This paper focuses on between-country differences in clinical decision making in CHD. Doctors' diagnostic and management decisions were compared in three countries (US, UK and Germany) with different health insurance systems. To analyse the country effect on doctors' decisions a factorial experiment with video vignettes was conducted. Working with vignettes is a sophisticated and valid method for measuring the quality of care provided by primary care physicians [17
]. This design was successfully used in previous studies to estimate the un-confounded effects of non-medical factors in clinical decision making [10
]. While this rigorous experimental study permits excellent internal validity, external validity remains a threat. Substantial effort was devoted to produce the video vignettes to be as realistic as possible. However, patient management decisions result from the interaction between physician and patient and from clinical examination. These issues could not be adequately addressed in a videotape-based experiment. Furthermore, doctors may have viewed the interview as a test situation. This could possibly bias the answers in the direction of social acceptability. To avoid this, the doctors were specifically told that the interview is not a test, and that we are interested in their daily work and not in textbook answers.
While there are some methodological limitations, the design has considerable strengths. The factorial experiment allows the estimation of independent and un-confounded country effects as it simultaneously controls for different types of patient (age, gender, social status, and race) and provider influences (gender and length of clinical experience) on clinical decision making. Although there are more physician and patient characteristics, these are considered important for the decision making process. In addition, the experimental approach with videotaped patients offers the possibility to integrate non-verbal signs such as the 'Levine fist'. To enhance generalizability doctors were randomly selected.
Even though all patients reported exactly the same symptoms, results indicate differences between countries, especially between Germany on the one hand and the US and the UK on the other. Minor non-consistent differences were found in patient management between British and American physicians (see also [5
]): British doctors gave lifestyle recommendations regarding alcohol and smoking more often and referred their patients to a cardiologist or other medical specialist more frequently. American doctors were more certain about a CHD diagnosis, they would request more additional information from the patient, and they provided more prescriptions appropriate to CHD. Physicians in Germany showed a significantly different pattern of behaviour. They asked fewer additional questions, diagnosed CHD less often, and were also less certain with their diagnosis compared to UK and US doctors. Among German physicians, a minor proportion would order a CHD specific test and if so, they would order fewer tests. Medications were prescribed to a lesser extent, but patients would see the doctor again sooner. Finally the total number of pieces of lifestyle advice given to the patient and the kind of advice significantly differs from the UK and US. Overall, German physicians would be less active in terms of diagnostic and management strategies. The reasons for this discrepancy may be structural.
In our study physicians were asked how much time they do have for a patient consultation (e.g., for a routine patient consultation, American physicians are allocated on average 18 minutes, British physicians are allocated 10 minutes, and German physicians are allocated 5.5 minutes). Generally, German physicians have the least time for patient consultations compared with UK and especially with US doctors. At the same time, German physicians would like to see the patient again much sooner. Thus, physicians in Germany seem to have the smallest time allocation for a single consultation while they would see the patients in smaller intervals. These different time restrictions might cause varying treatment strategies. Furthermore, there is a relationship between organizational structure and clinical performance [20
]. Differing structures of health care financing and reimbursement are likely to influence clinical decision making. Thus, characteristics of the health care system in each country might be one explanation for the observed country differences.
Although we consider these structural reasons most important, other explanations can't be ruled out. First, there might be cultural reasons for our results indicating that German physicians are less active in terms of diagnostic and management strategies than their US and UK colleagues. For example, physicians' behaviour may reflect different patient expectations. Secondly, there are possible methodological reasons for the different physicians' behaviour in Germany, as data were collected 3 years later than in the other two countries and vignettes in Germany were dubbed. The slightly different way that the videotape was made may have altered physicians' perception of the patient in Germany. However, more than 80% of the German physicians considered the patient on the videotape either very typical or reasonably typical compared to patients they encounter in everyday practice (in the UK and the US this rate was about 90%).