Table 2.

Details of papers identified from the systematic review

Reference	Purpose of study	Methodology and data collection method	Sample and size	Method of analysis	Findings most relevant to this review	Quality including Kmet score and any limitations
Brett et al. (2012)a (AUS)	“To determine the reported frequency of individuals presenting to genetics services after undertaking DTC genetic testing, to explore the opinions and experiences of genetic health professionals regarding DTC genetic testing” (p2) (“and to explore the perceived opinions and experiences of clients who have undertaken DTC genetic testing”.)	Quantitative;online survey	Population: members of the Human Genetics Society of Australia (HGSA) n = 130 (genetic counsellors) n = 38 (clinical geneticists)	Descriptive statistics. Chi-square tests to identify any difference between professional groups. Chi-square tests to identify relationships between firm opinions	Respondents had mixed opinions regarding DTC genetic testing (and a third did not have firm opinions). The majority did not consider DTC testing useful for clients who wanted anonymity, were curious or geographically isolated (but genetic counsellors more likely than clinical geneticists to consider it useful for anonymity or geographical isolation). Only 7 % were confident in accurately interpreting and explaining the results. Eleven percent had had one or more client referred to them following DTC genetic testing.	0.78; this was a well conducted study but would have benefited from some qualitative data concerning the opinions of health professionals. The authors acknowledge that the design of the survey could be improved by the addition of a five-point Likert scale rather than the use of ‘possibly’ responses.
Giovanni et al. (2010) (US)	“To query clinical geneticists about their experience with individuals who consulted them after DTC genetic testing.” (p817)	Quantitative;online survey	Members of the National Society of Genetic Counselors (NSGC), the Adult Genetics Special Interest Group and the American College of Medical Genetics (ACMG)n = 121 genetic counselors n = 6 medical geneticists n = 6 other healthcare providers	Descriptive statistics of questionnaire responses	Over 50 % of the health care providers considered the DTC tests to be useful (but note that of these, 85 % considered that BRCA (breast cancer susceptibility gene) testing was useful). NB. This is only a pilot study, numbers are small, response rate was poor and only 41 % were SNP (single-nucleotide polymorphism) testing (others were single gene). The study also looked at ‘downstream costs of further healthcare referrals’.	0.66; the poor recruitment rate (3.3 %) and strategy (no incentives or reminders following the email invitation) resulted in only 22 eligible participants in the study by Giovanni et al. (2010). However, of these 22, six health professionals had seen patients whose tests were from companies that only provide tests via health professionals, so cannot be considered as DTC tests
Haga et al. (2011) (US)	“To assess physician attitudes and uptake of genomic risk profiling among an ‘early adopter’ practice group.” (p835)”	Quantitative;online survey.	Population: physicians in the MDVIP network n = 157	Descriptive statistics for each question; Fisher’s exact tests for association between two categorical response variables. Regression analysis to identify predictor variables.	The odds of having ordered personal testing were over tenfold higher for those who felt well-informed about genomic risk testing. The odds of ordering testing for their patients were over eight-fold higher among those who had ordered tests for themselves. Clinical utility was a concern for both physicians who had ordered tests and those who had not.	0.83; in the study by Haga et al. (2011), participants were physicians who were part of a collaboration with Navigenics in 2008; these physicians were offered free genomic profiling for themselves and discounted prices for their patients, and were also given access to four online educational modules on genomic risk. We consider that this renders the findings less robust and less generalizable.
Hock et al. (2011) (US)	“To assess genetic counselors’ experience, knowledge, and beliefs about DTC genetic testing ” (p326)	Quantitative; online survey	Population: members of the National Society of Genetic Counselors (NSGC) (US) n = 312	Chi-square and Mann–Whitney U test to assess differences between sample population and the NSGC 2008 Professional Status survey. Comparison between groups re knowledge scores using Welch’s t-tests. Remainder was descriptive statistics. Quotations used to illustrate responses to open-ended questions.	Limited experience of DTC testing: e.g. 81 % stated that the no. of times patients had raised the issue of DTC testing in the last 2 yrs was < 2. Although 75 % had visited a DTC website, 15 % or less suggested it to patients or referred patients. Low level of knowledge of DTC testing-for example, what tests are available; those genetic counselors who worked or consulted for DTC companies scored significantly higher on knowledge section (p < 0.001) Opinions divided on whether genetic testing should be limited to clinical setting, but 56 % agreed that it is an acceptable method if counselling is provided.65 % disagreed that individuals can obtain adequate information on the Internet. More agreed with statements about risk than agreed with statements about benefits of DTC testing.55 % felt they should be knowledgeable about DTC testing, 48 % to be able to interpret DTC test results. Participants expressed the fact that the quality of DTC companies varies, and that their acceptance of DTC services depended on many factors such as availability of counselling.	0.88; a copy of the survey questionnaire was not included in this study, so it was difficult to evaluate whether the results had been fully reported. The authors acknowledge that there was some ambiguity in the design of the survey with reference to the definition of a DTC test and this may have confounded their findings.
Mai et al. (2011) a (GREECE)	To ascertain the views of the general public and physicians on the genetic testing services currently available in Greece; to understand both the wishes and needs of (the general public and) physicians with regard to the genetic testing industry and identify regulatory deficiencies and gaps in the existing legal provision.	Quantitative; cross-sectional survey.	Population: physicians attending a national Greek medical conference in May 2010n = 496	Frequency tables and chi-square tests. Contingency tables to display relationship between two or more variables, using chi-square test. Probabilities of <0.05 considered statistically significant.	Only a small proportion (12.7 %) of physicians was in favour of direct access genetic testing.	0.66 Using the Kmet appraisal criteria, we considered that the study design would have been improved by collecting qualitative data as well as quantitative. The conclusions were not were not fully supported by the findings.
Ohata et al. (2009) (JAPAN)	To identify problems associated with the expansion of ‘new’ genetics service, in particular, direct-to-consumer.	Quantitative; survey	Population: physicians from the Medical Association of Kanagawa Prefecture, members of the Japan Society of Human Genetics and the Japanese Society for Genetic Counseling n = 1145 physicians n = 294 clinical geneticists.	Descriptive statistics and t-tests for associations.	Awareness of DTC genetic testing: 40 % GPs, 70 % clinical geneticists. GPs rated the benefits of DTC significantly more highly than clinical geneticists, and the latter were significantly more concerned about the risks. In particular concerns were re reliability of results, provision of information and counselling, understanding of results.	0.77; similarly to other studies, this study would have benefited from the addition of some qualitative data. The authors acknowledge that because the data was collected two years prior to publication, in the early days of DTC testing, their findings may not be robust.
Powell et al. (2011) (US)	“To assess primary care physicians’ (PCPs’) awareness, experience, opinions and preparedness to answer patients’ questions regarding direct-to-consumer genetic testing." (p1)	Quantitative; survey: choice of online, fax or paper	Population: Members of the North Caroline Medical Society (NCMS) n = 382	Descriptive statistics for personal characteristics and survey responses relating to self-reported awareness, experience with patients, opinion and preparedness. Bivariate associations calculated using cross tabulations, and odds ratios to explore the strength of the associations. Model produced using step-wise forward regression.	38.7 % of respondents were aware of, and 15 % felt prepared to answer questions about DTC tests. Age (50 or older) was a predictor of awareness of DTC testing, and male providers more likely to be prepared to answer questions about DTC. Among respondents who were aware, family practitioners more likely than internists to think DTC testing was clinically useful. 18 % of those who were aware had patients who had questioned them or brought in results from DTC tests.	0.94; a limitation of this study was the low response rate (16.2 %)
Powell et al. (2012) (US)	“To assess the education needs of North Carolina PCPs about direct-to-consumer (DTC) genetic testing. Other aims were to ascertain: (1) PCPs’ preference for delivery of educational materials; (2) barriers to PCPs’ participation in a continuing education program; (3) PCPs’ preference for topics to include in an educational program on DTC genetic testing.” (p1)	Quantitative; survey	As Powell (2011) paper above; same study.	Descriptive statistics for personal characteristics and survey characteristics such as self-reported preparedness, desire to learn more. Bivariate associations calculated using cross tabulations. Step-wise forward regression analysis used to refine model.	85 % of respondents felt unprepared to answer questions about DTC tests (see above). Seventy-four percent wanted to learn about DTC genetic testing. PCPs who felt either unprepared to answer patient questions (p = 0.01) or that DTC testing was clinically useful (p = 0.00) were more likely to want to learn about DTC genetic testing.	0.78; the same limitation as that for Powell et al. (2011) applies.

^aIn these papers, participants included consumers/member of the public as well as health professionals; but in this review, we report only the findings relating to the health professionals