three primary care practices in one Canadian province.

Introduction

Despite the push for evidence-informed healthcare practice, the transfer of new knowledge into practice continues to be a struggle.¹ It is essential to know whether an intervention can effectively improve outcomes before deciding whether to implement it in practice; it is also important to understand how to best facilitate successful implementation in real-world settings. A programme’s effectiveness does not mean it is feasible to implement in practice.² Implementation research studies focus on understanding what leads to successful implementation at not only the patient level but also at the provider, organisation and policy levels of healthcare. Implementation research projects apply knowledge from established theories and frameworks to identify what facilitates or hinders successful implementation of an innovation and what strategies can be used to increase its adoption in a practice setting.^{3 4} The innovative practice for this implementation research project was implementing a palliative approach to care in primary care practices.

A palliative approach to care has been defined as having primary care teams with the ‘skills, resources and processes in place to recognise, assess and manage basic palliative care needs in a timely fashion in a community setting’.⁵ Evidence has shown that a palliative approach to care improves patient quality of life and symptom management, decreases depression and may increase life expectancy.^6–10 This evidence has led to the creation of indicators and guidelines^{11 12} to facilitate implementation of a palliative approach to care rooted in primary care. Primary care providers (PCPs) provide first-contact services for patients and can include family physicians, physician assistants, nurses and nurse practitioners.¹³ PCPs are positioned well within healthcare systems to develop ongoing trusting relationships with patients while co-ordinating and providing personalised end-of-life care.¹⁴ An essential element to high-quality palliative care is the early identification of appropriate patients.¹⁵ Given primary care often has an ongoing relationship with patients, PCPs are well situated to identify patients early in their palliative trajectory. The role of a palliative approach to care in primary care is crucial to keep patients at home as long as possible.¹⁶ It is commonly understood that staying at home has a high value for individuals and lowers costs for the healthcare system, thus PCPs are starting to have a more significant role in early palliative care.¹⁷ It is clear that a palliative approach to care is an important element of primary care, however implementation barriers can limit adoption. Thus, it is important to understand what factors facilitate or hinder implementation of this type of care.

The objective of this study was to examine the barriers to and facilitators of successful implementation of a palliative approach to care in primary care practices by applying an implementation science framework.¹⁸

The study addressed the following two research questions:

1. How do multilevel factors facilitate, or hinder, successful implementation of a palliative approach to care in three primary care practice settings?

2. How can learnings from this study facilitate implementation of innovations in other primary care settings?

Methods

The innovation: Early Palliation through Integrated Care (EPIC)

Our project worked in collaboration with a quality improvement project led by the health authority that hoped to spread a protocol and associated implementation strategies they developed to support earlier access to palliative care in primary care settings. The protocol, called EPIC, embodied processes and guidelines previously established for a palliative approach to care in other countries.^{19 20} EPIC provided primary care teams with resources and tools to aid in successfully adopting a palliative approach to care. Box 1 lists what was provided to practices involved in the EPIC quality improvement project. The EPIC project leads had piloted and refined the protocol in one primary care practice. The findings indicated there was a successful shift in practice towards implementing the EPIC protocol and positive patient satisfaction with the care received using the protocol.²¹

Box 1. Early Palliation through Integrated Care for primary care providers.

• Tools for practices to use (eg, symptom assessment, functional assessment and goals of care).

• Improved access to information on how to improve the experience of care for patients who could benefit from an early palliative approach to care.

• Training in a palliative approach to care (Pallium Canada’s Learning Essential Approaches to Palliative Care Core training curriculum, https://www.pallium.ca/course/leap-core/).

• Identified champion at each individual primary care practice.

• Dedicated implementation facilitator embedded within the health authority to work with individual primary care teams.

The EPIC protocol had PCPs use and implement multiple tools to help identify individuals who were appropriate for early palliative care. These tools included the Edmonton Symptom Assessment System (ESAS),²² Palliative Performance Scale (PPS)²³ and the Supportive and Palliative Care Indicators Tool (SPICT).²⁴ EPIC emphasised that providers needed to document tool results so patients could be monitored over time and results could be used to inform decisions on changes to care.

The EPIC protocol also facilitated practices’ access to health system information on ways they could improve patient experiences of early palliative care. This included resources to share with early palliative care patients such as information on accessing streamlined emergency health services programmes for patients considered to be palliative.

Learning Essential Approaches to Palliative (LEAP) Care Core training²⁵ was provided to ensure all providers involved in the EPIC project were adequately prepared and understood the rationale behind early identification and initiation of palliative care. LEAP training aims to teach providers the importance of self-awareness when providing palliative and end-of-life care, how to identify patients who could benefit from a palliative approach to care early in their disease trajectory, how to assess and manage pain and how to identify physical symptoms of delirium, gastrointestinal symptoms, hydration, nutrition and respiratory issues. LEAP training included providers from all disciplines: physicians, nursing, pharmacy and others. Providers involved with EPIC were required to complete LEAP training in-person (an online option was not available at the time) at the beginning of the project. If they had previously completed LEAP training, they were exempt from doing the training a second time.

Finally, as part of EPIC implementation strategies were added to promote successful implementation of EPIC. These implementation strategies included the use of champions and an implementation facilitator. Each practice had to identify a practice-based champion. In addition, a project co-ordinator also functioned as an implementation facilitator. The implementation facilitator had clear roles and strong connections to EPIC project leads who were embedded in the health authority. The role included continual auditing and feedback done in conjunction with the practice champions.

Theoretical framework

Identification of multilevel factors that could affect successful implementation data collection and analysis was guided by the Consolidated Framework for Implementation Research (CFIR). The CFIR is composed of evidence-based constructs designed to explain implementation outcomes.²⁶ CFIR V.1.0 has 39 constructs organised into five major domains that have been found to influence the successful implementation of innovative programmes. The domains assess the (1) Intervention/innovation characteristics (8 constructs), (2) Outer setting (4 constructs), (3) Inner setting (14 constructs), (4) Characteristics of individuals (5 constructs) and (5) Process (8 constructs).²⁶ The CFIR domains for this project were defined as: EPIC innovation (Intervention), health authority and community context (Outer setting), primary care practices (Inner setting), individuals involved in implementing EPIC (Characteristics of individuals) and process of implementing EPIC (Process).

Setting

In Canada, primary care is predominantly funded through public funds that are allocated by the health authorities within each province. The provincial health authority where this project was conducted encompasses urban, suburban and rural service locations. The practices in this study were in either a suburban or rural location. Prior to EPIC implementation, the health authority had established an integrated palliative care strategy, had identified an increased readiness from primary care practices to implement a palliative approach to care and supported LEAP palliative care education within primary care practices. The EPIC Study was planned and initiated prior to the COVID-19 pandemic, but primarily conducted during the pandemic.

Primary care collaborative practice teams that participated in this study were comprised of health providers with different professional backgrounds (ie, physician, nurse practitioners and nurses). Physicians could be renumerated through various payment plans; however, all nurse practitioners and family practice nurses were employed by the health authority.

Study participants and practices

For this study, purposive sampling was used to ensure that all participants were involved in implementing EPIC. Participants included policy-makers, EPIC quality improvement personnel and PCPs (nurses, nurse practitioners and general physicians) from the primary care practices. Participants were given time to review consent forms and ask the interviewer any questions prior to providing verbal, informed consent to participate in the study before each interview was conducted. As part of the verbal, informed consent, participants agreed to audio recordings before each interview.

Data collection procedures

Expression of interest (EOI)

To be part of the EPIC quality improvement project, collaborative primary care practices submitted an EOI questionnaire. The EOI was developed by the EPIC quality improvement team prior to the implementation study. The EOI captured information on primary care practice infrastructure, readiness to implement the EPIC project, how the practice currently cared for palliative patients, if a palliative approach to care was needed in their region and willingness to work collaboratively. The EOI was disseminated to family practices across the provincial health authority; six practices submitted an EOI and three practices were chosen by the EPIC quality improvement team to participate in the project. The EOI information was used by the quality improvement team to choose primary care practices they felt had the highest readiness to participate. For this study, the EOI was analysed in conjunction with interview data to provide insights into practice-level contextual factors that could affect successful implementation of EPIC. EOI questions are listed in box 2. EOI questions 3, 5, 7, 8 and 19 were not used in our analysis because either (1) they did not provide information pertinent to understanding successful implementation or (2) practices all recorded the same response, therefore was no differentiation between practices (eg, all practices could access patient records).

Box 2. Expression of Interest (EOI) questions asked of practices to be part of the EPIC project.

EOI questions

1. Team composition: Please identify the full primary care team, including names and disciplines. Response options: Open-ended question.

2. Champion contact: Please identify your practice champion and their email address/telephone number. Response options: Open-ended question.

3. Implementation awareness: Please confirm that you have read the introductory page and that all members of your team are committed to participate in this initiative. Response options: yes/no.*

4. What is the total practice size? Please identify the total number of patients served by the team. Response options: Open-ended question.

5. Flexibility: Palliative patients sometimes require attention outside of regular business hours. Would this be feasible for providers at your practice? Responses were yes/no.*

6. How are physicians in the practice renumerated? Response options: different types of payment systems.

7. Charts: Which system does the practice use to manage patient records? Response options: different local types of electronic medical records.*

8. Charts: Can patient charts be accessed by all team members? Response options: yes/no.*

9. Culture of improvement: Please describe your openness to identifying areas of improvement and making changes. Response options: not at all open, somewhat open, moderately open, or very open.

10. Teamwork: Which term best describes the working relationships within your practice. Response options: independent, colocated, co-operative or collaborative.

11. Palliative care role: Please describe your team’s current role in managing palliative and end of life care for your patients. Response options: Open-ended question.

12. Quality indicators: How do you measure success of the palliative care you are providing? Response options: Open-ended question.

13. Palliative care is an approach that is beneficial for those with advanced life limiting conditions. Response options: Scale from 1-10 on how much they agree.

14. Our team is committed to implement a new approach to palliative care that is rooted in primary healthcare. Response options: Scale from 1 to 10 on how much they agree.

15. Patient engagement in decision making and planning is an important aspect of patient care. Response options: Scale from 1-10 on how much they agree.

16. There is a higher level of planning and co-ordination required for patients with multimorbidities and life limiting conditions. Response options: Scale from 1-10 on how much they agree.

17. There is willingness to work collaboratively with a palliative care team and other community providers to jointly meet patient needs. Response options: Scale from 1-10 on how much they agree.

18. Are there any challenges limiting your team's participation in proceeding with an approach to palliative care rooted in primary care? Additionally, is there more information you need in order to fully commit? Response options: Open-ended question.

19. Is there any further information you would like to share, or other comments you would like to make about your interest in this project? Response options: Open-ended question.*

Interviews

For the implementation study, a semistructured interview guide based on the CFIR was created. The initial semistructured interview guide is provided in online supplemental file 1. Interviews were conducted between September 2020 and November 2021. Policy-makers and EPIC quality improvement personnel were interviewed first, followed by PCPs. Interviews were conducted during work hours. Physicians who worked in a fee-for-service renumeration model were given monetary compensation for their time to complete the interview. This was done to reimburse them for the time not spent caring for patients. Other providers were not provided with compensation as they were salaried. Interview questions probed participant narratives to understand multilevel factors that manifested during implementation, explore factors that facilitated or hindered implementation and what was needed to successfully spread the initiative across the province. Interview questions covered perceptions of the programme, organisational and health system impacts and details of the planning and process of implementation. Open discussion was encouraged to identify and explore evolving themes. A single person (JL) conducted all interviews over Zoom or telephone; this was necessary as the project was conducted during the COVID-19 pandemic. All interviews were audio recorded and transcribed verbatim by an external transcriber. After preliminary interviews were completed and data were coded and analysed, three additional repeat interviews were conducted (one representative from each practice) to probe the legitimacy of potential themes.

Analysis and interpretation

Interview transcripts were uploaded into NVivo for analysis. A framework analysis approach²⁷ was used during the study to determine the applicability of CFIR constructs and domains. Framework analysis is an iterative process that involves both deductive and inductive coding. The analytic process began once the initial interviews were conducted to help determine when new information was needed, or no longer being generated, from participants. The EOI and interview data were analysed concurrently and initially coded deductively using CFIR, followed by a more inductive approach to coding. CFIR coding was helpful in identifying, categorising and interpreting barriers and facilitators to achieve the outcome of successful implementation. See online supplemental file 2 for CFIR determinants. The codes and themes were reviewed by analysts (JL and GW) multiple times to check for potential biases, ensure they reflected participants’ words and improve the credibility of reviewers’ interpretation. These reflective processes are associated with trustworthiness.²⁸

Additional interviews were added when new themes emerged. This was done to ensure the resulting themes fully represented participant-identified concepts. These interviews helped confirm that enough information power²⁹ was reached with our sample size. Study parameters used to confirm information power were: keeping the focus of our study clear, sampling individuals with the experiences and knowledge to answer our research questions, applying an established implementation science theory, eliciting a strong dialogue during interviews by an experienced interviewer and examining one case rather conducting a cross-case analysis. Member checking involved secondary interviews with key participants and consultations with the broader research team (RU, CT and EC), these methods provided additional perspectives on whether the final themes were sufficiently supported. A convergent mixed methods analysis of EOI responses and interview coding was conducted to develop the final themes.³⁰ Triangulating the two data sources (qualitative interviews and EOI responses) allowed analysts to interrogate the credibility of the final themes. The use of concept mapping was used in both qualitative interviews and EOI responses to integrate the data.

Patient and public involvement

There was no patient or public involvement in the design or conduct of this study. However, knowledge users and healthcare knowledge users who were key to the implementation of the EPIC quality improvement project were involved in the design of the evaluation.

Results

In total, 14 individuals were invited to take part in interviews, with 71% agreeing to participate (n=10/14). The four PCPs who did not agree to participate did not answer the email invitation (n=3) or initially expressed interest but declined later due to time constraints (n=1). In total, 3 of the 10 interviews were conducted with policy-makers and research personnel involved in implementing the EPIC project; the remaining 7 were with PCPs (physicians, nurse practitioners or nurses). Three individuals from each practice underwent a second interview to discuss emerging themes. Most providers did the LEAP training as part of the EPIC project; two did it prior to EPIC. The provider in Practice A completed the LEAP training because they saw a need for it in their practice; conversely, the provider in Practice B took LEAP training before they started practising during their residency training.

The interview data was initially used to define successful implementation of EPIC. Then, an analysis of the EOI responses was done to examine initial differences across practices (table 1). Finally, a mixed methods analysis of the EOI responses and interviews was conducted to identify facilitators of and barriers to implement EPIC. This analysis resulted in five themes listed in table 2.

Table 1.

Answers to EOI data broken down by practice

	EOI questions	Practice A	Practice B	Practice C
1.	Team composition	PHY:1	PHY:2	PHY:2
		NP:1	NP:1	NP:2
		FPN:1	FPN:2	FPN:2
2.	Team champion	NP/FPN	FPN	PHY
4.	Approximate number of patients in practice	1600	4650	3710
6.	Physician renumeration model	Fee for service	Fee for service	Fee for service
9.	Culture of improvement	Very good	Very good	Very good
10.	Teamwork-working relationships within your practice	Collaborative	Co-operative	Collaborative
11.	Palliative care role	Working as a team, bringing in other experts as needed	Work independently	Work independently
12.	Quality indicators measuring success of the palliative care you are providing	Patient comfort/family satisfaction	Patient comfort/family satisfaction	Patient comfort/family satisfaction
18.	Challenges limiting team participation	Time available for FP to take this on*	Logistics of timing and resources*	No challenges
		Level of agreement for quest 13–17 1=do not agree at all to 10=agree very much
13.	Palliative care is an approach that is beneficial for those with advanced life limiting conditions	10/10	10/10	9/10
14.	Our team is committed to implement a new approach to palliative care that is rooted in primary healthcare	9/10	10/10	8/10
15.	Patient engagement in decision-making and planning is an important aspect of patient care	10/10	10/10	9/10
16.	There is a higher level of planning and co-ordination required for patients with multimorbidities and life limiting conditions	10/10	10/10	9/10
17.	There is willingness to work collaboratively with a palliative care team and other community providers to jointly meet patient needs	10/10	10/10	10/10

*Answers summarised.

EOI, expression of interest; FPN, family practice nurse; NP, nurse practitioners; PHY, physician.

Table 2.

Mixed methods themes: facilitators and barriers by practice

Themes	Practice A	Practice B	Practice C	Associated CFIR determinant
Timing of implementation	Barrier	Barrier	Barrier	Outer setting
Interpersonal collaboration within a practice	Facilitator	Barrier	Barrier	Inner setting: culture, networks and communication Characteristics of individuals: knowledge and beliefs
Established versus new practice	Facilitator	Barrier	Barrier	Inner setting: structural characteristics
EPIC project supports	Facilitator	Facilitator	Not mentioned	Outer setting: incentives Process: engaging (formally appointed internal implementation leaders)
Perceptions of EPIC as a type of care versus a limited term project	Facilitator	Barrier	Not mentioned	Intervention characteristics: complexity Characteristics of individuals: knowledge and beliefs
Champion position	Facilitator	Facilitator	Barrier	Inner setting: implementation climate Characteristics of Individuals: knowledge and beliefs Process: engaging (champions)

CFIR, Consolidated Framework for Implementation Research; EPIC, Early Palliation through Integrated Care.

Defining successful implementation of EPIC

Discussions with policy-makers and EPIC quality improvement personnel helped identify how successful implementation was defined when the EPIC project was initially funded. From the discussions it was evident that the intent was for practices to demonstrate adaptation of the palliative approach to care by identifying ways EPIC elements (eg, identification, assessment, documentation) could be incorporated into their practice routines. The quote below indicates that EPIC project leads viewed EPIC as best practice.

this is about improving palliative care and improving access and improving your ability to provide basic palliative care… this is quality improvement. If you are in primary care, you are definitely working with people who are very sick and dying. And so if you want to be able to provide the best care possible to them, then here’s a way to do that. And here are some tools, here’s some education. DM 002

Furthermore, the EPIC project had envisioned a culture change within practices that would indicate this type of care could be sustained after the project was completed. However, it was acknowledged that sustainable implementation of the protocol may not be completely accomplished during the project.

…what we're asking for though requires a culture shift. And that’s the part that I think is maybe bigger than the scope of this project. DM 002

A key element identified in the EPIC funding proposal, which was an essential part of a palliative approach to care, was documentation of patient tool results (ie, ESAS, PPS, SPICT) and goals of care in the electronic medical record. Documentation was considered key because it ensured a process was in place for the practices to continuously monitor quality of care. Tool results could be tracked across time, and it facilitated team communications about patients’ trajectories towards palliation. Furthermore, this information was used to inform implementation progress. Documentation was easily taken up by some individuals; however, a lack of time and difficulty with technology created barriers to others documenting as thoroughly.

Not all practitioners embraced the key role of documentation. One practitioner felt they were implementing EPIC because they were talking about it with their patients even though they were not fully documenting goals of care. Another practitioner from a different practice noted they used the ESAS and other tools, but they did not feel they had to document the results.

…so I think the goal for documentation was good. But I will be honest, I didn't do a very good job of documenting… I had in my mind where they’re at but I didn't actually every time use the actual number value… For me, it was more of a sense of where they were. PHC 006

A lack of documentation ended up creating work for other team members. One of the practice champions ended up spending valuable time going through a team member’s patient chart to see if the elements of EPIC project had been done.

So [I was] kind of like reading through all these visits, trying to find the information… we really do kind of need someplace to document it… PCP 005a

The EPIC project definition of successful implementation was used as our outcome because this was the intended outcome identified in the EPIC funding proposal. In summary, for our project, successful implementation of EPIC was defined and evaluated by the EPIC quality improvement personnel as practices that demonstrated (1) adaptation of the palliative approach to care by identifying ways EPIC elements could be incorporated into their practice routines, (2) a high probability this type of care could be sustained after the EPIC project was completed and (3) documentation of patient tool results (ie, ESAS, PPS, SPICT) results and goals of care. Using this definition, the three practices fell on a spectrum from successful (Practice A) to somewhat unsuccessful (Practice B) to not successful (Practice C) implementation of EPIC.

Practice comparison of EOI responses

Many of the EOI responses were similar, in fact most respondents had very positive responses scoring 10/10 on most items (questions 13–17) related to beliefs and team perceptions. However, some revealed important structural and cultural differences between practices (table 1). Applying the definition of successful implementation, the practice classified as having successfully implemented EPIC was Practice A. There were some noteworthy differences when comparing Practice A to the other two practices. Practice A had two PCPs (nurse practitioner and family nurse) coleading the role of champion (EOI question #2), and the other two practices had only one. Practice A had the least number of patients per staff of all three practices. Although the number was smaller for Practice A, when it was broken down by the ratio of patients to practitioners who cared for patients independently (ie, family practice nurses assisted physicians and nurse practitioners with their patients) Practices A and C were similar (1:800 and 1:928, respectively). Practice B however had a worse ratio of 1:1550. It should be noted that Practice A’s answer to EOI question #11 stated their team’s current role in managing palliative and end-of-life care for patients was working as a team and bringing in other experts as needed. In contrast the other two practices answered it was acting individually. Finally, of particular significance were responses to EOI question #18. Practice A stated they may have a challenge making time for the physician to take on the responsibility of implementing EPIC. Practice B identified there may be logistical issues of timing and resources, and Practice C responded they had no challenges to participate.

Comparing the EOI data from question #18 to interview responses for the following question: ‘On a scale of 1–10, with 1 being very easy and 10 being nearly impossible, how difficult has it been to implement the different components of the palliative approach to care project?’ Participants in Practice A had an average score of 3.75 (range 2–5) compared with Practice B average of 7.75 (range 7–9) and Practice C average of 5.5 (range 4–7). These responses indicate Practice A had a lower perception of difficulty when interviews were conducted after implementation even though they noted more challenges in the EOI data gathered prior to implementation.

Barrier and facilitator themes

Table 2 compares the themes identified during our mixed methods analysis of the EOI and interview data for each practice and whether they were perceived as facilitators, barriers or not mentioned in the interviews. In addition, the table identifies the associated CFIR domains and constructs associated with each theme. Although our analysis identified individual CFIR domains and constructs associated with barriers to and facilitators of implementation, our themes reflect the overlap and interaction between them. Briefly comparing practices based on their level of successful implementation of EPIC, it can be seen in table 2 that Practice A had the most facilitators.

Timing of implementation

The EPIC Study was conducted during the start of the COVID-19 pandemic, which had an adverse impact on practices implementing the EPIC protocol. Although primary care was an essential service that remained operational during the pandemic, many patient visits were modified to phone or Zoom appointments in alignment with COVID-19 guidance from public health. PCPs felt patient telephone appointments reduced their ability to identify individuals in their practices who might benefit from the EPIC protocol.

…we probably would have done a better job if people were…going to be coming in for little things because then that’s often when you see them, that face time, and identify…they really are having more problems with pain management or they're having more shortness of breath. PCP 005a

As well, the COVID-19 pandemic and online or telephone appointments made administering assessment tools that were part of the EPIC protocol more difficult.

And our process had been for people to be identified in the EMR [Electronic Medical Record], for it to pop up as an alert. And so when people checked in, they were given the ESAS in the waiting room… once people were no longer coming into the clinic, that process…broke down…and it became the responsibility of practitioners to remember to do it… PCP 007

Due to the COVID-19 pandemic changes, practitioner schedules decreased the number of providers (and therefore patient volumes) on site at any one time. For example, some practices split into two groups that worked on alternative days to align with COVID-19 guidance from public health infection prevention and control protocols. This had a negative effect on PCPs working collaboratively.

After the initial…enthusiasm, it was very difficult…to engage both of the physicians in what needed to be done and in trying to keep it going… especially with COVID, once we weren’t all working, because we broke off into sort of bubbles… At that point, it was almost non-existent, trying to get them engaged or trying to get any deliverables from them… PCP 005

Another factor related to timing of implementing EPIC was Practice C delayed rolling out the project until they completed their transition to a new electronic medical record system as they felt they could not take on the project at the same time as changing systems.

Interpersonal collaboration within a practice

As part of the inclusion criteria, all practices were considered by the health authority to be a collaborative family practice team defined as having physicians, nurse practitioners and nurses working in the practice in a collaborative way with shared responsibility for a practice population. In EOI question #10, practices chose the responses ‘collaboratively’ or ‘co-operatively’ to describe their team, and interviews provided more in-depth details on their level of collaboration. A provider in Practice A stated that they saw collaboration while helping each other with individual cases, managing EPIC patient lists and discussing different approaches to care.

…my colleague, … been great through it. Because… [they’ve] helped to manage the patient list and keep that up-to-date with regards to when we identify a patient who…meets the criteria. And then keeps track… Like we make sure we do the PPS and ESAS scores with each visit, each patient encounter. I do forget that sometimes. But [they are] really good at keeping track of the patient list, and when [the patient] died, date of death, where they died. PCP 003

Practices that did well implementing the EPIC initiative had multiple contextual factors identified during the analysis of their interviews that could classify them as being collaborative, this included: positive teamwork, clear designation of roles and strong communication among the team. Several interviewees mentioned that physicians often felt more time constraints working collaboratively, due to fee-for-service reimbursement pressures, “it sounds terrible, but you are limited…if you basically have every appointment…30 minutes” PHC006. However, when physicians collaborated with nurses on their team who were salaried, it allowed them to see patients and the EPIC protocol to be implemented.

And I think for us as nurses… and we don't have the same need to have high volume days… for some providers, the high volume of patients and efficiency is paramount above some other things…if some of the work could be done by somebody else then they see another few people and, you know, deliver health care to other people. PCP 005

Among the three practices, providers described distinct differences in lines of communication, collaboration on patient care and practice champions involvement. Teamwork and collaboration were demonstrated in Practice A by having a positive working relationship between the family practice nurse and nurse practitioner.

We just kind of made it part of our practice… You know, once we decided that we were going to do it, we just kind of did it. PCP 004

In response to a physician indicating they did not have time for aspects of EPIC that involved ordering palliative care drugs, the family practice nurse in Practice A took on the role of filling out the necessary forms. This designation of roles reduced the time needed to integrate new tasks into their practice routines.

…the previous physician that was working here with me, when she needed palliative orders, she could do the orders but I would submit the form for the palliative drugs to be covered through Pharmacare. PCP003

It is noteworthy that providers in Practice C struggled with team communication such as making time for discussions between team members. This was magnified during the COVID-19 pandemic when practitioners were often not physically in the practice together. A provider saw this as a barrier to integrating the EPIC innovation into their practice.

We don't particularly see a lot of each other. We will cover each other often when necessary… There’s not a lot of interaction. It’s unfortunate. It’s one of the deficits… PCP 008

As noted previously, there were also differences across the practices in their level of documentation. Documentation was more consistent and valued as an important method of team communication in Practice A.

…you're documenting it so that it becomes a source of communication within your team and also external to your team. (DM001)

Established versus new practice

Providers in less established Practices B and C indicated they found it difficult to navigate providing a new type of care while they were simultaneously facing the challenges of being a new practice and team.

Practice B had several individuals who were new to their career. Although the nurses in Practice B had a strong relationship and good communication, being a new practice meant they did not have organisational structures in place to facilitate regular team meetings. This made communication difficult.

Another challenge I thought for us as a team, a challenge to implementation was that we were not at a point where we were…having…collaborative team discussions about patient management…Other than the occasional hallway discussion, you don't really know about what’s happening with other patients in the practice… PCP 005a

Alternatively, Practice A had worked together as a team for 5–7 years, with providers indicating they had time to establish good communication routines in the practice. As well, their practice champion had been practising in the community for 15+ years. Providers in this practice felt it was their personal responsibility to promote a palliative approach to care in the practice, as exemplified in the quote below.

…even if a patient is not choosing to die at home, primary care providers can end up being involved with that patient’s care up until the time they're admitted to hospital…being involved in advance care planning, goals of care, what their wishes are, having those difficult conversations. PCP 003

There was a clear indication from the EOI and interview data that practices started with different experiences with a palliative approach to care. Some providers felt their practices were already doing this type of care, whereas other providers expressed curiosity about what a palliative approach to care entailed. EPIC project leads commented in their interviews that providers who were at different timepoints in their career may have influenced whether the practice ultimately decided to adopt a palliative approach to care. One advantage of the more established teams was a better understanding and historical relationship with palliative care services that could lead to a deeper understanding of the role primary care has in palliative care.

The palliative care service… is actually very receptive… There’s greater understanding from the primary health care teams about how those referrals for the specialist services come in. DM 001

EPIC project supports

Providers perceived EPIC project supports as facilitators to implement the EPIC initiative. Providers noted that the time to discuss EPIC and regular check-ins provided by the implementation facilitator was vital to their success. Most teams identified having funding for someone who acted as an implementation facilitator as a fundamental resource.

I feel like… [the implementation facilitator]…took a role to keep momentum going… And…talk about the data, talk about practice changes and challenges and things that were going well. So she was always…raising it to the top of people’s agenda by the mere fact that we had some ongoing contact with her. PCP 004

EPIC project lead interviews indicated that the lack of consistent individuals in the facilitator position may have mitigated the benefits. This was mentioned by practices that struggled, “I think it was difficult because there was a switchover of the project managers” PHC 007. However, practices that took ownership for the initiative did not mention it as a barrier to their implementation, even when it was probed in interviews. They noted that early in the project many checkpoints were key, but as time went on the providers and implementation facilitator were open to transitioning to less meetings.

And like those meetings were just good for reminders…it was more just that initial introduction…check-ins were more of a reminder for us than anything else just to…keep up what we're doing…once you start and you know what you're doing, it’s more…being accountable for your own practice. PCP 004

Individuals found the implementation facilitator provided a communication channel between themselves and EPIC project leads who could then relay information to additional policy-makers and managers in the health authority. Providers noted it was important that EPIC project leads knew of their struggles.

For me, certainly knowing that we could discuss ongoing challenges with [implementation facilitator], and then that [they] would take it forward as a concern. So that other people in leadership knew of the struggles we were having. PCP 005b

Other important EPIC supports mentioned by providers were the funding for LEAP training and the networking opportunities within LEAP. Although there were networking opportunities between practices involved with EPIC that helped providers troubleshoot issues with patient care, technology and how to ensure best access to patient services, there were no formal networking opportunities with practices outside the EPIC project besides those within the LEAP training at the beginning of the project. Networking with other practices was a positive part of the LEAP training. It provided an opportunity for practitioners to get together (in-person) before initiating implementation of an early palliative approach to care.

The incentive for me for this project was the networking opportunity we had [in the LEAP training]…any time that you get the opportunity to be brought together and network and gain knowledge and share experiences with other people like that. PCP 005a

However, LEAP did not provide continued connections between practices once the training was completed. Practices were left to work through struggles implementing this new care with providers connected to the EPIC project. When providers were asked if they knew of practices outside of EPIC doing this type of care, many were unaware, “No, I don’t know of any. And I don’t believe we’re very integrated” PCP 008, even though they had done LEAP training with some of these practitioners.

Interestingly, the practitioner who noted in their first interview that they enjoyed LEAP training for the networking opportunities subsequently remarked in their second interview (this individual moved to a different practice between their first and second interview) that moving to a different practice made them realise that practices outside of EPIC were already doing this care. They had not realised this because other than the LEAP networking opportunity most practices were siloed in their communities.

[new practice]…doing the early approach to palliative care just organically on their own, the way that they work. So I don't think there’s a whole lot for them to change… PCP 005a

Perceptions of EPIC as an approach to care as opposed to a limited term project

Providers had very different viewpoints on what the EPIC project meant to their practice. Those that viewed EPIC as an opportunity to make a shift in their practice to provide this care long term were seen as more successful by the EPIC project leads. Providers in Practice A considered EPIC as a way of providing care, as opposed to being only a limited term project, stating that it did not take much extra time to provide this care. They continually noted that they “just made it a part of their practice” PCP 003 and over time they did not have to think too much about the research side of things, as it became natural to do EPIC as part of implementing this type of care.

Well, because you're meeting with the patient anyway because they have needs. So you're just… asking a few extra questions. And if you're doing the ESAS and the PPS score… It’s kind of something we should have been doing all the time anyway with those patients…this has just become normal now. PCP 003

Although based on our definition of successful implementation, Practice C did not successfully implement EPIC, the physician in the practice perceived implementing this project as “not a big deal” PCP 007 because they had implemented similar EPIC processes previously when practising in another province. Furthermore, the nurse practitioner in the same practice perceived EPIC as a short-term project as opposed to a long-term change in care. They stated in the interview that they wanted to restart the project as opposed to just making a practice change as they felt unsuccessful in their first attempt.

But it would be interesting to do more on it now that there has been a more of an introduction to it. Because I feel like if I could re-start again…let’s put this into implementation. And a new batch of patients, because it does tend to fade a bit. But like now I’m kind of interested to re-do it again. PCP 008

Champion position

Providers placed a lot of merit on the role of the champion and how it facilitated implementing EPIC. In the EOI, each practice was asked to name an individual within the practice to champion the project. Although many factors contributed to successful implementation, nurses and nurse practitioner champions had a higher success rate in implementing changes in care routines to integrate the palliative approach to care into current practice routines than physicians, as exemplified by the following quote from an EPIC project lead.

…and I don't know if it’s a coincidence because we have such a small number of teams, but nurse practitioners are particularly accepting and they're very receptive. And part of that, I think, is the paradigm in which they work. DM001

As well, a physician acknowledged that their workload created a barrier to initiating a new way of providing care, comparing their workload to nurse practitioners. This may have been due to the fee-for-service model the physician worked within. They were understanding, but envious, that nurse practitioners and nurses had the time to take on implementing new practice innovations.

I think it would be hard [for family doctors] because I think someone like a family practice nurse… has the time and all that stuff to do it. I think it would be more difficult if it was…just family doctors. Because…they're going to go down to the time perspective, and how many things they already have on their plate… PCP006

In contrast, the nurse practitioner from Practice A did not feel the project took any extra visits so they did not see time as a barrier and felt this type of care would be easy to integrate into their practice.

When I found out it wouldn't involve any extra patient visits, it’s just a few extra questions that would be asked during patient visits, we felt that it was attainable and something that we could do. PCP 003

Discussion

The goal of this mixed methods study was to see how multilevel factors facilitated or hindered successful implementation of a palliative approach to care within primary care practices by applying an implementation science framework. In addition, the hope was that learnings from this project could inform implementation of future innovations within primary care. Overall, there were implementation barriers at multiple levels that caused some practices to struggle. However, these barriers were mitigated when practices were more established, collaborated to deliver care, had committed practice champions, leveraged innovation resources and perceived the innovation as an opportunity to make a sustained change in care. Some of these facilitators have been identified in prior research, such as using a team-based approach, practice champions, leveraging training resources and attitudes supporting a long-term change in practice.³¹ These facilitators also align with CFIR constructs associated with successful implementation (noted in table 2) such as practice-level communication, organisation, implementation climate and maturity, along with individual beliefs regarding EPIC adaptability and complexity.

Facilitators and barriers at different levels led to perceptions that there was not enough time to integrate changes into practice routines, document care on electronic systems and communicate or collaborate within the team. Time constraints were the result of individual PCP priorities as well as structural and cultural factors within practices, institutional factors associated with the health system and external conditions, such as the COVID-19 pandemic. A well-documented institutional factor is that PCPs in a fee-for-service renumeration system feel pressured to accomplish their workload in the allotted time³² and that this pressure can lead to increased stress.³³ In this study, nurse practitioners were salaried, whereas physicians were working on a fee-for-service basis, this renumeration method may have reduced physicians’ available time to implement EPIC. Leysen et al³⁴ support this conclusion, they found that the unpredictable and busy daily workloads of physicians were a large factor in their lower implementation of research protocols.

A practice-level factor that has been shown to reduce the time for implementing innovations is using electronic medical records linked to clinical decision support systems.³³ However, despite these time-saving benefits, it is a cultural shift for clinicians to use these tools.³⁵ The literature has shown that primary care physicians often resist following strict protocols,²⁰ preferring reflective frameworks. Electronic medical record systems, although considered best practice, can also be seen as burdensome.³⁶ Documentation burden for physicians has been demonstrated in the literature, physicians can spend twice as much time on electronic documentation and clerical tasks as compared with direct patient care.³⁷ Similar results have been seen in regard to nurses, where more than half of their time can be dedicated to electronic medical record data entry and retrieval^{38 39} These factors may explain why some PCPs in our study felt there was not enough time to document tool results and goals of care in the electronic medical record. However, documentation burden could be offset by improved communication and co-ordination of care.

Sharing the workload within the practice is another way to address time constraints and reduce the associated stress. Having a team-based approach can facilitate implementation,³² and having dedicated staff to help physicians reduces burnout.⁴⁰ In addition to reducing burnout, it may also increase team morale and improve organisational culture.^{41 42} Our study supports these finding. Practices that did well implementing the EPIC initiative had qualities that would classify them as being collaborative including good teamwork, working well interprofessionally, clear designation of roles and strong communication between team members. Among the three practices, there were also distinct differences in the practice champion involvement that contributed to successful implementation.

There is strong evidence that champions play a pivotal role in improving the success of healthcare implementation.⁴³ The champion role was pivotal to practices maintaining momentum and truly integrating EPIC into their practice and is an acknowledged strategy to help overcome implementation barriers.⁴⁴ However, the selection of an effective clinician champion can be difficult; there is no established method for identifying or training champions and historically they have come to light naturally in an environment given their influence, ownership, physical presence at the point of change, persuasiveness, grit and participative leadership style.⁴⁵ In the EPIC project the practice champions self-identified. The implementation science literature shows that champions can increase the adoption of an innovation at multiple levels, but their success often depends on educating and persuading others through social influence.⁴⁶ This is confirmed in our study, the most successful EPIC champions were collaborative and had a strong passion for implementing a palliative approach to care in their practice. The pre-existing relationships and readiness demonstrated within nurse/nurse practitioner teams may have increased their influence as practice champion.

The EPIC implementation facilitator was a key support in our study. The concept of facilitation has been well described in the implementation science literature and is especially prominent in the Promoting Action on Research Implementation in Health Services (PARIHS) framework.⁴⁷ PARIHS is a conceptual framework that proposes that successful implementation is a function of the complex interplay between the evidence to be implemented, the context of implementation and the way in which the implementation process was facilitated. Evidence gleaned from applying the PARIHS framework has demonstrated that facilitation is key to practitioners being successful implementers, by ensuring that they are ready, able and passionate about changing current practice.⁴⁷ The role of the facilitator is a cornerstone of successful implementation, as the facilitator acts as a resource to support practitioners and reduce the burden of implementation. This has been further supported by our study findings.

An important indicator of successful implementation identified in CFIR is implementation readiness.²⁶ The EPIC Study assessed practices’ implementation readiness by using the EOI form. The EOI data were useful; however, based on our findings, additional factors could be assessed such as how long the practice has worked together or how long providers have been practising. Readiness for change is well documented in the literature with half the failures to implement organisation change occurring because organisational leaders failed to establish a prior level of readiness.⁴⁸ In addition, our analysis allowed us to understand how Practice A’s initial hesitancy and recognition that implementing EPIC would be a challenge may have helped them to be successful, as they already knew that implementation would be time consuming and require practice changes. Acknowledging or anticipating challenges to implementation could be assessed as an indicator of better implementation readiness.

Limitations

There are acknowledged limitations to this research. First, there was a limited number of practices involved in the study and these practices had been chosen to be part of the EPIC quality improvement project. Furthermore, EOI responses may have been positively biased as this was used as a readiness tool to decide which practices would be involved in the EPIC project. Second, the COVID-19 pandemic impacted the way primary care practices were able to deliver care. Given EPIC project implementation coincided with the start of the COVID-19 pandemic practices faced additional barriers as they rolled out EPIC, some of these contextual barriers may not affect future iterations of EPIC implementation. Third, we collected data in one Canadian province; therefore, our results may not be transferable to other jurisdictions or settings. On the other hand, our in-depth description of the distinct barriers and facilitators practices faced increases the transferability of our findings to other settings. A final limitation is we used the EPIC project definition of successful implementation because it aligned with the intent of their quality improvement project. This definition was slightly different from practitioners’ perceptions of successful implementation. If our definition had not included documentation, the other two practices may have been considered more successful. In summary, by applying an implementation science lens, we were able to identify barriers and facilitators that may be transferable to other settings.

Conclusion

In this project, facilitators related to successful implementation were associated with team composition; how strongly they prioritised the initiative; and years of experience working together. Future implementation work should consider assessing these facilitators to better gauge primary care pre-implementation readiness. Our results highlight how important collaborative teams and champions at the practice levels are for mitigating barriers to successful implementation. As well, providing a dedicated facilitator at the health system level enhanced practices’ ability to successfully implement the innovation. Future quality improvement initiatives should embrace the utility of champions and facilitators. Also, PCP perceptions impact successful implementation, practitioners who made it ‘part of their practice’ did better. Finally, future initiatives should respect that culture shifts in primary care practices take time, but barriers to implementation can be mitigated by appropriate facilitators.

Data availability statement

No data are available.

Ethics statements

Patient consent for publication

Not applicable.

Ethics approval

This study involves human participants and was approved by the institutional and local ethical review committee of Nova Scotia Health (Nova Scotia Health REB File #1025792) and the institutional review board. Participants gave informed consent to participate in the study before taking part.