12 GPs and 14 psychiatrists took part in the study. Five physicians initially interested in the study did not participate, giving no reason. 54% of the participants were female with the average age of the participants being 54 years. On average, participants had 24 years of work experience. Further details of the demographics of our participants are presented in Table 2.

Table 2 Demographic data of study participants compared to physicians in the whole of Germany [39, 40]

The interviews lasted 30 min on average. GP interviews lasted 32 min on average (between 19 and 53 min) and psychiatrist interviews lasted 27 min (between 20 and 36 min).

The content analysis of all interviews revealed four major thematic domains: (i) Motives and modifying factors for giving or rejecting a psychiatric diagnosis, (ii) Methodological aspects of finding a diagnostic conclusion, (iii) Subjectively perceived diagnostic and therapeutic expectations in the medical system (physician’s perspective) and respective interprofessional cooperation issues, and (iv) Expectations of patients with psychiatric symptoms in the general medicine and psychiatric setting.

In this paper, we concentrate on the results of the first domain to focus on our main research question. In this domain, we identified three main categories in the process of giving or rejecting a diagnosis with five subcategories illustrated in Fig. 1 and detailed with several codes in Table 3.

Fig. 1

Motives and modifying factors that play a role for giving or rejecting psychiatric diagnoses. Model consisting of main and sub-categories of qualitative content analysis

Table 3 Analytical category system of the content analysis including main categories, subcategories and codesMotives for diagnosing psychiatric disorders

Three different main categories could be identified that positively motivated physicians to attribute a psychiatric diagnosis:

Diagnosis as an objective matter

In our inductive analysis process, we defined diagnosis as an objective matter related to the determination of a fact. When asked why diagnoses were allocated, some of the responses addressed its objective nature.

In general, diagnoses serve to categorize patients, i.e. disease patterns, in order to distinguish them from one other. Other physicians mentioned that generating a diagnosis is indispensable for international research and communication.

“The aim is to form groups that are as uniform as possible so that they can be researched in the broadest sense.” P11 (P = Psychiatrist).

Functional and performance-related factors

Functional and performance-related factors refer to different functions resulting from the diagnosis. In this context, it means that the content of the function has an intended mode of action or purpose.

Functional therapeutic factors

Most importantly, the diagnosis was seen as the basis for treatment by some physicians. This means that the diagnosis implicated specific action following guidelines and recommendations for the physicians like medications, illness prescriptions, psychotherapy and much more. Especially the sub-category of diagnosis as a condition for medication was mentioned frequently. In some statements physicians pointed out that there were situations where they were not convinced of the correctness of the diagnosis but still needed it in order to prescribe a certain drug.

“Or he has anxiety attacks that I do not yet consider sufficient to diagnose an anxiety disorder, but feel that I want to give him something for the exam so that he can simply pass it and get over this hurdle, so I make the appropriate diagnosis so that I can prescribe the medication. So that’s a functional diagnosis, if you like, and not a factual one.” P4.

Functional administrative factors

Administrative tasks involve overseeing and organizing personal affairs or those of someone else, typically within a structured setting such as government agencies or organizations, in this case within the health care system as well as with insurance companies.

A related point was the requirement of a diagnosis for billing purposes from health insurance companies or the Ambulatory Health association, which practically forces physicians to make a diagnosis. Many physicians emphasized that they often felt they had to diagnose diseases they were not convinced of because of this billing pressure.

“Because you are really forced by the system to at least commit to one (…) so you can’t write: he came to see me but I don’t know what he has. That means you can’t bill for that.” P14.

“You can write a suspected diagnosis first. But after a quarter, i.e. after three months, the KV demands that you check it and make a confirmed diagnosis or drop it.” P13.

(KV ◊ Association of Statutory Health Insurance Physicians)

In the German healthcare system some diagnoses generate more money for the treating physician than other diagnoses, which can lead to incentivization of specific diagnoses.

“And then there was a figure that implied money per patient per quarter if the patient had certain rather more serious diagnoses. Of course, this has, how shall I say, given a slight distortion to the more severe diagnoses.” P1.

Some GPs emphasized the great influence of administrative needs of insurance companies on allocation of a diagnosis, one example being disability insurance.

“So if someone is unable to work because of a mental illness and this drags on for a certain period of time, then this also forces a diagnosis.” GP2.

However, health insurers also required psychiatric diagnoses for certain services, which led physicians to make them.

“I can’t code it under a flu-like infection or normal exhaustion R53 for example, it wouldn’t get waved through, so it really has to be an adjustment disorder, it has to be, yes exactly, it has to be an F-diagnosis, so that the health insurance company says: “All right, it’s justified, we’ll pay for the psychotherapy”.” GP12.

Information-providing factors

The term “information-providing factors” refers to the social or informative benefit for the recipients of the diagnosis, which is created by passing on the diagnosis and the information it contains. This means that the diagnoses include different treatment or therapy options, which can be helpful when communicating with other colleagues, but also with patients, and can help the patient understand more quickly.

“So if I say: “Someone has schizophrenia”, this is different from me saying to a colleague: “He has severe depression with psychotic symptoms”, for example. I think both have psychotic symptoms, but they have a different status and therefore a different value in treatment.” TNP10.

Another aspect mentioned by the participants was that a psychiatric diagnosis can be a relief not only for the patient but also people around them, such as friends and partners, since the diagnosis may help them understand the patient’s behavior better. As an example, one physician spoke about the relieving effect of a diagnosis for the patient’s relatives, who as a consequence no longer blamed themselves.

Individual factors

Individual factors do not primarily relate to objective facts or specific functions of a diagnosis, but rather to individual framework conditions across situations.

Individual physician-dependent factors

In the interviews, it became apparent that there were also diagnostic styles that differed not only between the two specialist domains, but also within them. Some physicians justified this different way of thinking about and making diagnoses with individual differences of the physicians themselves. As an example, the individual personality of the physician was mentioned. Psychiatrists in particular seemed to have an intrinsic claim to make a diagnosis.

“When I interview someone and want to find out what they have, then for me, somehow the requirement is that I want to have a diagnosis” P5.

Other structural differences, such as the self-image of the physician to be tolerant towards psychiatric diagnoses, also had an impact on making a diagnosis. Other individual factors were experience and expertise, as well as the generation of the physician.

“For example, with young general practitioners (…) I have the feeling that it is different, that they are already more informed about (…) that psychiatric diagnoses are just more of an option.” P2.

The school of training also had an effect. In some universities, for example, some diagnoses were categorically excluded, which led the physicians to continue this practice in their later work.

“Well, I grew up as a purely behavioral psychiatrist. Also because I studied in < city>, where psychiatrists and psychosomatics are, at least on the face of it, mortal enemies. And as a consequence, I diagnose very few of these so-called somatoform disorders.” P11.

One psychiatrist described a situation where other psychiatrists had given a diagnosis of a personality disorder to a difficult patient because they themselves were frustrated with the treatment, and diagnosing a personality disorder offered an excuse for the treatment failure.

“I have also often experienced that specifically borderline disorders, or also narcissistic personality, personality disorders – were given, more as a reaction to the annoyance of dealing with the patient for weeks, that you didn’t get on properly: “Ah, he must have a personality disorder.”” P11.

Motives for not diagnosing

It also became apparent that there were several motives for not attributing diagnoses.

Objective matters

Many participants criticised the inaccuracy of psychiatric diagnoses as they often represent a mixture of norm variants and disease-like conditions. The different classification systems of the ICD and the Diagnostic and Statistical Manual of Mental Disorders (DSM) and their different categorisations were often cited as evidence of the model-like nature of the diagnostic categories. Syndrome diagnoses, which are common in the psychiatric field, were often taken up as a construct.

“Diagnoses are always, especially in psychiatry (…) almost always syndrome diagnoses, which means that there is often a certain vagueness in it and it is always a construct.” P8.

Many physicians mentioned the problem that there are no intermediates between normal and pathological ratings in the coding systems. This led to physicians assigning diagnoses that they themselves doubted were real in order to obtain certain services for patients. Some physicians wanted alternative diagnoses that would better reflect these intermediate states.

“If I think that the patient has a condition that, unfortunately, if it were coded, could be a mental disorder, I mean now like a grief reaction or, you can be in a bad mood, that this is then immediately an F-diagnosis, there is nothing that can be coded so reasonably.” GP12.

Some stated that many clinical pictures to be diagnosed according to the guidelines were often explainable and normal in the individual context. However, the health system would often turn this into a disease by forcing a diagnosis, even though physicians would normally regard this as a normal variant of health.

Functional and performance-related factorsFunctional administrative factors

Some psychiatrists reported that insurance status also influenced whether or not the diagnosis was made. E.g. as patients being privately insured may receive their coded diagnosis with their invoice directly after contacting the physician for billing purposes together with the report on diagnostic findings. Whereas this is not the case for patients with a statutory health insurance.

Many physicians reported that they were cautious when making diagnoses, as the diagnosis could no longer be removed from the insurance companies’ patient records. Access to certain insurances, such as occupational disability or life insurance, could be more difficult or no longer possible. In addition, certain diagnoses can make it more difficult or even impossible to take out a loan or to become a civil servant.

“Then, of course, a confirmed diagnosis from my side would have consequences, not only in terms of stigmatization, but also, for example, for insurance companies or something like that later on. Once something like that is in there, it’s hard to get it out again.” GP2.

Individual factorsIndividual physician-dependent factors

The issue of stigmatization played a major role in the interviews. Physicians from both specialist areas reported that the fear of stigmatization influenced them when making a diagnosis. First, some physicians mentioned that especially the patients were afraid of experiencing stigmatization. At the same time, however, there was also mentioning of potential prejudice by other physicians if they were to read the diagnosis in the patient’s file.

“Let’s say a physiotherapist gets a prescription for physiotherapy and it says: F45.1 chronic pain disorder with psychological stress, then this might prejudice them against the patient.” GP12.

Alternatives to not diagnosing psychiatric disorders

Because of the above-mentioned stigmatization on many levels, alternative ways of dealing with diagnoses were reported in the interviews. Often, for example, a more harmless diagnosis from the psychiatric classification system was chosen instead of the more valid one:

“Then I do try to merely classify it as an adjustment disorder, which is of course also a psychiatric diagnosis, but certainly the least disabling one for someone when it shows up in the health insurance documentation.” P8.

Sometimes, however, the physicians chose a completely different diagnosis in order to guarantee the functional goal, e.g. sick leave.

“If I have a student who comes to me with love sickness, then I tell him quite clearly: “I’ll give you (…) a sick note. But I’ll put down stomach pain or flu.” GP12.