Article Text
PDF
Abstract
Introduction General practitioners often criticise clinical trials for their poor applicability in primary care, which may at least partially explain why their engagement in primary care research remains limited. In order to enhance primary care research, the German government has funded six regional practice based research networks (PBRNs). Within the Bavarian PBRN (BayFoNet), two cluster-randomised pilot trials will be conducted. This paper presents the protocol of the process evaluation accompanying both trials, which aims to explore relevance, feasibility, acceptability and credibility of clinical research in primary care from the perspectives of BayFoNet researchers, general practitioners, and patients.
Methods and analysis The BayFoNet will be established by recruiting general practices (GPs) as prospective research collaborators in two cluster randomised pilot trials. Research teams will provide training in good clinical practice, and support practices in patient recruitment, data collection and documentation. Our process evaluation explores barriers and facilitators in the set up of the BayFoNet PBRN and both cluster randomised pilot trials, under the application of the consolidated framework for implementation research and the theoretical domains framework. In a mixed-methods concept, we will use qualitative and quantitative approaches to evaluate both pilot cluster-randomised trials as well as the BayFoNet itself: focus groups with researchers, semi-structured interviews with general practitioners and questionnaires for patients participating in the pilot cluster-randomised trials at three different time points.
Ethics and dissemination Research ethical approval for this study was granted by the Ethics Committee of the Medical Department, Ludwig-Maximilians-University Munich (AZ 21-1135). Results will be published in international peer-reviewed journals and summaries will be provided to the funders of the study as well as other PBRNs, GP teams and patients.
Trial registration numbers DRKS00028805, NCT05667207.
- PRIMARY CARE
- GENERAL MEDICINE (see Internal Medicine)
- Clinical Trial
- QUALITATIVE RESEARCH
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: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
Strengths and limitations of this study
This comprehensive multi-centre process evaluation will examine all perspectives of the involved key stakeholders in research within primary healthcare.
The longitudinal mixed-methods study design will elaborate information concerning maintenance of study participants over time and development of research processes within primary healthcare.
A multi-professional research team (ie, general practice (GP), pharmacy, nursing, sociology and health service research) will combine quantitative and qualitative methods to provide multiple perspectives and analyses.
Access to GPs and their patients might be challenging.
The generalisability of findings is likely to be compromised by participation bias.
Introduction
Primary care plays a key role in the provision of medical care in healthcare systems globally. However, it faces a number of challenges, which are mainly attributable to an increasing demand due to ageing societies with an increasing prevalence of chronic diseases and multi-morbidity. First, such patients are often under-represented in clinical trials that inform clinical practice guidelines, which limits its applicability in primary care. Second, there is a need for service innovations to meet increasing demands in primary care, which require evaluation. The implication is an urgent need for both clinical and health services research in primary care, for which the engagement of general medical practices is a pre-requisite.
PBRNs are networks of general practices (GPs), which cooperate closely with specific coordinating centres, often academic departments of GP or specific publicly funded research institutes for primary care research.1 They enable primary care practices to conduct studies with complex designs, disseminate results and implement evidence-based strategies in daily clinical practice.2 Practice based research networks (PBRNs) have demonstrated increasingly high-quality research output, but also improvements in the quality of patient care3–5 by ‘shifting efforts from a single disease focus per study to practice systems transformation’.6
In comparison to countries with a longer history of PBRNs (including the UK, the Netherlands, the USA and Australia3–6), the research output from GPs in Germany remains unsatisfactory, which is at least partly attributable to deficits in research infrastructure.1 7 As German GPs work in a market-based, competitive setting of small private practices, they have no protected time or funding for research. Furthermore, they have typically no or little research training or experience. Academic departments of GP mainly depend on public research funding, which has traditionally been scarce.1 A further practical barrier is that GPs in Germany use a wide variety of practice software systems, which are generally poorly suited to support the efficient identification of eligible research participants and data collection.
In order to enhance primary care research, the German government has funded the set up of six regional PBRNs, each with a coordinating university department of GP and research staff to conduct pilot studies. The Bavarian PBRN (BayFoNet) initially comprised four university departments and conducts two pilot cluster-randomised trials within the initial 5-year funding period.8
Aim of the study
This paper presents the protocol for the process evaluation accompanying the set up of the PBRN BayFoNet as well as the implementation of two pilot cluster-randomised trials. The aim is to explore the barriers and facilitators to recruitment and research processes from the perspective of key stakeholders (research teams, GP teams, patients) to identify opportunities for improving PBRNs in Germany and beyond.
Methods/design
Setting
At the start of the funding period, the initial four regional network centres (RNCs) (ie, two departments of GP at two universities in Munich, and one each at the universities of Würzburg and Erlangen) had established working relationships with local GPs. However, the collaboration had focused on teaching of medical students, and none of the RNCs had a formal research network with defined qualification programmes and contractually bound practices. Each RNC had access to clinical trial centres providing methodological support in the design and conduct of clinical trials. The RNC Augsburg became part of BayFoNet as it was founded in November 2022. The main characteristics of the RNCs are described in table 1.
Characteristics of all participating RNCs of BayFoNet
Implementation of the PBRN BayFoNet
Infrastructure of the PBRN BayFoNet
As patient care in German GP is solely carried out in private outpatient practices, academic departments depend on a trustful relationship with local GPs when performing research. Therefore, BayFoNet has to operate at regional level but cooperate closely across Bavaria. Each RNC has its own research team for operational tasks. They are responsible for practice recruitment, ongoing contact and equipment for educational courses, trainings and study material. The coordinating unit (located in Würzburg) is responsible for the overall network activities in BayFoNet and manages any requests to the network, for example, collaboration, patients, publicity and funding.
Recruitment, accreditation and incentives for GP teams
Some of the already collaborating practices had previously participated in primary care research. In the first step, these practices are invited to participate as partners in BayFoNet by invitations letters. In the second step, practices will be invited to participate in the pilot cluster-randomised pilot trials and other projects in each RNC. Invitation to participate in a project will comprise an additional invitation to participate in BayFoNet. Participating in high-quality clinical research can place a heavy burden on GPs. To be accredited, physicians and staff members need to obtain the necessary qualifications in training courses, prepare for studies, adequately inform patients, document according to standards, host monitoring visits, etc. All of these activities compete with the conduct of routine care. In order to establish a sustainable network performing high-quality research, high motivation of practices has to be maintained over years. Financial compensation will primarily be implemented for additional efforts within the studies by paying case-based allowances.
Training of GP teams
We will develop and implement a comprehensive training programme for participating GPs and their teams. The aim is to achieve common levels of research competency in order to enable practice teams to conduct high-quality research including clinical trials and to enable particularly engaged GPs to develop their own research questions. To reach rural GP teams in underserved areas, we will use e-learning facilities.
Data management within the PBRN BayFoNet
So far, there is no publicly available data repository on practice characteristics across RNCs. BayFoNet will develop a central dataset on practice characteristics, research experience and qualification of practice staff as well as on current research activity.
Pilot cluster-randomised trials
Apart from the primary purpose of informing a possible definitive evaluation of intervention effectiveness, the pilot cluster-randomised trials will provide opportunities to identify current weaknesses in the infrastructure to support the implementation of this challenging but important study design in German primary care. Each pilot trial has a coordinating RNC, which will collaborate with the other RNCs in the recruitment and training of GPs, data collection and data management. The latter will use an established software system for electronic data capture including electronic case report forms. All trial procedures and data will be handled according to national and international clinical trial standards. The following paragraphs briefly describe the two pilot trials. Both trials are registered as followed: IMONEDA is registered at the German Register of Clinical Trials, MicUTI is registered at Clinical Trials.gov.
Pilot cluster-randomised trial 1: dipsticks and microscopy to reduce antibiotic use in women’s urinary tract infections (MicUTI)
The aim of the microscopy to reduce antibiotic use in women’s urinary tract infections (MicUTI) is to evaluate the effects of a point-of-care diagnosis and treatment algorithm on antibiotic use in women with symptoms of an uncomplicated urinary tract infection. Twenty GPs affiliated with BayFoNet will be randomly assigned to the intervention arm (women with symptoms of an uncomplicated urinary tract infection will be diagnosed using phase-contrast microscopy and urinary dipsticks) or to the usual care arm. In total, 200 patients should be included and followed up using a patient diary completed until day 7–14 and through telephone calls at day 28 to assess antibiotic prescriptions (number, dose and appropriateness), as well as symptom burden, relapses and recurrence of urinary tract infections, need of re-consultations due to urinary tract infections and the occurrence of upper urinary tract infections.
Pilot cluster-randomised trial 2: implementation of an online education programme for asthma patients in GP (IMONEDA)
The aim of this study is to examine the effectiveness of an online asthma education programme in terms of asthma knowledge, asthma control and unplanned emergency treatment of patients suffering from bronchial asthma in primary care. Twenty GPs affiliated with BayFoNet will be randomly assigned to the intervention arm (patients will have access to the online training) or to the usual care arm. In total, 100 patients should be included and followed up using a patient questionnaire on asthma knowledge, asthma control, patient autonomy, and attitudes and attitudes towards asthma medication after 3 and 6 months.9 10
Design of the process evaluation
In order to elicit the perspectives of key stakeholders (research teams, GPs and patients) on the set up of BayFoNet and the implementation of the pilot trials, we have designed a longitudinal mixed-methods study. Data collection will take place at three time points1: at the point of recruitment (where we also aim to elicit the main reasons for declining an invitation to participate in BayFoNet),2 during the implementation of the pilot cluster-randomised trials and3 after the completion of both pilot cluster-randomised trials.
Applied frameworks to examine barriers and facilitators
In order to elicit barriers and facilitators from the perspectives of research staff and GPs, we will draw on the consolidated framework for implementation research (CFIR). The CFIR provides a conceptual model of implementation drivers across five domains, namely intervention characteristics, inner setting, outer setting, characteristics of the individual and process.11
In order to elicit barriers and facilitators from the perspectives of patients, we will draw on the theoretical domains framework (TDF). This framework assumes that three key drivers (namely motivation, opportunity and capability) determine individual behaviour (such as participation in research). The TDF is very useful to examine domains of behaviour change in individual persons and will be applied to analyse the patients’ perspectives towards participation in pilot cluster-randomised trials.12 Unfortunately, we do not have the possibility to contact patients who refused an active invitation to participate in a pilot cluster-randomised trial. The process evaluation will be conducted between August 2022 and December 2025.
The study protocol follows the reporting guidelines of Standards for Reporting Implementation Studies to differentiate the intervention and implementation strategy of interest13 (online supplemental additional file 1).
Supplemental material
Exploring the perspectives of researchers
The research teams and coordinating unit of the PBRN BayFoNet will meet online for a focus group at three-time points: before, during and after the interventions of both pilot cluster-randomised trials. For logistic reasons, the number of participants may vary between 6 and 15 participants, with at least one participant representing each regional study centre of BayFoNet. An experienced scientist (HK) in qualitative data collection will moderate the discussion and display the previously developed key questions (see table 2). After informed consent of the participants, the discussion will be recorded to support the subsequent transcription.
Applied research questions referring to the CFIR domains concerning the focus groups of researchers and the coordinating unit
Exploring the perspectives of GP
To examine the perspective of the GPs, a convenience sample of GPs will be interviewed until data saturation (we expect 3–4 GPs per regional location; about n=16) before, during and after the interventions of both pilot cluster-randomised trials (see table 3). The semi-structured interviews will be conducted via telephone or web-based video conference, based on the GP’s preference to enable a high degree of feasibility and acceptability in daily practice.
Applied research questions referring to the CFIR domains concerning the interviewed general practices
In addition to qualitative data collection, we will also collect data on practice characteristics. We will examine the practice size and rurality of the GP, as well as experience in teaching, further education of medical students and medical doctors and postgraduate training status of the practice owner. We hypothesise that practice organisational characteristics and adoption will be associated with the level of reach, delivery to the patient and maintenance achieved. Lower levels of these might be associated with lower effectiveness of the delivered interventions. The specific hypotheses to be tested will be based on findings from the interviews with the GPs. Participating and non-participating practices will be compared using organisational information, as far as possible.
Exploring the perspectives of patients
We will conduct a parallel process evaluation, where data will be collected simultaneously to the implementation of both pilot cluster-randomised trials. The paper-based questionnaires will be provided to every enrolled patient (n=300) of both pilot cluster-randomised trials after informed consent by the GP team before, during and after the interventions of both pilot cluster-randomised trial (see table 4). Patients are invited to send their completed questionnaires to the study centre (LMU). The questionnaires for the subsequent time points are sent from this study centre directly.
Applied research questions referring to the theoretical domains framework concerning the patient questionnaires
For this purpose, we have developed a paper-based questionnaire through iterative procedures and discussion between three researchers (LS, TD and AH). These discussions were informed by specific domains of the TDF.12 14 Answers will be provided on a 5-point Likert scale (from ‘1=I do not agree’ to ‘5=I totally agree’) and a descriptive, exploratory data evaluation is planned for data analysis.
Data analysis
The recorded interviews and focus groups will be transcribed using the transcription software ‘F4-audio transcription (Windows)’. Established transcription rules will be applied, which focus on a semantic-content transcript and smoothing of the language.15 In view of the research interest and the data to be generated, the evaluation will be carried out applying a qualitative content analysis according to Kuckartz and Rädiker.16 Especially with larger amounts of text, this rule-based procedure allows a qualitative evaluation, but also opens up possibilities for quantifying partial aspects. Within Kuckartz’s qualitative content analysis, the CFIR framework will be applied. In addition to these deductively obtained categories, there will be the possibility of forming categories inductively in order to make aspects of implementation practice that have not yet been described accessible to theory building. Two different researchers (LS and TD) will read and analyse the data, the third researcher (JG) will solve disagreement. Both the intracoder and the intercoder reliability will be checked, above all to eliminate any ambiguities in the categorisations and thus support the reliability of the analysis. Different stakeholders will be interviewed and analysed independent from each other.
Quantitative data analyses will be performed in IBM SPSS Statistics V.19.0 using descriptive methods.
Patient and public involvement
Patients and/or the public were involved in the conduct and dissemination plans of this research.
Discussion
With our process evaluation study, we aim to gain insights concerning expectations and experiences of all potential stakeholders in primary care research during the development of the PBRN BayFoNet. Both pilot cluster-randomised trials investigate areas of uncertainty about the feasibility of future definitive randomised controlled trials in this setting.17 Our process evaluation will be performed alongside and reported separately. The comprehensive longitudinal mixed-methods study design considers individual-level behaviour change aspects (assessed by the TDF) as well as organisational aspects (assessed by the CFIR) at different time points of a healthcare intervention. We are aware, that both frameworks might address constructs at the individual and collective levels.18 By combining them during data collection and analysis, we aim to define the multi-level nature of behaviour change in healthcare organisations than either of these frameworks alone. We did choose the TDF to understand the patients’ behaviour, whereas the CFIR is a valuable tool to get information about the implemented interventions from the perspective of primary care providers and researchers.
Strengths and limitations
The study design of upcoming interventions and studies will sufficiently benefit from our insights, besides the organisational infrastructure of BayFoNet itself. Our multi-professional research team (ie, GP, pharmacy, nursing, sociology and health service research) will provide multiple perspectives about the research processes within primary healthcare.19 Possible limitations could occur due to different effectiveness in the simultaneous implementation of the cluster-randomised pilot trials and the accompanying process evaluation by the GP. Whether the effectiveness in these practices will be higher or lower cannot be estimated prospectively. A post-hoc analysis will not be sufficient, as the effectiveness will be already known and this trial would be conducted sometime after both pilot cluster-randomised trials were completed. Consequently, we decided to use this prespecified protocol for our process evaluation, as the proposed methods are flexible to unexpected findings in the planned qualitative data collection and analysis will be iterative in nature. Furthermore, it will influence the choice of the actual hypotheses to be tested quantitatively. Another limitation of our process evaluation is that some of the quantitative data might be scarce due to variable access to GPs and their patients. For example, we will not be able to evaluate patients that are not interested in participating in both pilot cluster-randomised trials. Understanding causation will be limited as differences between practices may reflect different attitudes towards medical research in general. This participation bias will compromise the generalisability of our findings.
Next steps
These insights will be shared on a low-threshold level with other regional PBRNs to be able to derive indicators for the successful development of a Germany-wide PBRN in primary care. Based on the data obtained from the process evaluation, recommendations for the development and implementation of clinical studies in German primary care should be developed in accordance with good clinical practice. Furthermore, two sets of indicators for identification of ‘suitable research practices’ as well as criteria for ‘good research practice’ in German primary care are to be derived using a Delphi consensus process involving primary care researchers.20 To this end, we will assemble a list of candidate indicators (eg, appropriate study designs, needed resources, etc to conduct clinical studies in primary care) followed by each expert individually scoring each indicator for importance using a Likert scale. First round results will be discussed again with those experts before second round ratings are placed. In a final face-to-face meeting, the experts will consent the most important indicators and make final recommendations (eg, which types of GPs are needed for a sustainable PBRN and describe features of clinical studies for German primary care to make them relevant, accepted, credible and feasible for healthcare professionals and their patients).
Ethics and dissemination
This study protocol conforms to the Declaration of Helsinki.21 Research ethical approval for this study was granted on 21 February 2022 by the Ethics Committee of the Medical Department, Ludwig-Maximilians-University Munich (AZ 21-1135). Participating general practitioners and their patients receive both verbal and written information explaining the purpose of the study and provide informed consent. Participating general practitioners and their patients receive a minor compensation for answering interview question or for filling the research questionnaires. The interviews with the general practitioners are expected to take about 30–60 min; the research questionnaires for the patients will take 5–10 min to complete. Results will be presented at scientific meetings and published in international peer-reviewed journals. Summaries will be provided to the funders of the study as well as other coordination centres of PBRNs, GP teams and their patients.
Ethics statements
Patient consent for publication
Acknowledgments
The authors thank all study participants for their support. Active and actual members of the BayFoNet research group are (in alphabetical order): Andrea Baumgärtel (TUM), Melanie Bößenecker (Augsburg), Tobias Dreischulte (LMU), Stefanie Eck (TUM), Isabell Endrich (UKW), Maike Ermster (UKW), Ildikó Gágyor (UKW), Jan Gehrmann (TUM), Jochen Gensichen (LMU), Alexander Hapfelmeier (TUM), Susann Hueber (FAU), Merle Klanke (FAU), Christian Kretzschmann (UKW), Thomas Kühlein (FAU), Peter Konstantin Kurotschka (UKW), Klaus Linde (TUM), Klara Lorenz (Augsburg), Marco Roos (Augsburg), Linda Sanftenberg (LMU), Antonius Schneider (TUM), Stefanie Stark (FAU), Til Uebel (UKW), Fabian Walter (Augsburg).
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.
Footnotes
LS and TD are joint first authors.
Twitter @LMU_Allmed, @LMU_Allmed, @LMU_Allmed, @LMU_Allmed, @LMU_Allmed
Contributors LS and TD contributed equally to this paper and have drafted the manuscript and developed the concept of the process evaluation under the supervision of JG. They contributed substantially to the conception of the work and will mainly analyse the resulting data. AH developed the questionnaire for the quantitative data collection as well as the semistructured interview guideline (qualitative data collection) and discussed these materials with LS and TD. HK piloted the semistructured interview guideline and items for our focus groups. She will serve as an expert in qualitative data collection, moderate the discussion sessions and will acquire focus group data. IG and PKK developed and implement MicUTI and revised the manuscript substantially. TK and LB supported the development of the items for the upcoming focus groups and revised the manuscript. SE developed and implement IMONEDA and revised the manuscript substantially. MR contributed to the conception of the study and will support the analysis and interpretation of resulting data for the work, he revised it critically for important intellectual content and approved the final version to be published. All authors have approved the submitted version of the manuscript and have agreed both to be personally accountable for their own contributions and to ensure that questions related to the accuracy or integrity of any part of the work, even ones in which the author was not personally involved, are appropriately investigated, resolved and the resolution documented in the literature.
Funding This study is part of the practice based research network “BayFoNet”. BayFoNet is funded by the German Ministry of Education and Research (01GK1903A-D) 'Development of a sustainable structure for practice based research networks for the strengthening of general practice'. The funding body had no influence in study design, data collection, analysis and interpretation of data as well as in writing the manuscript.
Competing interests None declared.
Patient and public involvement Patients and/or the public were involved in the design, or conduct, or reporting, or dissemination plans of this research. Refer to the Methods section for further details.
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.