We present herein the first results of a questionnaire study regarding the awareness, experiences and views of European clinical geneticists about DTC GT. Moreover, this is, to our knowledge, the only study that specifically focuses on the views of clinical geneticists. Comparison with other published surveys' results of healthcare professionals on this topic [14–17, 19–22] is not very informative due, in part, to the differences in target audiences, and also due to the fact that the actual nature of the questions posed differs a great deal. In our questionnaire, we focused primarily on posing questions based on the specific characteristics of the DTC model of providing GT which make it different from the traditional model of test provision. In particular, we addressed the lack of an established physician-patient relationship, the lack of face-to-face consultation and the offer of genetic tests outside of the traditional healthcare system. Given the great heterogeneity between companies, posing such specific questions allowed us to obtain more detailed information than that acquired from questions referring more generally to the term 'direct-to-consumer genetic testing'.
Our results reveal that clinical geneticists value face-to-face genetic counselling over telephone counselling, however, they do accept the latter under certain conditions within the context of an established physician-patient relationship. Admittedly, in-person genetic counselling represents the traditional and presently most widely used model of providing counselling in a clinical setting. However, other models of provision of counselling are also being incorporated into clinical genetics, including telephone, group and telegenic counselling (via video conference or web-link, which can also be considered face-to-face) . Indeed, the lack of appropriate counselling and provision of information within the DTC GT context is a serious concern [14–17, 19, 20, 22, 29, 30].
Our data also show that European clinical geneticists have a nuanced view with regards to the different types of traits and diseases for which DTC GT are offered. For example, offering a consultation that is not face-to-face was less acceptable to clinicians for conditions with high penetrance than for low penetrance conditions or carrier testing. This is in line with the recommendations for genetic counselling related to GT as developed in the context of the Eurogentest project .
Furthermore, outside of the traditional healthcare system, without a face-to-face consultation, and without an established doctor-patient relationship, almost all clinicians surveyed strongly disagreed with the offer of genetic tests for conditions with serious health repercussions but less than half disagreed with offering genetic tests for conditions with limited health repercussions. This supports the notion of not addressing all different types of DTC GT in the same way. In the report 'A Common Framework of Principals for direct-to-consumer genetic testing services' (2010), the Human Genetics Commission also categorizes genetic tests into different groups, and then specifically makes recommendations for some groups and not others . Some authors  have also suggested that not all DTC GT be treated equally when it comes to attempts to regulate them and address the social and medical needs of consumers. How feasible this is in practice remains to be seen.
Interestingly, our data also offer information regarding how well-known or established different companies are within the European clinical geneticist population. Consequently, this could be a sign of how effective different companies' advertising efforts have been. When we asked clinicians to recall which companies were used by the patients who had purchased tests, few clinicians could recall the companies used, nonetheless, those who could recall the names, most frequently expressed that patients had purchased genetic tests from 23andme, deCODE, GHC and Myriad. Along the same line, when we asked clinicians which companies they could name 'off the top of their heads' the companies most often mentioned by geneticists were 23andme, deCODE and Navigenics. Brett et al. reported that almost half of the 25 patients who purchased DTC GT had purchased the services from 23andme or Counsyl  while Giovanni and co-authors reported that 50% of the testing was purchased from four companies: 23andMe (22.7%), Navigenics (9.1%), DNADirect (9.1%) and Genelex (9.1%) . These results suggest that these are, indeed, the companies that have most successfully advertised their services to consumers as well as to healthcare professionals in the USA, Australia and many European countries.
Our results also reveal that there may be some confusion about the exact meaning of DTC GT. A number of respondents mentioned the names of companies that are not strictly speaking, based on information present on their websites, considered DTC GT companies (that is, not offering nor advertising DTC). An alternate explanation may be that some companies are not necessarily presenting all their services on their websites (that is, not explicitly stating that they sell DTC but will sell tests when presented with requests from consumers). Either way, if this confusion can be generalized to clinical geneticists and counsellors, it could undermine some of the responses of surveys that simply use the term 'direct-to-consumer' GT in their questionnaires without providing a definition for the term.
From our survey, 86% (110/128) of European clinical geneticists reported being aware of DTC GT while 68% (198/291) of clinical geneticists in a Japanese study stated they were aware of DTC GT . The higher level of awareness of our respondents may be due to the fact that Ohata and colleagues conducted their survey three years prior to ours (September and October 2007), at a time that can be considered relatively early in the development of the DTC GT phenomenon. They also posed the question specifically about awareness of particular types of DTC genetic tests: 'DTC genetic testing kits for the classification of predisposition to obesity or the prediction of susceptibility to hypertension/diabetes mellitus/osteoporosis have become commercially available. Do you know about the sale of such test kits?' Therefore, answering in the negative would not necessarily mean not knowing about DTC genetic tests in general. Forty-two percent of the 1880 DocStyles respondents (n=790) were aware of DTC GT services and this difference in awareness may be also due to the fact that this survey was conducted in 2008 but also perhaps due to the fact that none of the healthcare professionals involved in the survey were genetics specialists per se. The increasing awareness among healthcare providers is in line with increasing awareness in the general population .
As mentioned by Brett et al. (2012), the size of the DTC GT market and therefore its impact on clinical services is difficult to assess. Based on our study, 42% and 54% of respondents had at least one patient contact them regarding DTC genetic testing without having purchased and after having purchased testing services respectively. Six percent of the clinical geneticists who responded to Ohata and colleagues survey reported having a consultation with a patient regarding DTC GT . Hock et al. report that 46% of genetic counsellors (146/312) had worked with patients who brought up DTC GT and 14% had patients who wanted to address their DTC GT results . While Giovanni et al.  and Brett et al.  reported that 16.5% (22/133), and 11.3% (19/168), respectively, stated having at least one patient consult them regarding the results from DTC GT companies. About 6% (n= 118) of the respondents of the DocStyles survey reported having at least one patient bringing in DCT GT results to discuss in the past 12 months . Again, our higher percentage may simply be due to the fact that we collected data later, except that the Australian study took place months after ours. The above results are also in line with studies that measured the impact of DTC advertising for hereditary breast cancer testing on genetic services which showed an increase in patient interest in the test as well as an increase in the number of genetic tests ordered by medical professionals [35–37]. Furthermore, five and four of the respondents in our study reported to have been contacted by over 100 patients regarding DTC GT before and after having purchased testing, respectively (Table 3). This is a very high number and we speculate whether this is (partly) due to a confusion on the part of these physicians regarding the exact meaning of DTC GT. Alternately, some countries may have companies that are advertising (if not also selling) some tests DTC and this has resulted in the high number of consultations. This issue is particularly important when trying to assess how much of the resources from the public or traditional healthcare system will be consumed by customers of DTC GT companies in the form of referrals to physicians within the public system.
The response rate of 44% (131/300) is somewhat less than the mean response rate of 56% calculated in a review of surveys conducted with healthcare professionals . The corresponding 131 returned questionnaires did not allow for the power needed to detect subtle differences in subcatergories (that is, differences in responses based on years in practice or based on regional location). Furthermore, we are aware that those clinicians who did respond to our survey may be geneticists who have particularly stronger (negative) opinions towards DTC GT and/or have had more experience with patients requesting information about these services than non-respondents. In this way, our results may not be generalizable to all European clinical geneticists and additional studies looking at non-responders' characteristics should be performed. Moreover, we sent invitations mostly to the directors of clinical genetics institutes whom are likely to be older, and have practised clinical genetics under more traditional models of testing and counselling and therefore may not be as open to newer models. That being said, the average year of birth of respondents is 1957 (SD, 8.1; range, 1941 to 1983) and the average years in practice is 21 years (SD, 8.3; range, 2 to 40 years) hence there is still a good range of years of experience of respondents represented in this study.