Identifying barriers and facilitators to audit & feedback in general practice in Italy: a qualitative study

Abstract

Background

Audit & Feedback (A&F) interventions are recognized as effective strategies for enhancing healthcare quality, yet they remain relatively underutilized in Italian primary care settings. Recent literature highlights the importance of examining the factors influencing the implementation of these interventions across diverse clinical and organizational settings. Thus, this study aimed to identify such factors by exploring the experiences of professionals involved in an A&F intervention tailored for General Practitioners (GPs) treating chronic illnesses.

Methods

Using qualitative research methods, this study organized focus groups (FGs) separately for each professional role involved in A&F: GPs, GPs acting as internal group coordinators, Public Health Physicians (PHPs) who served as liaisons between the Local Health Districts (LHDs) and groups of GPs, and Directors at LHD involved. All participants were invited to contribute to the FGs. A moderator guide was prepared, and sessions were audio-recorded and transcribed verbatim. Initial open coding led to development of a preliminary codebook, refined through iterative recoding cycles. Resultant codes were organized into 3 categories: barriers, facilitators, and recommendations and suggestions.

Results

Participation rates among FGs varied, ranging from 9.3% (8/86 GPs) to 77.8% (7/9 PHPs). Participants identified barriers and facilitators to the implementation of A&F strategies. Factors described both as obstacles and facilitators to adoption included the time and organizational commitment; the participants’ motivation; the credibility of data sources; the characteristics of the GP coordinators and the PHP liaisons; and the perceived usefulness of participating in A&F. Furthermore, barriers included the excessive theoretical load of the education and training program; some contextual factors; personal characteristics of GPs; heterogeneity in available technology, and the habit of GPs working as individuals. Conversely, facilitators included some logistical and organizational factors; the involvement of recipients in the topic’s identification; social interactions with other professionals; and incentives. Participants proposed some recommendations and suggestions on how to improve A&F interventions involving GPs.

Conclusions

FGs allowed participants to share valuable insights into the barriers and facilitators essential for shaping improvement strategies in A&F interventions within general practice. Implementing these findings may enhance the acceptability, feasibility, and effectiveness of future A&F interventions and increase their uptake.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12913-026-14141-w.

Text box 1. Contributions to the literature
• This study identifies obstacles and facilitators in Audit & Feedback (A&F) implementation, providing suggestions on how to enhance the effectiveness of this strategy and to reduce variability.
• It collects the experience reported by General Practitioners (GPs) participating in an A&F intervention: GPs are a critical target either because they are pivotal figures in Primary Care and because they have frequently less knowledge, attitudes and experience in quality improvement strategies than hospital physicians.
• It provides an example of collaborative work between GPs and Local Health Districts.
• It shares a methodological experience in applying a qualitative approach to explore A&F barriers and facilitators.

Background

Audit and Feedback (A&F) is defined as “any summary (verbal or written) of clinical performance of health care over a specified period of time aimed at providing information to health professionals to allow them to assess and adjust their performance” [1]. These interventions offer a multifaceted strategy that improves the quality of care and encompasses adherence to guidelines, health outcomes, safety, and equitable access to healthcare services. Among the quality improvement strategies, A&F is particularly appropriate when the aim is to compare clinical performance against an established standard of care. Furthermore, it is a flexible tool that can be used in different settings, also in combination with other strategies, such as clinical pathways or educational interventions. A recent systematic review of 292 studies found a mean absolute increase in professional practice of 6.2% and a 95% confidence interval ranging from 4.1% to 8.2%, confirming previous evidence of A&F efficacy [2, 3]. Despite decades of experience and evidence supporting the efficacy of such interventions, available data indicate considerable variability in the magnitude of the effects [3, 4]. The most recent works invite researchers to focus their further studies on discerning ways to optimize the effectiveness of A&F interventions [2–4]. “EASY-NET,” a network research program co-funded by the Italian Ministry of Health and participating Italian Regions, sought to assess the effectiveness of A&F in enhancing care across various clinical conditions within diverse organizational and legislative contexts. This initiative, detailed by Acampora et al. [5], involved seven Italian Regions, each developing specific projects targeting health conditions ranging from chronic to oncological and acute diseases. These projects spanned multiple care settings, including hospitals, primary care, integrated healthcare, medical facilities, and surgical procedures. The overarching objective across all initiatives was to compare different A&F strategies’ effectiveness in diverse contexts and care scenarios, with the aim of refining clinical practices and mitigating health disparities. Moreover, an additional objective was to identify obstacles and facilitators in A&F’s utilization and in enhancing its effectiveness. Understanding how optimizing A&F could, eventually, catalyze its broader adoption.

Within this framework, the Lazio Region piloted an innovative A&F strategy focused on primary care for patients with chronic diseases, specifically chronic obstructive pulmonary disease (COPD) and type 2 diabetes mellitus (DM2), involving General Practitioners (GPs), clinical specialists, public health physicians (PHPs) and Local Health District (LHD) directors, as previously described by Nardi et al. [6]. GPs, who are self-employed professionals working as independent contractors in a Local Health Authority (LHA), are pivotal figures in the Italian National Health Service (NHS) and play a critical role in managing patients with chronic illnesses effectively. To bolster primary care’s centrality, it is imperative to engage GPs in enhancing clinical practice’s appropriateness and effectiveness in managing patients with chronic conditions. This necessitates novel service management and organizational modalities to achieve genuine clinical governance in close collaboration and coordination with hospital services. In Italy, LHDs are functional subdivisions of a single LHA and represent administrative units responsible for healthcare delivery within specific geographic areas: LHDs emerge as pivotal players, acting as steering committees in planning, coordinating, implementing, and monitoring various prevention and care activities pertinent to primary care.

In this context, coordinating clinical governance requires overcoming the traditional individual approach of GPs to enlist their participation. Recent Italian regulations, such as DCA U00565/2017 [7], aim to promote and fortify group medicine through initiatives like Primary Care Units (Unità di Cure Primarie, UCP) and Complex Primary Care Units (Unità Complesse di Cure Primarie, UCCP), the former including only general practitioners and the latter including general practitioners, specialists, and other professionals. One of the aims of these initiatives is to promote the use of strategies to improve the quality of care, such as diagnostic and therapeutic care pathways and A&F. In particular, the comparison of clinical performance enabled by A&F may be more robust in group-based practice than in individual practice, given opportunities for shared reflection and peer comparison.

In addition, while A&F strategies aim to enhance guideline adherence, reduce inappropriateness, and minimize care variability, they also provide an avenue for disseminating a culture of continuous healthcare quality improvement. They foster awareness regarding critical aspects of care pathways and enhance collaboration among professionals across different settings and disciplines [4]. However, these strategies remain relatively underutilized in Italian primary care and territorial settings and may offer a path to engaging GPs in coordinated efforts. Hence, this study aims to explore the perceptions of obstacles and facilitators to the implementation of the A&F intervention among physicians who participated. To this purpose, we organized focus groups with the different professionals involved in the intervention, to bring out barriers and facilitators that can act during the whole implementation process. Ultimately, disseminating the potential of A&F by fostering knowledge enhancement will shape attitudes, influence behaviors, and improve coordinated care.

Methods

Audit & feedback intervention

The pilot Audit & Feedback (A&F) Intervention for Professionals involved qualitative methodology, using focus groups (FGs). The A&F intervention was implemented in two LHDs in the LHA Roma 1 in the city of Rome. The intervention was tailored for GPs and included directors from LHDs and PHPs who served as liaisons between LHD and groups of GPs. The A&F intervention started in April and lasted in December 2022, whereas the qualitative study started at the end of the intervention, in January 2023 and ended in January 2024 (Fig. 1). The A&F intervention focused on the care of patients with one of two diseases, each with high prevalence and care burden: diabetes mellitus type 2 (DM2) and chronic obstructive pulmonary disease (COPD); the diseases were selected based on evidence indicating low adherence to the guidelines and variability among districts and among GPs in DM2 and COPD care [8]. Overall, GPs received external feedback—once at the beginning and again at the end of the intervention. The feedback included a set of indicators evaluating healthcare for patients affected by DM2 or COPD and were calculated at LHD level and at single GP level. The feedbacks were generated by the Department of Epidemiology (DEP) of the Lazio Region utilizing data from the Health Information Systems (HIS).

Fig. 1.

Timeline of the A&F intervention, and FGs. Timeline of the A&F Intervention and FGs. The A&F intervention started in April 2022 with the series of theoretical lessons. Practical activities of the clinical audit started in May and were carried out in three subsequent encounters (May, June, and December 2022). Following the pilot A&F intervention (January and February 2023), qualitative data collection was planned. Four FGs were conducted between April and June 2023, involving the professionals who participated in the A&F intervention. Transcripts (data) from the FGs were coded and analyzed between August 2023 and February 2024

A personal report providing individual feedbacks was sent by e-mail to each GP. Indicator results were presented using text, graphs and tables. The report included a comparison between the performance of the GP and the Regional, LHA, and LHD means, as well as with the anonymized results of the other GPs of the group. Indicators detail, as well as a detailed description of the study design, have been reported elsewhere [9].

GPs had to engage in structured clinical audit activities as part of an accredited education and training course. This course involved clinical audit experts, clinical specialists in diabetology and pneumology, LHDs’ directors and PHPs, as well as epidemiologists from the DEP. First, GPs attended frontal lectures (in groups of 30–40 individuals) covering clinical audit methodology, feedback techniques, and clinical guideline updates pertinent to COPD and DM2. Subsequently, they engaged in practical activities encompassing all stages of a clinical audit cycle [10]. These stages included identifying criteria, indicators, and standards; defining data sources and collection methods; conducting data analysis; critically discussing results; and formulating improvement actions.

For the training activities, participants were organized into groups of 10–15 individuals, preferably affiliated with the same UCP. In each group there were one of the GPs acted as an internal coordinator and a PHP working at the LHD. The GPs coordinators were group referents who supported GPs in their practical part of the course if needed and facilitated the communication between GPs and PHPs. The PHP was responsible for supporting interaction and communication among groups and between groups and the LHD. Furthermore, PHPs supported group meetings organization and the analyses of data retrieved from GPs management software. Eventually, the LHD directors provided general supervision. To streamline practical activities, each group was assigned one of the two diseases of interest (DM2 or COPD), based on group’s preference.

As mentioned before, participating GPs received details about the feedback reporting process and outcome indicators calculated by the DEP using data from the HIS. Additionally, supported by PHPs, participants were asked to use their own practice management software to extract local data about patients affected by DM2 and COPD, and to calculate the same kind of indicators as those from the DEP using data collected on their own. The objective was to enable them to compare the same indicators calculated using different data sources.

Focus groups procedure

To gain insights into the participants’ experience, which would not be possible to obtain with surveys or common quantitative approaches, qualitative data collection techniques were employed [11]. FGs are ideal for gaining insight into the experiences and perspectives of various stakeholders, such as program participants [11–14]. This approach was used with the aim of collecting spontaneous opinions from participants about possible obstacles and facilitators for A&F (starting from their own experience in the pilot intervention) and, consequently, of defining improvement proposal for future A&F strategies. The research team conducted four FGs in Italian.

Figure 1 depicts the timeline of the A&F intervention, FGs; Table 1 describes the structure of the FGs.

Table 1.

Invited professionals for FGs

Qualitative meeting	Professionals
Focus group n. 1	GPs from LHD 2 who were learners and trainees
Focus group n. 2	GPs from LHD 13 who were learners and trainees
Focus group n. 3	GPs who served as internal group coordinators in work and discussion meetings
Focus group n. 4	PHPs who served as liaisons between LHD and groups of GPs

To report the results of the FGs, the study team followed the COREQ (Consolidated Criteria for Reporting Qualitative Research) guidelines in the development of this manuscript [15].

Researcher characteristics and reflexivity

One researcher (AA) moderated all the FGs. The moderator was a medical doctor (MD), specialized in Hygiene and Preventive Medicine, with a doctoral degree (PhD) in Public Health and Epidemiology, with knowledge and experience in qualitative research and A&F application. She was working at the DEP and coordinated the implementation of the pilot A&F intervention. Two researchers (AM and ELG) collaborated as moderator assistant in the FG sessions. Both of them were MD residents in Hygiene and Preventive Medicine collaborating with DEP. Data collection, coding, and analysis were conducted in consultation with a qualitative research scientist (PhD) with over 20 years of experience (RF).

Different FGs were organized for each type of professional, separating coordinators and PHPs working at LHD from GPs, in order to facilitate the free expression of participants’ views. Each FG started with a brief presentation and informal conversation to help put participants at ease. Discussions were guided by open-ended questions, and all participants were encouraged to share their experiences. To reduce subjectivity, the codebook was developed by three researchers, who performed the analyses independently; disagreements in coding were discussed and resolved by consensus (see “Data analysis” section below).

Participant recruitment

All the individuals involved in the A&F intervention were invited to participate in the FGs (Table 1).

FGs were organized by participant role with a maximum of 11 participants per FG.

The first FG involved PHPs. A second FG involved GPs serving as coordinators. Finally, two separate FGs were conducted with GPs in the role of learners/trainees in each LHD. Since there were only two LHD Directors, the format was adjusted and a dyad interview was conducted to gather their perspectives.

The recruitment of participants in the FGs was facilitated by the LHD Directors to ensure that everyone was informed about the opportunity to participate. PHPs collaborated in the recruitment of GPs. First, LHD Directors reached out by telephone to each GP involved in the intervention presenting the initiative and emphasizing the importance of their participation in the upcoming FGs. Second, a formal invitation was sent via e-mail. During recruitment, prospective participants were informed that the goal was to gather their perspectives on the A&F experience and to understand their interest in being involved in future similar activities. Prospective participants were informed that their names would not be linked to any reports or papers related to the study, that participation was voluntary, and that we sought their honest opinions, so there were no right or wrong answers to the questions they would be asked.

Instruments

A semi-structured guide was drafted for the FGs, consisting of opening questions designed to foster a welcoming atmosphere where all participants felt comfortable, initiate discussion, and encourage interaction (Appendix 1 in Supplementary Materials). Transition questions followed, which facilitated the shift to key questions. Key questions focused on identifying facilitators and barriers to implementing A&F interventions in primary care involving GPs. The key questions were accompanied by a series of prompts designed to assist the moderator/interviewer in exploring emerging themes and eliciting details or seeking clarifications from the participants.

The guide was formulated during the study design phase, tested during the first FG, refined based on emerging elements, and adapted to the specific participant groups. The guide iterations were designed to allow participants an opportunity to reflect on their experience and share their perceptions. If a topic arose organically during the conversation, the corresponding question was omitted, and topics were explored in the sequence that naturally unfolded.

The guide development process is detailed in Appendix 2 [13, 15].

Setting and data collection

FGs were conducted in Italian with a moderator/interviewer and an assistant who was responsible for taking notes. The FGs were organized in person and in locations selected based on participant preference and ease of access.

Throughout the discussions, the moderator facilitated active interaction among the participants and encouraged them to engage with one another rather than solely responding to the moderator’s questions.

The FGs were audio-recorded and transcribed verbatim to facilitate a thorough analysis. Additionally, both the moderator and assistant took field notes. Participants were informed about the audio-recording of the discussion, its purpose, its intended use, and who would have access to it; they were assured that their names would not be linked to the transcripts or any study materials. Furthermore, it was clarified that the recordings would be destroyed after transcription and analysis. Informed consent to participate and have the conversation audio-recorded was obtained from all the participants before the start of the FGs.

Data analysis

A codebook was developed inductively. Initially, one researcher (AA) transcribed the four audio files using the Microsoft Word © automatic transcription tool. Subsequently, transcripts were checked for accuracy and deidentified. Then, three members of the research team (AA, AM, ELG), independently open-coded one transcript each to inform the development of a codebook. The team met to discuss proposed codes and establish a preliminary codebook to be applied to the entire data set. The transcripts and the code book were then entered into Taguette [16], a software program used to facilitate qualitative coding and analysis. Subsequently, each transcript was coded independently by two of three members of the research team (AA, ELG and AM). AA coded all four transcripts, whereas ELG and AM each coded two transcripts. During the coding process, proposed changes to the codebook were reviewed at team meetings and, if deemed appropriate, integrated into the codebook. Inter-coder reliability was discussed at each meeting and coding discrepancies were reviewed and resolved through consensus. When coding was complete, the study team met to organize the codes into thematic categories [13]. Data collection, coding, and analysis were done in consultation with the qualitative research scientist (RF). The final codebook is shared in supplementary materials (Table S1) and shows all the analytic codes and their respective descriptions.

The results from the analysis of the FGs are presented in a narrative format and supplemented with schematic representations of the coding scheme. To synthesize the results, the obstacles and facilitators to the implementation of the A&F intervention that emerged from the FGs were categorized depending on whether they referred to recipient, feedback, or context variables, based on the framework proposed by the Clinical Performance Feedback Intervention Theory (CP-FIT) [17].

Results

The A&F intervention recruited a total of 142 participants. During the course of the intervention there was 10% drop-out rate. Following the intervention, four FGs with 35 participants were conducted, to elicit their perspectives and opinions about the A&F intervention carried out at two LHDs in Rome, Italy.

Participants

Table 2 shows, for each type of professional involved in the A&F, the number of individuals invited to join the FGs, the number of those who agreed to participate, and the participation rate.

Table 2.

Number of individuals invited in the FGs, number of participants and participation rates

Professionals	Invited n	Participants n	Participation rate %
GPs from LHD2 who were learners and trainees	86	8	9
GPs from LHD13 who were learners and trainees	27	9	33
GPs who served as internal group coordinators	18	11	61
PHPs who served as liaisons between LHD and groups of GPs	9	7	78

The number of participants in each FG ranged from seven to 11. The participation rate varied from a minimum of 9% (8 out of 86 GPs from LHD 2) to a maximum of 78% (7 out of 9 PHPs).

Table 3 presents characteristics of the participants in each FG. Due to the number of GPs who accepted the invitation, two FGs were organized, one for each LHD they belonged to.

Table 3.

Characteristics of participants and focus groups

Participants	N of participants	Age (mean ± sd)	Years of experience (mean ± sd)	Gender (%)	N of patients (mean; total)	LHD (%)	Duration
PHPs	7	40.1(± 6.9)	2.9 ± 1.1	F 57 M 43	-	LHD2 86 LHD13 14	1 h:32 min
GPs coordinators	11	60.9 ± 10.5	28.8 ± 10.5	F 36 M 64	1450 (tot 15967)	LHD2 82 LHD13 18	1 h:24 min
GPs LHD 13	9	54.4 ± 11.5	17.7 ± 10.2	F 33 M 67	1280 (tot 11535)		1 h:10 min
GPs LHD 2	8	56 ± 11.2	17.3 ± 10.7	F 38 M 62	1400 (tot 11190)		1 h:07 min

Abbreviations: LHD Local Health District; GP General Practitioner; PHPs Public Health Physicians; sd standard deviation

Across the four FGs, participants were predominantly male, except for the FG with the PHPs; the majority of professionals belonged to LHD 2. The years of experience were measured by the number of years since commencing clinical practice for the GPs and since starting employment within the LHD for the PHPs. The GPs coordinators were older and had more experience compared to the other GPs.

The duration of the FGs ranged from 1 h 7 min to 1 h 32 min.

Themes

After coding was completed, analytic codes were organized into three thematic categories: barriers, facilitators, and recommendations and suggestions. Barriers and facilitators were grouped depending on whether they referred to recipient, context, or feedback characteristics, according to the CP-FIT classification (Table 4). When discussing the A&F intervention, respondents also shared several recommendations and suggestions that were generally linked to the discussed obstacles and/or facilitators. We reported these recommendations and suggestions separately in Table 5.

Table 4.

Obstacles and facilitators grouped according to the clinical performance feedback intervention theory (CP-FIT) in referred to recipient, feedback, or context

Themes	Obstacle	Facilitator
Recipient variables
Participants’ motivation.	X	X
Perceiving usefulness of participating to A&F.	X	X
Personal characteristics of GPs.	X
Context variables
Time and organizational commitment	X	X
Characteristics of GPs coordinators and PHPs	X	X
The excessive theoretical load of the education and training program	X
Contextual characteristics	X
Heterogeneity in available technology	X
The habit of GPs working as individuals.	X
Logistical and organizational factors.		X
Involvement of recipients in the topic’s identification		X
Social interactions with other professionals		X
Incentives		X
Feedback variables
Credibility of data sources	X	X

Table 5.

Recommendations and suggestions from the A&F experience

• Optimize the theoretical contents regarding to the A&F methodology in short education information concerning the clinical practice. • Increase the time for discussion meetings among GPs, with clinical specialists and with the LHD. • Make A&F part of the daily practice provided for by the Agreement between GP and the NHS. • Involve figures specialized in data collection and analysis. • Compare and discuss the results calculated from the data extracted by the GPs and the indicators calculated from the HIS.

• Make available technologies suited to the audience. • Involve GPs since the identification of the specific topics of the A&F. • Offer financial incentives.

Each barrier, facilitator, recommendation and suggestion is described below, and exemplar quotations translated to English are also shared.

For each group identified by the CP-FIT model (recipient, context, and feedback) we first described the themes that were shared as both obstacles and facilitators, depending on the perspective. Subsequently, we presented the themes that emerged only as obstacles, followed by those that emerged only as facilitators.

Recipient variables

Analysis revealed three themes related to the GP experience with A&F: participants’ motivation to participate in A&F activities; the perceived usefulness of participating in A&F for themselves, their clinical practice, patients, and for the entire system; and the personal characteristics of GPs. Participants’ motivation to deploy A&F and the perceived usefulness of the process were discussed consistently across the focus groups, being framed as both as an obstacle and facilitator to adoption of A&F.

Personal characteristics of GPs emerged only as obstacles. They included different factors, such as age, having poor attitudes or previous experience towards A&F, and low computer skills. No theme emerged solely as a facilitator.

Participants’ motivation

While a lack of motivation to participate in A&F activities was a barrier, motivation to be engaged was a facilitator. Participant considerations focused on the consequences of higher or lower motivation, rather than on its causes or its relation to specific aspects of the intervention. Participants shared that keeping GPs motivated to participate in A&F was important if the program was to continue. Participants also noted that motivation seemed to vary with GP age, and self-motivation was cited as a factor influencing engagement in A&F activities.

Some were very enthusiastic, especially the younger ones, …, and particularly those who were not part of an UCP. (PHP, liaison)

Participants also noted that the presence of less motivated individuals in the A&F session had, in some cases, made group work more challenging and resulted in variable effectiveness. Coordinators shared that they felt it was important for there to be a common desire to learn from the experience within a group.

…I realized that the improvement was due to the efforts of 6–7 colleagues who were already doing well and have improved. But those who started poorly remained the same. (GP, coordinator)

…some individuals were a bit detached, as they were somewhat disconnected from the group. Partial participants. Some were very involved in the matter, while others were a bit more withdrawn… (GP, trainee)

Perceiving usefulness of participating to A&F

Participants expressed their opinions about the usefulness of the actual A&F strategy. They described this aspect either as a potential obstacle when the activity or its particular aspects were perceived as not useful, or as a facilitating factor when the opposite was true. The intervention was deemed useful because participants viewed it as an opportunity for training and professional growth, to improve themselves, to create or increase awareness of their own clinical practice, and to interact with other GPs, both on clinical and practical aspects related to the intervention, specialists, and the LHD. Moreover, participants highlighted how it has been useful for enhancing GPs’ knowledge about data, their awareness on the potential of their practice management software, and on the importance of accurately entering information.

… I think they (GPs) acquired the data literacy, I mean, they had not it as an idea, and instead, they realized the usefulness of what A&F could give… and what they could give, as well. I mean, they could also contribute to the discussion by sharing their data. (PHP, liaison)

… this experience allowed us to reconnect relationships with the LHA, I mean, at the end of the meetings it happened us that they (GPs) stopped to ask us about specific A&F aspects or to discuss data. This was useful. (PHP, liaison)

We are fundamentally as islands, the general practitioners, the family doctor is inclined to be an island, so finding moments of confrontation, finding moments in which it may be possible to share data and analyze them together is a really important thing. (GP, coordinator)

Regarding the data, the intervention was deemed to have enhanced the GPs’ understanding of data-related topics, and the perception of this gain was seen as a facilitator.

What I think they have acquired, though, is data literacy. They didn’t have it as a concept before, but now they realize its usefulness, what it could provide, and what they could contribute themselves. They understand that they can contribute a lot to the discussion with their data. (PHP, liaison)

Moreover, the participants highly valued the opportunity, through participation in A&F, to increase their awareness of their current clinical performance. Also, the training allowed them to understand the potential of their practice management software and learn to utilize its tools to enhance their practice and provide better care for their patients. This awareness was considered a facilitating factor.

When it came to verifying what had been done, and what indicator was achieved, there was a bit of an impact of ‘I was convinced that I was doing better.’ Someone, in fact, told me directly, ‘I was convinced I wouldn’t achieve these results, to perform better even for clinical competence ‘. (PHP, liaison)

…I realized how different the feeling is when faced with reality, in the sense that I realized you are convinced of doing a good job when you realize that some things escape you. (GP, trainee)

Personal characteristics of GPs

Some personal characteristics of each GP were shared as obstacles to participation in A&F activities. Among these were the ways in which GPs use their practice management software.

One of the problems that has arisen, and it’s a methodological issue, is precisely the way of working, obviously working with diagnosis and working without diagnosis. Obviously, colleagues who work without a diagnosis, who therefore use the software basically as a typewriter, to print or send prescriptions, are the ones who have had more difficulties in extracting data and participating (GP, trainee)

Also reported as obstacles were, among personal characteristics, having poor attitudes or previous experience towards A&F, having low computer skills, the GPs’ data “literacy,” and the lack of understanding about the sources of data used to calculate indicators.

Another aspect examined referred to available technologies. As part of the A&F intervention, GPs extracted data to calculate indicators using their personal practice management software. The quality of data entered into the program was often described as poor because this software is primarily used for prescriptions and is not systematically populated with diagnoses and other clinical data. Also, GPs shared that they can choose the type of program to use, leading to heterogeneity in their utilization.

Context variables

Among the context variables, time and organizational commitment and the characteristics of GPs coordinators and PHPs were described in various FGs both as an obstacle and facilitator. Furthermore, it is noteworthy that time and organizational commitment were consistently related to other themes, such as the excessive theoretical load of the education and training program.

Factors related to the context that emerged only among obstacles were the excessive theoretical load of the education and training program, contextual characteristics (for example the fragmented nature of the health services in which they operate or the excessive bureaucratic burden in GPs’ practice), and the heterogeneity in available technology. Additionally, as many GPs work independently, they are not used to the collaborative nature of A&F.

Those described only among facilitators were logistical and organizational factors, working in small groups, alternating in-person and online meetings, and the presence of GPs coordinators and PHPs within the group. Furthermore, participants reported that facilitators included involving recipients in the topic’s identification, the possibility of establishing social interaction with other GPs, clinical specialists, and the LHD/LHA, and incentives.

Time and organizational commitment

Time commitment emerged as a prominent concern, particularly among GPs coordinators and PHPs, who also faced organizational challenges. When asked for spontaneous impressions of the activities, a commonly used word was “demanding,” reflecting the time, organization, and efforts required to maintain high motivation among GPs.

Regarding the logistical commitment, especially us as liaison, we were actually invested in a series of logistical problems not of little importance, bearing in mind that we also carry out other activities already in the LHD. (PHP, liaison)

For most of coordinators, there was a huge amount of work that is not repeatable. I mean, if I were to be involved in another audit project as a coordinator, I would say ‘no, thanks’. (GP, coordinator)

Another term frequently used to describe the A&F experience was “frustrating,” since, despite organizational efforts, the response from GPs was not always adequate.

I would say in some cases it was also a bit frustrating because the organizational commitment, especially on our part, was not always supported on the other side by a great motivation, on the part of the doctors or by significant results in understanding the methodology which was then the cornerstone of the course. (PHP, liaison)

GPs themselves cited “lack of time” as a potential barrier to their participation.

They also complained that they had to follow the clinical activity for which they did not have excessive time to dedicate, so they said, let’s do something, let’s do it small but good. (GP, trainee)

Characteristics of GPs coordinators and PHPs

The characteristics of these figures (GPs coordinators and PHPs) were cited as important elements that could serve as both facilitating factors and barriers, depending on the individual. As reported below, the presence of coordinating figures was seen as a facilitator. However, participants noted that careful consideration must be given to selecting these figures to prevent potential conflicts arising from their lack of recognition by the group. Thus, participants shared that there is also a need to clarify roles properly. Perceiving the mentor or tutor not as an equal but as a figure of command or control was seen as a barrier, whereas viewing them as equals was seen as an advantage.

I mean… At least as far as I’m concerned, I have to say we have coordinators but it’s not like we see it as if we’re subordinate, I mean, we see it on par with us, I mean, of course… we fundamentally see it not as a coach but as a player on the field. (GP, trainee)

Participants discussed how these figures facilitated the activities, particularly if they were well accepted by the groups, with personal characteristics suitable for the role, i.e., with predisposition for dialogue, ability to manage group dynamics, and high computer skills.

The excessive theoretical load of the education and training program

Theoretical workload emerged as an obstacle, characterized by an excessive number of theoretical meetings and lengthy lessons during the experience. This workload was perceived as burdensome and exhausting, making participation challenging. In accordance with the perceived usefulness code, participants found it difficult to understand the usefulness of some theoretical content, further adding to the strain. Conversely, participants acknowledged the initial theoretical training sessions as a facilitating factor but suggested that future sessions should strike a better balance between theory and practice. They advocated for less theoretical information and more practical implications, with a greater emphasis on applying concepts to real-world practice.

They were definitely more interested in the results rather than how to achieve them, not in the whole methodological aspects to achieve the results. The moment they saw the results, much more dialogue was generated, between me and them and among themselves. (PHP, liaison)

Extremely verbose, excessively repetitive, redundant, so it could have been much more concise. We would have still had the same useful stimuli, or we could have refined the theoretical part a bit and perhaps done something more with our patient records, our computers, even among ourselves as colleagues, and engaged in more discussion. (GP, trainee)

Contextual characteristics

Among the contextual characteristics, the perceived fragmentation of healthcare services was highlighted as an obstacle. Additionally, poor communication between different professionals operating in various settings, the efforts required to coordinate patients across silos, regulatory expectations, and the resulting excessive bureaucratic burden were all seen as obstacles.

Unfortunately, we struggle to act as a group, to speak with one voice, for a series of reasons, and, for better or worse, whether a lot or a little or nothing at all, we always suffer from the original sin that the patient is still an important part of income for many of us. So, you’re always caught between a rock and a hard place trying to sort things out, and then you collect, sketch out on a series of aspects. (GP, coordinator)

So, there’s a little hurdle to overcome with the Region, but it would be quite simple to resolve. We just need to have a roundtable where the specialist, the general practitioner, the healthcare institutions, and the Region come together, discuss the issues, and find the solution. (GP, trainee)

Logistical and organizational factors

From an organizational standpoint, having the PHPs as liaison was seen as a strength. Similarly, the role of the GPs coordinator, a GP within the group, was deemed necessary.

Perhaps the strength at this point was the figure of the liaisons who, however, we were already prepared on the topic, so we were able to guide them somehow. (PHP, liaison)

If you have a district that supports you and also a coordinator who helps you, these activities will eventually help you improve, they will help you get out of that bottleneck you’ve ended up in… (GP, trainee)

The choice of working in small groups for the training activities was also reported as a positive theme, as well as the decision to integrate a mixed-method approach for the meetings, alternating between in-person and online sessions.

Social interactions with other professionals

GPs from all the groups recognized that participation in A&F presented the opportunity to establish social interactions among themselves, as they typically work as individuals and with the LHD/LHA and clinical specialists. Indeed, they reported that this experience allowed many to get to know each other.

…and then it also put us in relation. Let’s say a certain number of us, even if aggregated, in the study tend to always be such units. (GP, coordinator)

A positive aspect was the teamwork. Being able to work together among individuals who, objectively, in many cases didn’t even know each other, and this work was also done with the district. It was a collaborative effort, consistently carried out. (GP, trainee)

Continuing Medical Education (CME) credits were awarded to people who participated in A&F activities; however, although this was mentioned as a facilitator, some of the participants did not consider CME credits a sufficient motivation to participate, as it is relatively easy for GPs to reach the required number of credits through other sources.

Feedback variables

Concerning feedback variables, participants focused on the credibility of data sources, that was described as both an obstacle or facilitator. Although it was the only theme that emerged in this category, the credibility of data sources was described as a prominent issue, as it represents a pivotal point to engage GPs in the activities.

Credibility of data sources

Regarding data sources, when participants extracted data firsthand it was considered a facilitating factor as this information was perceived as more reliable and credible. Additionally, the availability of the Electronic Health Record (EHR) was seen as another facilitating factor, enabling the integration of general practice data with other information from the HIS during the A&F. However, participants expressed reservations about the available data, noting HIS data seemed to focus on healthcare finances, not clinical effectiveness or patient experiences. Trust in this type of data is reported as low, also due to the perceived distance between GPs and the collection and processing of this data. In particular, PHPs shared that this distance may also have diminished the credibility of indicators calculated using HIS. Nevertheless, participants acknowledged the potential of HIS data, contingent on greater involvement of GPs, specialists, and the Region.

At the moment, pathology data are only available through Diagnosis-Related Group records and exemptions. However, even exemptions are consistently underestimated because a patient with diabetes, Chronic Obstructive Pulmonary Disease (COPD), or arterial hypertension may be 100% disabled, and therefore, the exemption is related to disability, not the pathologies themselves. (GP, coordinator)

We need to see where this data comes from, ‘consider the source.’ We must ensure that this data is reliable, consistent. If we start from the initial structure, it should be understandable and not absolute, there should be something that also indicates how it was extracted. (GP, trainee)

The data provided by the DEP… the problem is that the collection was very distant then in reality compared to them, that is, it was totally different because maybe the way it was collected is obviously not what the general practitioner directly collects. (PHP, liaison)

Recommendations and suggestions

During the FGs, participants proposed some recommendations and suggestions on how to improve A&F interventions involving GPs.

To address the obstacle related to the excessive theoretical load of the education and training program, many suggested optimizing the theoretical contents regarding the A&F methodology as short messages. All of them appreciated the possibility to have social interactions among GPs and between GPs and other professionals, then recommended that there be more time for discussion among GPs, clinical specialists, and the LHD. To overcome obstacles related to the context in which they work, several participants recommended that A&F be integrated into their daily practice, formally regulated by the Agreement between GP and the NHS.

Regarding the data sources and analysis, many participants appreciated the possibility of integrating also GPs’ firsthand collected data and recommended the training give more emphasis to comparison and discussion of results calculated from these data and those from the HIS. Additionally, some of the participants highlighted the need for involving figures specialized in data collection and analysis for support. The opportunity to integrate general practice data with other health data through the EHR was also seen as a potential avenue for improving A&F by many of them. In this pilot A&F intervention, GPs extracted data from their practice management software, then manually entered them into Excel sheets. Many participants commented on the technologies used: some suggested integrating the use of apps to streamline the process, while others recommended adapting technologies to meet the participants’ skill level.

Some participants shared as a facilitator having involved GPs in the identification of the specific topics (key points of care pathway for patients affected by diabetes type 2 or COPD) of the A&F. They suggested involving them also in identifying diseases that they consider a priority (the pilot intervention was proposed already focused on diabetes type 2 and COPD). Finally, while participants shared that CME credits were a motivator to participation, some of them suggested that financial incentives would be better received as CMEs are easy to obtain.

Discussion

This study captured the experiences and perspectives of PHPs and GPs who participated in a pilot A&F intervention and qualitative data collection that identified barriers, facilitators, and improvement proposals that may influence uptake and effectiveness of A&F among GPs in Italy who are treating chronic conditions.

Conducting distinct FGs for each professional involved at various levels in the A&F activities allowed us to capture different points of views on obstacles and facilitators. Among the recipient variables, in addition to the well-recognized perceived usefulness of A&F in improving quality of care and patient health, our findings highlighted other aspects relevant for GPs. Indeed, the A&F intervention was seen as an opportunity for training and professional growth as well as a facilitator of the interaction with GPs and LHD colleagues. Although this could depend on the nature of our intervention, that included also an education component, it could be interesting to consider these aspects in the design of a traditional A&F interventions.

Most of the themes that emerged concerned context variables. In general, facilitators appear more easily modifiable and implementable than obstacles. Interestingly, facilitators were mainly related to the intervention characteristics, whereas among obstacles there also emerged factors related more generally to characteristics of the setting not specifically linked to A&F, highlighting the need to make more effort at organizational and central levels to remove or mitigate some of these obstacles. This could be particularly relevant in contexts where factors identified as obstacles are predominant, such as those with professionals working alone (for example in some primary care settings), and those facing an excessive bureaucratic load.

Among the feedback variables, only credibility of data emerged as both an obstacle or a facilitator. As expected, credibility was perceived as low when data came from HIS, whereas it was higher when data derived from general practice software. When data did not align with expectations, GPs attributed discrepancies to the way the indicators were calculated from HIS. However, despite these critics on indicators from HIS, GPs recognized the importance of source integration and suggested involving specialized figures in data collection and analysis. According to this, the availability of the Electronic Health Record (EHR) was considered another facilitating factor, enabling the integration of general practice data with information from HIS.

Most of the themes that emerged during this study are consistent with obstacles and facilitators showed in previous research on A&F. Time commitment is one of the most frequently reported barriers in previous studies on quality improvement strategies, including in the GP setting [18]. Indeed, in our intervention, time commitment was particularly impactful given that the intervention was composed of both practical and theoretical sessions, with the latter considered particularly burdensome by GPs. In their 15 recommendations to design practice feedback, Brehaut et al. emphasize that feedback can be more effective if it is perceived as credible by the recipients, and when it’s delivered by a supervisor or a colleague [19]. Although GPs expressed more credibility in data comes from their management software, it was highlighted the difficulty of calculating indicators among the GPs who use the software merely as a typewriter. Taking into account advantages and limits of both sources, it is noteworthy that the GPs shared concerns that there was no effort to integrate HIS data with data collected directly from the practice site. Indeed, participants suggested that integration of practice level data would benefit A&F interventions.

In addition, participants shared a belief that the impact of A&F was influenced by the GPs’ attitudes and motivations to engage in the intervention, which was consistent with what has been seen in prior studies [20]. In our study, low attitudes and motivation of GPs came out mainly in the FG of PHPs, which could be partially due to the high number of GPs enrolled in the intervention (respectively about one third and two thirds of all GPs in the two LHDs). From a public health perspective, the high number of GPs enrolled in the pilot A&F represents a strength of the intervention and demonstrates strong organizational commitment by the LHA. However, this could lead to the enrollment also of a group of GPs less motivated and confident with A&F [6]. Strategies to enhance motivation can be the involvement of the GPs in the developmental phases of the intervention [17] or the organization of group discussion, which was found to be particularly relevant to enhance motivation to change clinical practice subsequent to an A&F intervention [21].

GPs involvement in the choice of the focus of an A&F intervention emerged also among facilitators, and it is an example of a bottom-up approach frequently reported in the literature on A&F as a highly recommended strategy to design effective interventions [22, 23]. In this study, although GPs were involved in selecting the key points of care pathway for patients affected by diabetes mellitus type 2 or COPD, the topics (diseases under evaluation) were already defined in the main research program EASY-NET [5, 24, 25]. Future research may benefit from involving them also in the identification of diseases perceived as priorities. Furthermore, involvement in the choice of topic may be linked to and enhance other enabling factors that emerged also in our FGs, such as improving the perception of intervention usefulness for themselves, for their clinical practice, for patients and for the entire system. Indeed, perceived usefulness, particularly in improving patient care, was described as a factor that can lead to a higher sense of urgency in respect to the intervention [22] and thus motivate the GPs to actively participate. Both the bottom-up approach and the perceived usefulness can explain the low rate of drop-out (10%) during the course of the intervention [6]. Another facilitator that emerged in these FGs was the opportunity to interact with other GPs, specialists, and with the LHD/LHA. Other studies pointed out that A&F can be seen as an opportunity to share ideas and knowledge, and, in the context of a safe organizational culture, to discuss data and performances with organizational leaders [22] and peers [26]. This aspect is noteworthy as A&F can be valuable not only in efforts to improve patient outcomes, but also as organizations attempt to strengthen the collaboration between GPs and LHD. This is particularly useful in a context such as the Italian NHS, wherein GPs are private employees affiliated with the NHS.

Other factors that emerged in these FGs are less described in literature, probably because they are linked to the specific context, recipients, or strategies characteristic of this intervention. Those factors that related to the context in which the intervention was delivered were bureaucratic burden in professional practice, fragmentation of the services operated in, and availability of adequate technologies for the audience. They can be considered as general issues related more to the healthcare system than to A&F and are applicable to various aspects of the Italian general practice setting. Those factors pertaining to recipient characteristics included age, limited past experiences, low computer skills, GPs’ habit of working individually and offering economic incentives. Low computer skills and GPs’ habit of working individually could be, at least partially, related to GPs’ age, which was particularly high (mean 58 years) in our sample, confirming that a deep evaluation of recipient characteristics is relevant during the development phase of an A&F strategy. Excessive theoretical load and a mixed (presence and online) mode of conducting meetings are related to the nature of the specific type of intervention, composed of both theoretical and practice sessions. One of the most mentioned enabling factors introduced in this intervention was the presence of a liaison between LHD and group of GPs (PHP) and an internal group coordinator (GP), a finding in agreement with the existing literature [2]. Their role can act as facilitator at different levels, from the design of the intervention to the establishment of interpersonal relationships, optimizing organizational and logistic factors, strengthening the dialogue with LHD, and demonstrating the commitment of the LHD. Therefore, from this experience, a more structured identification of their role and activities should be considered in the development of future A&F interventions in this setting.

Focus group participants also provided recommendations and suggestions, mostly related to the identified themes. To address time constraints and the excessive theoretical load of the program, for example, participants suggested optimizing the time dedicate to theoretical contents regarding to the A&F methodology by delivering it as short educational information. To improve facilitators, it was proposed to involve GPs since the identification of the specific topics of the A&F. In addition to the content of the comment, spontaneous suggestions highlighted the added value of involving motivated participants in the development of the intervention.

Results of these FGs have implications from both a public health and a research perspective. From a public health perspective, these findings offer evidence on the main obstacles and facilitators to A&F implementation in primary care settings, adding important insight and proposal to design more effective interventions. Indeed, our FGs proved to be a valid and feasible strategy to study physician opinions, barriers, and facilitators to the implementation of quality improvement strategies, encouraging their use also in the design phase of future A&F activities. From a research perspective, obstacles and facilitators that emerged from this study could be considered in the development of surveys or other instruments to evaluate the effect of A&F with other methodologies different from FG. The use of different data collection approaches might be useful to capture a wider range of participant views (for example, surveys or Delphi approaches), providing a more generalizable understanding of the mechanisms underlying the efficacy of A&F interventions. Furthermore, both public health and research implications could be particularly relevant in a context similar to that of the present study, where the recipients of the intervention are self-employed professionals working as independent contractors in the LHA, such as GPs in the Italian NHS. Furthermore, as A&F strategies may also be implemented among other community-based professionals (community nurses) or multidisciplinary teams (physicians, nurses and therapists), FGs should be conducted to examine their experiences and to better understand how the intervention operates within different professional contexts.

Limitation and strengths

This study has both limitations and strengths. One limitation is that the results are based on the personal opinions of professionals who volunteered to participate in the focus groups, likely representing a selected group of highly motivated individuals. This could have led to an underreporting of certain obstacles. Furthermore, it could explain the apparent discrepancies between some findings that emerged from the FGs, such as the perceived usefulness expressed by the participant and the low motivation of some of the GPs enrolled in the intervention but not in the FGs. Secondly, the number of LHD Directors was insufficient to organize a dedicated focus group. On the other hand, the FGs included various other professionals who participated in the activities, which represents a strength. Third, as the theoretical-practical intervention was demanding in terms of lectures and activities, it was not scheduled a structured meeting on the comparison between indicators calculated from HIS and those calculated from GPs management software. Thus, specific consideration regarding this point could not emerge from the FGs. Organizing dedicated FGs for the different roles has a dual strength. First, it allows for collecting comments from homogeneous groups of participants, enhancing the depth and specificity of insights shared. Second, it helps avoid or at least reduce the risk of having participants with varying levels of engagement or comfort expressing their thoughts, which may be influenced by their specific roles. In the present study, to create the most conducive environment for participants and facilitate attendance, FGs were scheduled based on participant preferences and held in familiar and comfortable locations.

Conclusions

From this study, several insights emerged regarding obstacles and facilitating factors perceived by GPs and PHPs working at LHDs who participated in a pilot A&F intervention aimed at improving healthcare for patients with chronic diseases. These findings have been shared with healthcare managers at both LHD and LHA levels to guide the continuation of A&F activities throughout the research period. In addition to informing healthcare managers on how to better design A&F initiatives in Primary Care involving GPs, these findings could also contribute to scientific knowledge by offering insights into optimizing A&F interventions within this specific context.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

A&F

Audit & Feedback

COPD

Chronic Obstructive Pulmonary Disease

CP-FIT

Clinical Performance Feedback Intervention Theory

DEP

Department of Epidemiology

DM2

Type 2 Diabetes Mellitus

EHR

Electronic Health Record

FG

Focus Group

GP

General Practitioner

HIS

Health Information Systems

LHA

Local Health Authority

LHD

Local Health Districts

NHS

National Health Service

PHP

Public Health Physician

UCP

Primary Care Units (Unità di Cure Primarie)

UCCP

Complex Primary Care Units (Unità Complesse di Cure Primarie)

Author contributions

The research team comprised seven researchers: AA and NA were responsible for organization and planning; AA, AM, LA and NA contributed to the development of guiding questions; AA served as the moderator; AM and ELG acted as moderator assistants; AA, AM, and ELG were responsible for defining the codebook, conducting analysis, and reviewing data; AA, AM, ELG, AN, LA, RF, and NA collaborated on writing the article.

Funding

This study was supported by the Italian Ministry of Health (EASY-NET project code: NET-2016-02364191). The funding sources played no role in the study’s design and implementation; data collection, management, analysis, and interpretation; manuscript preparation, review, or approval; nor in the decision to submit the manuscript for publication.

Data availability

The datasets generated and/or analyzed during the current study are not publicly available, as they are currently written only in Italian (the language in which the focus groups were conducted), but are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

The study is part of the EASY-NET research program in Lazio region (Italy), and it was conducted according to the Helsinki Declaration. The Lazio 1 Ethics Committee, Azienda Ospedaliera San Camillo-Forlanini, was notified and ethical approval was waived. Informed consent was collected before initiating audio-recording of the focus groups and interview. Data were analyzed in anonymized form and in accordance with the UE Data Privacy Regulation 2016/679 (“GDPR”) and Italian D.lgs. n. 196/2003, as amended by D.lgs. n. 101/2018.

Consent for publication

Not applicable: The manuscript does not contain individual data.

Competing interests

The authors declare no competing interests.