Study design

We conducted semi-structured interviews with GPs, cardiologists and neurologists to explore our research question in depth with the relevant medical specialists. We followed the COmprehensive consolidated criteria for REporting Qualitative research (COREQ) [18]. See S1 Table for the COREQ checklist. The standard of quality for this study was reflected according to Stenfors et al. (2020) [19].

The study was approved by two local ethics committees (University of Marburg, 17 May 2023, ref. 23–94 BO and University of Erlangen-Nürnberg, 12 June 2023, ref. 23–204-Bn). The study was conducted in compliance with the Helsinki Declaration [20]. We obtained formal written consent from all participants.

Study sample

Participants were recruited by two study centres of University of Marburg in Hesse and University of Erlangen in Bavaria, Germany. The recruitment period for this study was 20 June 2023–22 March 2024. Invitations for GPs to participate were sent out via email through the research practice network in Northern- and Middle-Hesse and Northern Bavaria. GPs were also approached and personally invited at institute events (Primary Care Day, Erlangen/Bavaria, Germany). Cardiologists and neurologists in Hesse and Bavaria were invited personally to participate via email, with addresses obtained from the practices’ websites. Cardiologists and neurologists from the local region were initially approached; if no response was received within a few days, recruitment was expanded to a broader geographic area within Northern- and Middle-Hesse and Northern Bavaria. Recruitment efforts were finalised when all participants appeared well-balanced regarding practice setting (region, rural/urban and single/joint), specialty, age and practical experience. We planned to conduct 15–25 interviews to ensure that all three professional groups were represented and that a reasonable range of perspectives and experiences could be captured in proportion to the overall study design.

Interested physicians contacted the study centre in Marburg or Erlangen by phone or email. We used purposive sampling by selecting a comparable number of physicians from both study sites, from rural and urban regions, of all genders and a wide age range. Interviewed physicians received a payment of 100 € as participation compensation.

Data collection

The development of the semi-structured interview guide was informed by existing literature on clinical monitoring [7,8] combined with the perspectives of clinical academics involved in the project (GPs and co-authors). This resulted in a qualitative design grounded in both conceptual and clinical expertise. The interview guide (S2 Table) was developed in a multidisciplinary team (with backgrounds in medicine, health sciences, psychology, and biology) and underwent extended discussion in the departmental qualitative research workgroups. The interview guide was pretested twice by the two interviewers (LR, JH) and two GPs (one was a female GP with several years of work experience who had recently started as a research associate at the Institute of General Practice, University Erlangen; the other was a male GP and professor at the Department of Primary Care, University Marburg).

The main topics of the interview guide were the importance of monitoring in general, monitoring routines of a fictional patient case (either with CHD or after stroke), motivations and reasons for specific monitoring examinations, collaboration with other physicians and patients, and consequences of monitoring (S2 Table). Half of the GPs were interviewed using the CHD case, and the other half using the stroke case. Cardiologists discussed the CHD case, whereas neurologists discussed the stroke case.

Authors JH and LR conducted the interviews, either in person in the physicians’ practices or per video call (https://docs.bigbluebutton.org/ or https://www.zoom.com/). JH and LR were both female research associates with experience in qualitative interviews. JH had a PhD in Biology and 6 years’ experience in primary care research, LR had a MA in Sociology, experience in medical market research and 3 years’ experience in primary care research. The interviewers did not know the participants prior to the study. The interviewers made it clear before the interview that they do not have a vocational education in the medical field and that they might ask more questions where medical knowledge is required. JH conducted the interviews with cardiologists and GPs with focus on patients with CHD; LR conducted the interviews with neurologists and GPs with focus on patients after stroke. Only the interviewer and the interviewee were present at the conversations. Field notes were made directly after the conduction of interviews. Physicians were interviewed until new interviews did not lead to new insights and themes, and the interviewers as well as the research team felt that sufficient data had been collected to answer the study questions. Given the broad study aim, the diversity of participants, and our intention to capture a wide range of perspectives and experiences, we decided to conduct more than the minimum number of interviews, even though the strength of the interview dialogue (supported by well-trained and experienced interviewers) would also have justified a smaller sample [21].

Interviews were audio recorded, pseudonymised and subsequently transcribed verbatim by an external professional transcription service [22]. Demographic data of the participants (gender, age, specialty, vocational experience and practice characteristics) were collected with a custom-made questionnaire.

Data analysis

Data were managed in MAXQDA (2022) [23]. Two authors (JH, LR) coded the interviews independently and discussed the coding process regularly with colleagues from the ChroMo team to receive feedback and resolve disagreements. Analysis was carried out following the reflexive thematic analysis according to Braun and Clarke [24–26]. Benefits of this method is its flexibility and the possibility to obtain a rich and detailed, yet complex account of the material. Despite its absence of a clear and concise guideline, thematic analysis advises to follow six steps [24].

The first step for the researchers was to familiarize themselves with the data by listening to the audio recordings, reading the transcripts and taking notes. Some transcripts were read out loud by two speakers to a group of other experienced qualitative researchers from our Institutes (Institute of Primary Care, University Erlangen and Department of General Practice, University Marburg). Notes were made and discussed in the group. In a second step, relevant text passages were identified and assigned to first codes, then organized in groups. The third step was to summarize and collate the codes to generate higher-level themes. Themes were generated from the data and participants answers (inductive approach) and from the interview guideline questions (deductive approach). The fourth step was to refine themes and build subthemes until the themes outlined/encompassed the entire dataset. At this point, the interviews were read again to ensure that the content of the interviews is reflected by the themes and to include any uncoded passages. In a fifth step, we finalized the title of each theme and subtheme and defined each theme with clear and precise sentences. Several quotations from the interviewees were chosen as examples for each theme and to support the interpretation to our findings (sixth step). We discussed the theme building process at each step within our research group and repeatedly revisited the transcripts and developing themes. This iterative and reflexive approach allowed us to refine our thematic structure and ensure that it remained well supported by the data.

Theme 1: Existing monitoring structures

Monitoring in Germany requires collaboration between different specialised practices, e.g., general practices and independent outpatient cardiologists for additional examinations. Physicians from both institutions reported that they actively shared the care of the patient.

“Patients come to the practice and often bring previous findings from their GP. And we/ If there are no previous findings from the GP, then we ask. If the patient says it hasn’t been done for a long time, hasn’t been checked, then we take care of it after consulting the GP, the check-ups, including cholesterol check-ups or blood pressure check-ups.” (C-15)

Only a minority of patients were referred to specialized practices for monitoring examinations and received care at their GP practice.

“So if I compare how frequent strokes are in the clinic, only a fraction of patients are seen by the neurologist. I would even say, just based on a feeling, maybe 5 to 10 percent of the patient population that’s actually out there.” (N-19)

In addition to a range of regulations, as DMPs and guidelines, monitoring was also determined by the structure of the German healthcare system.

Theme 2: Monitoring concepts has beneficial goals

In general, physicians viewed monitoring as beneficial for the patient and the health system. The goal of monitoring was seen as the early detection of deterioration in order to adjust therapy in a timely manner and thus delay progression of the chronic disease.

“Yes, to recognize the progression of a chronic disease at an early stage. And to be able to intervene if the target values are not reached, to improve them with medication.” (GP-04)

In the long term, physicians considered that monitored patients living with chronic conditions would be less likely to be hospitalised, thereby reducing healthcare costs.

“So for the monitoring of cardiac insufficiency, heart failure, there is good data that patients don’t necessarily live longer, but that the number of hospital stays can be reduced. Which benefits everyone, yes, the patient, the healthcare system, the costs, everything, yes.” (C-14)

Theme 3: Design of ultrasound monitoring for patients with CHD/after stroke

Specialized physicians reported that they carry out the ultrasound monitoring. Some reported that there was no disease-specific monitoring concept specifically for the diseases investigated. Instead, physicians mainly relied on individual clinical judgement rather than standardized, condition-specific processes.

“What if there was a process? Yes, wonderful, right? But I’m afraid that’s one of the areas where the/ let’s put it pathetically: the medical art still plays a role somehow.” (N-20)

Theme 4: Reasons for ultrasound monitoring are primarily control, management and adjustment of therapy

Reasons for ultrasound monitoring were diverse and individual, which led to the overarching theme of the main motivations for conducting these examinations. The factors primarily mentioned were progress control and management of the disease, as well as therapy adjustment if necessary. These factors also influenced the frequency of monitoring examinations.

“The greatest benefit or the primary benefit from the understanding that I have is that I can adjust the treatment in the event of unexpected progression or previously unforeseeable problems or improvements.” (N-20)

“Yes, on the one hand an exclusion of deterioration or progression. That would be essential. Then to document for the future: Such and such were specific issues [...]. And then, of course, to keep reminding the patient: “You had something. You are a certain risk patient. Please pay attention to this and that so that it remains as stable as it is.” (C-14)

Theme 5: Critical view of ultrasound monitoring in CHD and stroke patients

Physicians from all specialties were critical of echocardiography or carotid ultrasound in the care of patients with CHD and after stroke.

“There should be an indication to perform a carotid duplex. And not just like that, everyone has to have it at some point in their lives.” (GP-11)

Theme 6: Ultrasound monitoring usually has no therapeutic consequences

Consequences of echocardiography in CHD patients without further morbidities were seen as rare. The measure was rather indicated for complications such as heart failure, wall motion disorders or valve defects. The consequences of duplex sonography were also considered to be very rare and only to be expected if the stroke was caused by a stenosis of the cervical vessels. Physicians thus perceived limited benefit for the patient. When directly asked about experienced consequences of echocardiography in CHD patients or carotid sonography in stroke patients, physicians generally denied it, emphasizing that such cases were extremely rare.

“No, not in my case, because the cardiologists who are in the area [...] are very conservative.” (GP-02)

“I have to say quite honestly, generally, no, so not in the vast majority of cases.” (C-16)

“I’ve been in private practice for eight years now and it has happened to me exactly once.” (N-18)

Summary

In Germany, physicians from different specialties used different sources for guidance on monitoring patients with chronic conditions. GPs treated their patients along DMPs if available or based on their experience. In contrast, cardiologists and neurologists oriented their care procedures mostly based on guidelines supported by their clinical experience. GPs were seen as managers and gatekeepers in health care with patients being referred to other specialties from their GP. The German monitoring structure required collaboration across facility boundaries, which was characterized by pragmatic respect and understanding, but was also met with critical views regarding lack of coordination and differing treatment goals across specialties.

Monitoring was generally seen as beneficial for the patient and the health system. It was viewed as facilitating the early detection of deterioration, delaying disease progression through adaptation of therapy and leading to less hospitalisations. Physicians did bear in mind the risks of monitoring like overdiagnosis, overuse and overtreatment when examinations are done too frequently. Reasons and motivations for ultrasound monitoring in CHD and stroke patients were diverse and ranged from personal feelings and values, like aiming for the best possible care to own assurance and the patients’ safety to external factors like financial incentives. Despite this range of reasons, ultrasound as regular monitoring procedure was viewed controversially, as it showed insufficient medical benefits for the patient and it was a burden on resources like time and medical personnel. Echocardiography/carotid ultrasound rarely showed therapeutic consequences in everyday life of all interviewed physicians.

Comparison with existing literature

Despite the fact that monitoring in general was seen advantageous by the interviewed physicians, the regular use of echocardiography/carotid ultrasound as part of monitoring routine was criticized, often being not indicated, and rarely resulted in further treatment consequences. Nevertheless, it was described by the participants as an established practice in the monitoring of non-symptomatic CHD and post-stroke patients.

Reasons that justified high levels of monitoring were diverse. In other studies, GPs reported fear and pressure of missing a disease as reasons for overtesting [27–30]. Additionally, GPs would often strive to fulfil patients’ demands for investigations [31], even when being aware that excessive testing can cause harm, e.g., worsen patients’ anxiety or lead to false positive findings [27,28]. The importance of routine examinations is reflected in the language used by both patients and physicians. In our parallel interview study of patients, participants compared monitoring with an annual motor vehicle check, a metaphor that also emerged in our physician interviews and has also been reported elsewhere [17,32]. Patients tend to overestimate intervention benefits, feel safer with more examinations and have trouble understanding overuse [17,33,34]. Furthermore, GPs report about pressure from other healthcare professionals to perform unnecessary tests on asymptomatic patients [27], a factor that we also observe in our interviews.

A major consequence of a high level of monitoring is an increased workload and strain on healthcare resources. While current guidelines provide numerous recommendations to intensify rather than de-intensify routine services [35], it is unrealistic to implement all these recommendations in practice due to time limitations [30,36]. Declining a patient’s request could have significant consequences and GPs report that some longstanding and close doctor-patient relationships have been harmed or even severed [32,37]. Physicians need effective strategies to communicate the necessity of fewer examinations to their patients [32].

The German clinical guideline for chronic CHD introduced a new chapter on monitoring in their 2024 update where no diagnostic intervention (including ergometry and echocardiography) was recommended in asymptomatic patients to assess stenosing CHD [11]. The fact that the echocardiography is part of a cardiology billing complex (that could only be billed as a package) is an example of a wasteful service: it does not deliver recognizable benefits and the incentive misaligns with health care system goals [38].

To address the issue of overdiagnosis, several strategies have been proposed. Increasing public awareness through targeted campaigns and promoting education on the benefits and harms of early diagnosis could help individuals make informed decisions [39]. Additionally, continuous medical education for both students and practitioners are essential to ensure a balanced approach to diagnosis and treatment. The evaluation of diagnostic technologies is crucial for the appropriate administration of increasingly sensitive tests [39]. Identifying targets for de-intensification and synthesizing evidence on unnecessary treatments should be incorporated into clinical guidelines [40]. These guidelines should provide clear recommendations to avoid or discontinue treatments when evidence for their benefit is weak, and potential harms are more likely [9,40]. Furthermore, the concept of “time needed to treat” [36] is often overlooked in guidelines, making the implementation of certain recommendations unrealistic [36]. Political action is also necessary to reform healthcare incentives, shifting the focus from quantity to quality of care [38,39]. Policymakers should foster an environment that prioritizes the provision of appropriate services rather than their sheer volume, for instance, by adopting payment models that promote value-based care across all stages of treatment [38,41].

Strengths and limitations

An important strength of this study was that the interviews were held with physicians from different specializations (GPs, cardiologists, neurologists) who regularly performed monitoring routines to gain more insight into their beliefs and motivations, allowing us to capture perspectives beyond those of a single specialty. Interviewed physicians also had a wide age range and differed in their vocational experiences. Interviews were held in two federal states of Germany to ensure that procedures in the health care systems were not specific for a single state. Another strength was that the interview guide and the results of the analysis were discussed in a multi-professional team (GPs, psychologists and public health scientists that all had experience in qualitative research) not only within the project but also across two institutes of primary care. The self-reflection that the interviewees were willing to share with the interviewers about possibly unnecessary examinations in their own practices and their own personal beliefs and motivations, showed that the physicians did not want to please, but felt free and relaxed to talk. The non-medical background of interviewers probably led to physicians explaining more detailed, e.g., about the billing system and its influence on performing examinations. On the other hand, this could have had a negative effect: as the interviewers were not peers, some nuances of professional reasoning may not have been fully captured.

Further limitations had to be taken into account. Only physicians interested in research took part in the interview study which could have led to a selection bias. Certain character traits, such as being open to new things vs. traditional, or the placement of their doctor-patient relationship were not considered. It was possible that certain types of physicians were not represented and that their motivations for carrying out certain examinations would have offered alternative or differing points of view. Furthermore, physicians from other specialties also could have had additional perspectives. The results offer valuable insights for both clinical practice and health policy, and may help optimize monitoring strategies in CHD and stroke care.