Article Text

Barriers and facilitators to learning health systems in primary care: a framework analysis
  1. Georgia Fisher1,
  2. Maree Saba1,
  3. Genevieve Dammery1,
  4. Louise A Ellis1,
  5. Kate Churruca1,
  6. Janani Mahadeva2,
  7. Darran Foo1,2,
  8. Simon Wilcock2 and
  9. Jeffrey Braithwaite1
  1. 1Centre for Healthcare Resilience and Implementation Science, Australian Institute of Health Innovation, Sydney, New South Wales, Australia
  2. 2MQ Health General Practice, Macquarie University, Sydney, New South Wales, Australia
  1. Correspondence to Dr Jeffrey Braithwaite; jeffrey.braithwaite{at}


Background The learning health system (LHS) concept is a potential solution to the challenges currently faced by primary care. There are few descriptions of the barriers and facilitators to achieving an LHS in general practice, and even fewer that are underpinned by implementation science. This study aimed to describe the barriers and facilitators to achieving an LHS in primary care and provide practical recommendations for general practices on their journey towards an LHS.

Methods This study is a secondary data analysis from a qualitative investigation of an LHS in a university-based general practice in Sydney, Australia. A framework analysis was conducted using transcripts from semistructured interviews with clinic staff. Data were coded according to the theoretical domains framework, and then to an LHS framework.

Results 91% (n=32) of practice staff were interviewed, comprising general practitioners (n=15), practice nurses (n=3), administrative staff (n=13) and a psychologist. Participants reported that the practice alignment with LHS principles was influenced by many behavioural determinants, some of which were applicable to healthcare in general, for example, some staff lacked knowledge about practice policies and skills in using software. However, many were specific to the general practice environment, for example, the environmental context of general practice meant that administrative staff were an integral part of the LHS, particularly in facilitating partnerships with patients.

Conclusions The LHS journey in general practice is influenced by several factors. Mapping the LHS domains in relation to the theoretical domains framework can be used to generate a roadmap to hasten the journey towards LHS in primary care settings.

  • Primary Health Care
  • Delivery of Health Care
  • Health Services Administration
  • Patient-Centered Care
  • Information Literacy

Data availability statement

Data are available upon reasonable request.

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See:

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


  • The learning health system (LHS) concept is gaining traction in multiple healthcare settings yet remains relatively underexamined in primary care, particularly through the lens of implementation science.


  • This study uses an established implementation science framework to describe key facilitators and barriers to the cultivation of an LHS in a primary care setting.


  • We compare these factors to the small existing body of literature in this area and propose practical solutions to implement the principles of the LHS into primary care practice.


Primary care is the ‘frontline’ of healthcare; it is the first point of contact with the health system for most people1 2 and thus an essential component of care delivery. Primary care can reduce overall health costs and relieve pressure on other areas of the health system; for example, by reducing the number of preventable or unnecessary presentations to emergency departments.3 In many countries, including the UK and Australia, primary care is chiefly provided through general practitioners (GPs).4 5 However, general practice is under pressure. Ageing populations and an increase in chronic disease have heightened the demand for primary care services.6 7 Growth in the workforce has flatlined8 9 with fewer GPs providing care for more people,10 and many GPs unsure of the viability of their practice.11 More recently, the unprecedented challenge of a global pandemic has necessitated system-wide reorganisation12 and placed many additional stresses on GPs and the system in which they work.13 The solutions to these entrenched implementation issues are by no means easy or short term, but in the interim, general practice needs a viable framework to guide the steps towards a sustainable and high-performing primary care system. In response, the concept of a learning health system (LHS) has been proposed.14 According to the National Academy of Medicine (NAM; then the Institute of Medicine), an LHS is a system that ‘consistently delivers reliable performance and constantly improves, systematically and seamlessly, with each care experience—in short, a system with an ability to learn’.15

LHSs have been embraced by multiple providers who have reported a variety of benefits, including increases in evidence-based care delivery, improved clinical outcomes, higher levels of patient-centred care and reductions in adverse events.16 The core characteristics of an LHS identified by the NAM include: (1) science and informatics that provide real-time access to knowledge and digitally capture care delivery; (2) patient–clinician partnerships, where patients are engaged, empowered participants in care; (3) incentives that reward high-value care and transparency; and (4) a continuous learning culture that is supported by the system and its leaders.17 More recently, a fifth characteristic has been identified, structure and governance, that aligns policy and regulation to facilitate research, collaboration and learning.18 An LHS can manifest at the micro level of the practice right through to the macro level of the healthcare system. This makes the LHS model well suited for primary care, and able to help support the performance of individual general practices and their interactions with the larger healthcare system. However, despite their promise for primary care, most reports describe LHS in tertiary hospital settings, with few that focus on the unique context of general practice or its providers on the frontlines of care.19

Even less frequent in the literature are reports of primary care LHS that are underpinned by principles of implementation science, a field that aims to establish what works and why in the translation of research evidence into practice.20 The simplicity of the five-part LHS framework is somewhat deceptive; not only are its components multifaceted and their role unpredictable,21 but they must also be applied in the broader complex adaptive system of healthcare.22 Subsequently, there are many factors that affect the success of the LHS in the real world. Implementation science frameworks provide an evidence-based explanation of such factors, enabling us to leverage facilitators, and overcome barriers. An established method of doing so is via the theoretical domains framework (TDF), which brings together multiple theories of behaviour change into a single 14-item framework.23 In the present study, we used the TDF to conduct secondary analysis of data obtained in our previous investigation of an LHS in the general practice setting. We aimed to identify and describe the barriers and facilitators to adopting LHS principles specific to each of the five components of the LHS framework, and to provide evidence-based implementation recommendations for general practices who are making the journey towards an LHS.


This study is a secondary analysis of data generated in a qualitative investigation of an LHS in primary care.24 Our original investigation brought together researchers from the Australian Institute of Health Innovation (AIHI) and staff from MQ Health General Practice (MQGP) in a qualitative study that used an embedded research approach and that was codesigned by the research team from AIHI, and clinicians and senior clinic administrators from MQGP.

Study setting and context

MQGP is a not-for-profit, university-based general practice that operates in the northern suburbs of Sydney, Australia across two sites: one adjacent to a hospital on the university campus, and one in a suburban location.24 The practice is part of the broader entity of MQ Health, which also comprises specialist clinics, an inpatient hospital, and allied health, medical imaging, radiotherapy and on-site pathology services. Most MQGP staff are employees of MQ Health and have access to educational resources available to employees of Macquarie University. Due to its university affiliation, MQGP is actively involved in research and teaching activities and has a strong record of quality improvement initiatives. MQGP also participates in its local Primary Health Network (PHN), which is a government-initiated, independent organisation that aims to streamline and coordinate primary care services. At the time of the study, MQGP employed 17 GPs, 4 clinic nurses, 13 administrative staff and a clinical psychologist across both sites.

Embedded research approach

In our embedded research approach,25 a research assistant from AIHI (GD) was introduced to all MQGP staff at a clinic practice meeting in July 2021, and then worked alongside practice staff until December 2021. The embedded researcher was included on all staff emails, liaised closely with the practice’s business manager and GPs and attended the practice’s ‘strategy day’. The embedded researcher was also involved in the coordination and data collection of the present study.

Data collection and recruitment

We conducted semistructured interviews with MQGP staff. The research team used the modified five-characteristic NAM LHS framework18 to design the interview questions, which were then reviewed by multiple clinical and administrative staff at MQGP to ensure their clarity and relevance to the practice. All practice staff were invited to take part in the study via email, where they were provided with participant information and consent forms that outlined the purpose of the research study. There was no sample size calculation for the study. Instead, we aimed to interview a sample that was representative of all clinic staff. Interviews were conducted in October 2021 by a senior research fellow (LE) or the embedded research assistant (GD), either in person at the general practice or via teleconference. The interviewers had prior training in qualitative research methods and interviewing.


Interviews were audio recorded and transcribed verbatim. To deidentify the data, staff were given a unique code that consisted of their role (ADMIN, administrative staff; GP, general practitioner; NUR, nursing staff) and a random number. Deidentified interview transcripts were imported into NVivo V.20. The secondary analysis of study data was conducted by two members of the research team who were independent of the original study data collection (GF, MS), who conducted a deductive framework analysis26 with the TDF and LHS framework (the Independent Analysis Team, IAT). The components of each of these frameworks are detailed in table 1.

Table 1

Elements of the LHS and TDF frameworks

Both members of the IAT first coded an interview transcript together in real time, categorising the data into the domains of the TDF. Then, the IAT independently coded five transcripts, iteratively checking agreement and discussing conflicts after each. After the fifth transcript, the IAT’s mean±SD agreement across all TDF domains was 87.6%±10.1%. The IAT researchers then each independently coded half of the remaining transcripts. Next, they used the modified five-component NAM LHS framework18 to organise the data in each TDF determinant; this second deductive process ensured data coded to each TDF determinant were also described in relation to the key tenets of an LHS. Counts of the number of participants who made statements coded to each TDF determinant and each LHS component were recorded. Finally, the IAT researchers met and inductively generated belief statements that were relevant to each domain of the TDF and each component of the LHS. Final results were reviewed for validity by four members of the original study team: two senior academics from AIHI (LE, JB), one GP from MQGP (DF) and the embedded research assistant (GD).


A total of 32 out of 35 (91%) practice staff were interviewed, which included GPs (n=15), practice nurses (n=3), administrative staff (n=13) and a psychologist (n=1). Three clinicians were unable to attend their scheduled interview, and as data saturation was reached, these interviews were not rescheduled. Interviews lasted between 17 and 50 min (mean 35.5). Participating staff had been working at MQGP for between 3 weeks and 15 years.

Barriers and facilitators to an LHS

The environmental context and resources available to participants and their social and professional role and identity were key determinants in engagement with most domains of the LHS. Reinforcement was particularly important for the development of patient–clinician partnerships and engagement with incentives in the LHS, while a several domains of the TDF had a reciprocal relationship with the practice structure and governance; for example, clear policy facilitated the development of a strong professional identity, which then in turn facilitated access to and understanding of policy. Key barriers and facilitators that are relevant to each domain of the TDF are reported in table 2, according to each of the five LHS components. Figure 1 provides a visual summary of the framework analysis and the relative proportions of each TDF domain described in each LHS component.

Table 2

Key TDF determinants and associated belief statements for each domain of the LHS framework

Figure 1

Results from the framework analysis using the learning health system (LHS) and theoretical domains framework (TDF) domains. Each domain of the LHS is represented by a different colour. Each coloured circle represents a TDF determinant. The sizes of the circles represent the number of participants who reported a quote in the respective TDF determinant, which are also written on each circle.


The codebook and exemplar quotes on which these results were based are available as online supplemental material associated with this manuscript. The full study dataset is available from the authors on reasonable request, subject to ethical approval.

Supplemental material


In our original study, we presented a case study of an LHS within an Australian primary care setting and showed that it was operating within several dimensions of the LHS framework, and that its staff were willing to embrace additional elements of the LHS.24 In this secondary analysis, we used the TDF to describe barriers and facilitators to converting this willingness into reality. In all LHS domains there was a consistently reported influence of environmental context and resources; for example, the MQGP affiliation with a university was described as a strong facilitator of learning, and the unique general practice environment was reported to shape patient–clinician partnerships. The professional role of participants was a second consistently reported determinant, influencing access and attitudes to learning and incentives. The reported impact of other determinants varied across LHS dimensions; for example, continuous learning culture was mediated by social influences, where strong social relationships were reported to facilitate informal learning, while a lack of knowledge of clinic structure and governance was described as a barrier to its effectiveness. Overall, our results show that implementing the principles of an LHS in this primary care setting was influenced by many behavioural determinants, some applicable to healthcare in general, but most specific to the general practice structure and environment.

A key strength of the study was its codesign, which allowed it to reflect the goals of both the research team and the staff of MQGP. Further strengths included the high participation rate and broad recruitment strategy, which enabled a comprehensive description of behavioural determinants from the perspective of clinical and non-clinical staff. Additionally, this secondary analysis was conducted by an IAT that did not participate in the original study and were thus less subject to biases from their relationships with practice staff or from the original interviews. The primary limitation of this study was the inclusion of only one organisation, limiting the generalisability of our results to other primary care settings. Generalisability is also limited by the affiliation of the practice with a university which, while a facilitator of the uptake of LHS principles, is relatively uncommon in the Australian context. A final limitation was the timing of the present study, which was conducted during and after significant public health restrictions associated with the COVID-19 pandemic. These restrictions, and their removal, would likely have influenced the responses of participants.

Despite these limitations, our results have encouraging similarities with the few other empirical investigations of primary care LHS that are grounded in implementation science.19 Pestka and colleagues qualitatively evaluated the lessons learnt from their implementation of a primary care LHS in the USA.27 They, too, reported that clearly defined roles and the incentivisation of value-based care were facilitators to the development of an LHS, as was the use of a weekly newsletter to communicate essential information. However, their investigation took place in a system of 40 primary care practices, much larger than the two practices described in the present study. The facilitatory effect of a weekly newsletter was diluted by a larger LHS size, where at times people had ‘no idea what was going on at other stations’,27 a finding that was echoed by another investigation of a province-wide primary care LHS in Canada.28 The same study also reported that the perceived difficulty or cognitive load of a technology was a primary barrier to its use, and that a perceived increase in the quality and efficiency of patient care was a motivation for participants to engage in the LHS,28 findings similar to our results. However, a key difference between their investigation and our own was the type of incentives that motivated participants; in the Canadian province-wide primary care LHS competition or peer pressure were motivators for engagement,28 while our participants reflected that they were primarily motivated by the rewards of providing better patient care and developing a sense of comradery with their colleagues. These differences may reflect the different social contexts in which the studies were conducted, particularly the influence of the COVID-19 pandemic, in which healthcare workers likely banded together to deal with high levels of uncertainty and stress.

The results of our own and other empirical investigations suggest that while some barriers and facilitators are unique to certain contexts, others are common to many journeys towards a primary care LHS. These are summarised in box 1, which also describes possible strategies for primary care practices to facilitate their journey towards an LHS. A notable facilitator that likely applies to all contexts is external support, as many primary care providers work in small independent community practices which limits their access to resources.29 Affiliations with academic and professional institutions, including the use of codesign and embedded researchers, or collaborations of multiple primary care practices are viable strategies that cultivate a primary care LHS. Additionally, our results suggest that it is not only patient–clinician partnerships that are important in the primary care LHS, but rather that administrative staff also play an important role in the patient experience. As such, primary care practices that aim to become LHS should invest in training, involvement and retention of all staff, not just those in clinical roles.

Box 1

Summary of five key barriers and facilitators to a learning health system (LHS) in primary care and five proposed solutions.

Key barriers

  • Unclear policy and roles.

  • Poor data quality.

  • Complex learning requirements.

  • Physical distance between teams.

  • Poor communication with patients.

Key facilitators

  • Strong leadership.

  • Desire to help patients.

  • Shared organisational goals.

  • Culture of patient-centred care.

  • Communication of progress and goals.

Key solutions

  • Formal lines of patient communication and feedback (eg, online reviews).

  • Diverse modes of care and communication (eg, telehealth).

  • Weekly practice newsletter to share updates and progress.

  • Multidisciplinary leadership teams that model a learning culture.

  • Mentorship and ‘buddy systems’ between senior and junior staff.

  • Each point describes a barrier, facilitator or solution described in at least two of the three following papers: (a) Nash et al,28 (b) Pestka et al27 or (c) current study.


There are numerous benefits, success factors and barriers in primary care settings making the transition to LHS. These factors should feed into a roadmap to assist primary care settings that are at different stages of the journey towards an LHS.

Data availability statement

Data are available upon reasonable request.

Ethics statements

Patient consent for publication

Ethics approval

This study involves human participants and was approved by Macquarie University Human Research Ethics Committee (reference number: 52021905624229). Participants gave informed consent to participate in the study before taking part.


The authors gratefully acknowledge all interview participants who kindly gave their time to partake in this research.


Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.


  • Contributors JB, SW, JM, GD, KC and LE conceptualised the study. GD and LE collected the data. GF and MS analysed the data with input from LE and GD. GF created the study figures with input from MS, LE and JB. GF drafted the manuscript. JB is guarantor.

  • Funding JB is funded by multiple grants including the National Health and Medical Research Council (NHMRC) Partnership Grant for Health Systems Sustainability (ID: 9100002) and NHMRC Investigator Grant (ID: 1176620) on the learning health system and its applications.

  • Disclaimer The funders had no role in the design, analysis, and interpretation of the research, or drafting of the manuscript.

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.