In comprehensive prospective studies, there is always a problem with drop-outs during the course of the study, as was the case for this one. Nevertheless, we believe that the representation within our groups was satisfactory. Complete participation (i.e. responding on all four occasions) of 50 % in the student group and 42 % in the doctor group should be considered sufficiently satisfactory. The actual N value in the multi-level analyses is higher as all those who answered on at least two occasions are included in the correlation matrices (for the student cohort N ranged between 260 and 286). The data are strengthened by the fact that 91 % of participants in the student group and 89 % in the doctor group responded at least on one occasion. A further strength of this study is that it includes complete cohorts of Norwegian students and doctors, rather than random or stratified samples from each group.
The most conspicuous finding was the magnitude of change in individual preferences and choices regarding specialty, which continues long after completion of medical studies. However, at the same time the proportion of doctors who chose surgery, internal medicine, general medicine, or psychiatry as their career paths changed negligibly. This can perhaps be explained by actual demand. However, it appears that the pattern is established almost from the start of medical studies, and long before the students have begun to consider the options for their working lives.
Multi-level analyses of motivation scores show that there are also differences between the medical fields. It appears that men who were particularly concerned about their careers, independence in the work environment, the possibility for undertaking research, and other aspects of the «active» motive type, increasingly turned towards the professional specialties of surgery. Women who were concerned with organising work conditions so they could have a stable family life, and were interested in work with variety and diversity, as well as other aspects of the «integrated» motive type, turned towards psychiatry.
Psychiatry and laboratory-based medicine appear to be «winners» in that more physicians choose these specialities than what could be expected from those who indicated them as preferred specialties at the end of their medical studies or as house doctors. In addition, psychiatry seems to «win» further specialists in the period 4 - 10 years after qualification (the doctor group) - which is perhaps a positive signal when considering future needs. There might also be an association between this and a relatively large number of newly established positions created to strengthen the field of psychiatry. A Canadian study showed that those who chose psychiatry were doctors who were interested in mental health before they began to study medicine and doctors who made their choice through an elimination process based on positive experiences with psychiatry during the course of their studies and during specialisation, as well as negative experiences from other specialties (7). This tendency might also perhaps be associated with the increasing number of female doctors, who generally score significantly lower on the «active» motivation type.
The fields of surgery lose candidates during the physicians'' careers, but this study cannot give a realistic picture of the surgery group due to its heterogeneity. Orthopaedic surgery is one of the specialties with good recruitment (8). From this study, we are unable to analyse the patterns or the prognoses for the individual specialties because the data sets are too small. It is worth noting that within surgery, men tend to move away from this field to a greater extent than women. An American study demonstrated that the choice of a career in surgery was also associated with the extent to which the students were able to participate practically in operations during the course of their studies (9, 10).
It has been shown that medical studies in Norway do not amplify segregation by sex in medical specialty preferences (5). This research indicates that the same applies to the period as intern and the first years as a doctor. Nevertheless, the analyses show that women and men have different preferences and priorities regarding career and choice of specialty. This concurs with results from studies in Finland and Switzerland (11, 12).
Alterations in career preferences generally move towards less prestigious medical specialties than the ones originally considered. Whether the doctors actually become less interested in the more prestigious fields, or the cause for this change is more pragmatic cannot be explained by our data. However, studies from abroad show that this type of retraction from more prestigious fields is associated with the young doctors realising how choice of certain specialties or types of work may result in an extensive strain on their private lives (13) – (15).
Will the new knowledge obtained from this survey have an effect on recruitment to the specialties? The doctors who choose psychiatry have least interest in acute medicine and a good salary and are more interested in obtaining kindergarten places and having a normal family life. It can seem that recruitment to positions of low prestige in remote areas could be improved by taking into account that many young doctors are married to doctors, and that both would like to have both a satisfactory working life and a normal family life. It can also be expected that the increasing number of female doctors will result in a shift towards the more care-related specialties, which has been the case in Finland (11).