Study population and sampling

Eligible were all specialists in internal medicine, who (co-)owned an ambulatory practice in the state of Baden-Wuerttemberg and provided cardiology care. Physicians without a license from the Association of Statutory Health Insurance Physicians (which covers roughly 90% of the population), salaried physicians (neither practice owners nor co-owners) and physicians in hospitals were not included. Participants in the programme were extracted from a list provided by the health insurer, while non-participants were identified through internet databases (for details see [11]).

To ensure diversity of participants and statements, sampling for the qualitative study considered practices as a whole instead of individual physicians within each practice. Therefore, instead of inviting all physicians in a given practice, in every practice eligible for the qualitative study only one registered physician was chosen randomly and invited to participate in the study. Otherwise, a strong overlap of statements and attitudes was to be expected. This resulted in a total of 139 specialists’ practices in the cardiology programme and 84 non-participating practices. All of them were invited. The specialists in the programme received one reminder, the non-participating specialists received two because of the low response rate. All participating physicians provided written informed consent for the telephone interview.

In the quantitative study, contrary to the qualitative approach, instead of one physician, all physicians within each practice were approached, hence the difference in sample sizes. From updated searches and participant lists, we identified a total of 212 medical specialists participating in the cardiology programme and 99 who were not participating. All physicians received a structured questionnaire including an accompanying letter and a post-paid envelope. As the questionnaire was mailed anonymously, completing and returning it was interpreted as consent – a separate declaration of consent was not required here. Contrary to the qualitative study, reminders could not be mailed out directly, as the anonymous approach made it impossible to identify physicians who had not sent back their questionnaire yet. Still, we mailed out one letter to all physicians reminding them of the study in general and asking them to send back their questionnaire if it hadn’t been done already.

Data collection and measures

The qualitative study was conducted between September 2017 and July 2018. After the appointment for the interview had been made, participants received a short, written questionnaire in preparation of the interview and for reasons of sample description. For medical specialists participating in the cardiology programme it included 30 questions on the physicians, the practice they were based in, patient care and the cardiology programme. For medical specialists not participating in the programme it included 23 questions on the physicians, the practice they were based in and patient care.

Telephone interviews then were conducted supported by a semi-structured interview guide, which was tailored to the respective group of physicians. The interview guide addressed key components of the cardiology programme and aspects that emerged during pilot testing. Six topics were covered: a) reasons for (refraining from) participation in the cardiology programme as well as one’s own expectations, b) day-to-day tasks in patient care and changes one noticed since participation in the programme, c) cooperation with other care providers, especially general practitioners, d) how one organised their practices, e) factors that affected implementation of the programme and f) concomitants of the programme. Pilot testing of the guideline was conducted with 3 physicians. Interviews were planned to last 30 to 40 minutes.

The quantitative, structured survey was conducted between October 2018 and January 2019 through the questionnaires mentioned under “Study population and sampling”. These were constructed by the researchers using multiple choice questions, semantic differential scales, open ended questions and dichotomous questions, spanning four topics: a) general information on each physician and their practice, b) participation in the cardiology programme and its implementation, c) healthcare within one’s practice and d) cooperation with other care providers. Questionnaires were pilot tested with physicians and slightly adapted afterwards.

Analyses

All telephone Interviews were transcribed verbatim and analysed using MAXQDA 2018. We followed the method of a content-structuring qualitative content analysis as described by Kuckartz [15]. The code system we used was initially deducted from the research question and subsequently inductively complemented by new insights gained during the interviewing process. Texts were manually scanned for important and possibly relevant statements, which were noted as memos. Subsequently, main categories were developed, derived of the research questions, interview guidelines and emerging from topics mentioned by interviewees. All main categories were applied to parts of the empirical material and concerted between researchers to ensure quality and precision in definitions. Concerted main categories were applied to the whole material and further differentiated. Throughout the application, corresponding sub-categories were created inductively on the basis of statements within the interviews. This led to the final code system which was once again concerted between researchers, applied to all transcripts and interpreted subsequently with respect to the main topics of the study and in consideration of frequencies of mentions by the interviewees.

The quantitative data obtained was tested for normal distribution using the Kolmogorov-Smirnov-Test in SPSS 24. The analysis was largely descriptive with a focus on mean values and standard deviations (sd) (median values and corresponding interquartile range (IQR) where data was not normally distributed) as well as frequencies. For mean and median values, the corresponding n is given in square brackets where data was missing. Frequencies are given in absolute numbers with valid percentages respectively.

Sociodemographic characteristics

Table 1 contains sociodemographic characteristics of all participants in the qualitative study.

Table 2 contains sociodemographic characteristics of all participants in the quantitative study.

Motivation to participate in the cardiology programme

Results of the quantitative study regarding physicians’ motivations to participate in the cardiology programme are shown in Table 3. Physicians were asked to tick a maximum of 3 reasons why they decided to participate.

In the qualitative study, we identified a total of 7 categories with 14 sub-categories in the statements on motivation to participate in the programme. A complete overview on all (sub-)categories can be found in Table 4. For each (sub-)category we included one representative quote, the physicians’ names were replaced by an identification number. Other names, places and potentially unique identifiers were substituted by a place holder in upper case letters put in square brackets. Our additions for a better understandability of several statements were placed in lower case letters and put in square brackets as well. All quotations were translated by us from German to English carefully. Wherever a literal translation proved to be ambiguous, we rephrased it with the aim of conveying its original meaning.

In the subsequent sections, the results of the qualitative and quantitative studies are presented thematically.

Economic incentives

The vast majority mentioned higher reimbursement as compared to regular health care as a motivation to participate in the programme. This is supported by insight gained from the qualitative interviews. These showed that more than 75% of participating physicians deemed economic incentives important, especially financial benefits in terms of higher remuneration compared to regular health care and the discontinuation of caps on patient numbers. Physicians sporadically felt not only remunerated higher, but also perceived more (financial) recognition of their performance.

ID-3: “Because I saw a positive effect for me, regarding patient care and financially as well, of course, because it’s more attractive than the lump-sums I get from the Association of Statutory Health Insurance Physicians and I have more possibilities to check the patient, for example, and so on, and yes. And some things are remunerated, respectively remunerated at all, and remunerated better.”

A further economic aspect was the expectation that accounting would be easier than in regular health care, mentioned by 32.0% of physicians. In the qualitative study, however, this was only mentioned in isolated cases.

Structural incentives

Especially the quantitative study showed that structural and systemic reasons played a role for medical specialists. 66.7% stated that one of their reasons for participation in the cardiology programme was because it is an alternative to regular health care in Germany as it is offered by statutory health insurance for patients who do not participate in managed care programmes or similar types of care. In the qualitative study, isolated physicians brought this topic up as well and perceived the cardiology programme as a paradigm shift away from the current health care system or as being more transparent regarding accounting aspects. An additional structural aspect mentioned in the interviews was, among others, the wish to bind or keep patients who wanted to participate in the programme and were cared for by the practice in the past.

Expected improvements related to health care

Several health care-related expectations were deemed relevant for participation in the cardiology programme by medical specialists in the quantitative study. Especially the diagnostic possibilities available to participants in the programme were named in this regard. Cooperation with general practitioners, time available for individual patients and guideline-orientation of care played a further, yet less prevalent role.

In the qualitative study, reported incentives related to health care were mainly an (expected) optimization of patient-related services through the programme. The physicians justified this expectation with more diagnostic possibilities, a closer monitoring of the patients than in regular health care and higher availability of appointments. Diagnostically the possibility to offer services normally performed by hospitals in an ambulatory setting was highlighted. Such changes in health care within the programme were often derived from the increased remunerations for participating physicians, in the sense of an incentive or possibility to offer more encompassing services because of a higher remuneration.

ID-7: “[…] Furthermore the programme contains services we can only offer there and not for patients insured in the collective contract – be it ambulatory cardioversion, implantation of heart defibrillators, ICD box changes, these are all services suited to be offered ambulatory but normally not remunerated and they were included in the programme and this makes the programme interesting of course.”

Motivation through peers

Motivation to participate because of others’ recommendations played a role as well. The quantitative study showed that this was especially the case when it came to recommendations by the professional association the specialists were members of. Recommendations from colleagues were mentioned more seldom.

In the qualitative study, a more substantial part of physicians named the positive or encouraging attitude of colleagues towards the cardiology programme or their participation as an external incentive to participate themselves. Those who did not report direct influence of their colleagues still reported a majoritarian positive attitude towards the programme in their local medical setting. About a third of physicians perceived their colleagues’ attitudes towards the programme as highly diverse.

ID-1: “[…] And somebody should do it in [PLACE], it’s no good if all of us just wave it aside and see it as too much of an effort and yes, these kind of were the reasons and Mister [NAME] in [PLACE] actually also…motivated me to do it, so I thought: Ok, I’ll try it.”

Researcher: “So this means the opinion of colleagues was decisive for participation.”

ID-1: “Yes.”

Reputational benefits

Expectations related to an influence of participation in the programme on the physicians’ reputation were explicitly surveyed in the qualitative study (hence the high number of mentions) but seldom found – more than 50% of physicians reported that considerations on their own reputation had no effects whatsoever on their decision to participate in the programme.

ID-6: “Reputation is a big word. I would rather say no. […] No, if I had to, with one word, no. It’s definitely true that patients appreciate it [the medical specialist’s programme] but if you would want to call it reputation, it’s rather a positive attribute of a practice, I’d rather not call it reputation.”

Financial motives of physicians and aims of the programme

Even if there seems to be a divergency between the aims of the contract and physicians’ mainly financial motives to participate, this does not automatically imply insufficient implementation or that the programme’s goals cannot be attained: The strong focus physicians put on financial aspects can indicate a function of remuneration as a motivator to optimise one’s own patient care, as international research on incentives for an improvement of medical care quality [16] and integration of care already showed [17], and as the category “Principle: Better reimbursement for better quality” in the short survey mentioned in the introduction indicated [13]. On the other hand, financial incentives need to fulfil several criteria to achieve such effects [18]. Participating physicians named mainly and sometimes solely a higher remuneration as an incentive to participate. Many deemed the quality of their care to be reasonably good and not really influenced by the programme. These perceptions seem to reflect a lack of awareness of the widely prevalent coordination problems in German healthcare from the perspectives of patients and general practitioners [19]. They might lead to participants aiming solely for improvement of their financial situation without changing their own patient care or, in a worst-case scenario, subordinating patient care to financial intentions – a known risk for remuneration schemes in managed care [20]. It has been shown that physicians who receive higher, visit-related remuneration tend to increase the number of patients they see while the time for patients and quality of care decreases [21]. Both are aspects that would be directly opposed to the cardiology programme’s aims.

This situation also points to a possible difference in worlds of thinking: On the one hand there were the developers of the cardiology programme, which aimed at an improvement of quality, efficiency and provision of ambulatory cardiology care. On the other hand, there were the physicians participating in the programme. While the developers of the programme likely saw financial incentives through reimbursement as a necessary mean to attract participants, a good deal of physicians solely focused on this aspect. The developers’ ultimate aim was to improve medical care, the physicians seemed to place the programme largely outside the medical domain instead. A possible reason for this might be the approach of the cardiology programme itself: In recent years, integrated care became increasingly patient-centred with rather holistic approaches, including patients, nurses, social services and other non-medical staff [22, 23, 24, 25]. The cardiology programme however, even though it aims at an optimisation and integration of care in the ambulatory sector, mainly focuses on structuring and enhancing care provided solely by the medical specialist through higher remunerations. Patients and non-medical staff are barely included in this optimisation process. Just like regular health care in Germany, the cardiology programme seems to follow a rather physician-centred approach: Besides their communication with general practitioners, medical specialists in the programme still pretty much work on their own in a more or less unidirectional way towards the patient and do not necessarily need to change their medical daily routine significantly. Hence, from the get-go, the programme might not have been striking to physicians for its medical components, since they were already used to physician-centred care and may rather have appraised the monetary aspects. Here, a broader approach of the cardiology programme, exceeding a purely physician-centred care by including further health care providers and patients, might be a solution to foster integration of care and conquer the risk of participation for purely financial reasons, as the other providers involved would recognise and most likely sanction such behaviour. However, one needs to consider that the German health care system is typically driven by detailed medical and administrative requirements and financial aspects play an important role. Therefore, comprehensive implementation of such an approach seems unlikely at least in the short run. As it has been argued before, this would additionally most likely require a change in physicians’ professional culture beforehand [22].

Still, it remains unclear whether a financially driven participation is in fact problematic here: several physicians in the study explicitly perceived increased remuneration in the cardiology programme as an opportunity to specifically extend and improve their own scope of ambulatory care. A simple claim of financial benefits without resulting changes in patient care is hence not to be expected then. Here, higher remuneration may have influenced patient care positively. Financial incentives were able to make the cardiology programme appealing to some physicians and, at the same time, provided a basis for more regular and tight-knit services, therefore served the purpose of the programme in these cases.

Influence of colleagues

Physicians participating in the cardiology programme reported an influence of their local colleagues’ attitudes on the decision to participate. When it comes to the diffusion and acceptance of innovations, such interrelations between medical practitioners have already been shown in the 1950s works of Coleman et al. [26, 27]. On the other hand, our data did not indicate similar processes for physicians not participating in the programme. This might be a consequence of the differences we observed: Participants perceived rather positive local attitudes towards the programme, while non-participants perceived rather mixed attitudes. The mainly positive attitude in the participants’ environments could have functioned as an incentive for participation. For the non-participants, this aspect might have been absent because of attitudes difficult to assess and therefore suppressing such an external motivation. Another possibility is that the interviewed non-participants might have been less integrated into (informal) social networks with colleagues, which could inhibit such effects when it comes to acceptance of innovations [28] (see also [26]).

Fear for loss of professional autonomy

The statements of several non-participants showed that there were physicians who have been critical of the contractual binding to the health insurer within the cardiology programme. They especially perceived an emergence of unilateral dependence instead of cooperation. This is likely based on the balance of power within the programme: Health insurer and management organisation have a unilateral controlling function over physicians’ implementation of the programme. Meanwhile, the negotiation power of the individual physician is limited, also because the contract is arranged independent of the Associations of Statutory Health Insurance Physicians. Physicians only have indirect influence through the state-wide association of practice-based cardiologists. For some physicians, this constellation led to a fear of misuse of the contractual relationship in terms of interference in their own professional autonomy. Similar aspects have already been looked into in past research [29] and are comprehensible insofar as the cardiology programme involves a lot of requirements and regulations for participating physicians. It is important to ensure a beneficial behaviour of contractual partners and keep up mutual trust. Certain mandatory regulations, however, are unavoidable to coordinate care and set standards for participants – otherwise, there can’t be a basis from which health care could be optimized as intended by the programme.

Limitations

Limitations of our study result from the evaluation project itself: Data collection has been conducted between 2017 and 2019 – for some physicians this was eight years after they entered the cardiology programme. After such a time period it is unclear to what extent there are still memories left from the start of participation and preceding considerations. Unbiased and complete reflection on these aspects is especially questionable when it comes to non-participants who possibly never reconsidered their decision not to join and therefore probably did not pay any further attention to the cardiology programme or medical specialist’s programmes in general.

The number of participants in the quantitative study was substantially higher than in the qualitative study. This came as no surprise, as we assumed that filling out the written questionnaire is less time-consuming than completing an interview via phone. Furthermore, the quantitative study involved less personal information as it provided anonymity to the participants, while the qualitative study used pseudonyms and required written, informed consent as well as appointments for the interviews upfront.

The total response rate in the qualitative study was sufficient to reach a saturation of statements in the group of medical specialists participating in the cardiology programme – during the course of the interviews, certain aspects were repeatedly mentioned by various physicians and no additional topics requiring further interviews were brought up. With 35.5%, the quantitative response rate for specialists in the programme was not overly high, but within the scope of response rates studies with physicians typically achieve – especially considering that all medical specialists in the cardiology programme were approached instead of a sample, so the results cover over one third of the whole population. The response rate in the group of non-participating specialists was lower than the rate of participating specialists for both studies. While in the qualitative study we assumed a saturation of arguments in the group of non-participating specialists as well given that after several interviews no new aspects were mentioned, the response rate of 21,2% in the quantitative study left room for improvement and limited further quantitative comparisons between participants in the programme and non-participants.

Methodologically, the mixed methods approach can be seen as advantageous compared to using a single method. The qualitative data partly served as foundation of the quantitative study. Vice versa, it provided deeper insight into quantitative data in terms of possible explanations of quantitative results. The decision to use telephone interviews in the qualitative study was mainly driven by the fact that physicians normally only have little time and personal interviews had required longer appointments, possibly further reducing participation in the study. Still, personal interviews might have led to different results, especially when it came to rather sensitive topics.

The sample composition posed another challenge: It may be possible that physicians with strong opinions were more likely to participate – be it extremely in favour of the programme or extremely hostile towards it. Participants in the study furthermore might have answered socially desirable – regardless of our assurance of confidentiality and protection of the physicians’ identities it is possible that especially critical considerations were reported understated.