The survey and the questionnaire
Within the research activities promoted by the Italian Twin Registry (ITR) [6], a cross-sectional survey by mail questionnaire was conducted on all the Italian twins who had participated to Euroclot, a large international genetic-epidemiological study aimed at identifying genes associated with variations of the end-stage clotting process, and their role in the pathogenesis of stroke http://www.euroclot.eu/.
A questionnaire (Fig. 1) was developed to investigate three dimensions: (i) participants' "understanding" of Euroclot aim and method, (ii) their "attitude" and (iii) "awareness" towards blood donation for research.
(i) Understanding was addressed by two questions about Euroclot aim and method (Fig. 1, items n.1, 2). Response categories referred to the explanations provided in the letter sent for recruitment as well as during the informed consent procedure to the Euroclot participants. In these two occasions the main elements describing aims and methods were provided along with the key terminology reported later in the survey questionnaire (e.g. "physiological" and "pathological" mechanisms, "genetic factors", "environmental factors", "comparison of monozygotic vs dizygotic twins", "clotting process"). As several Euroclot participants thought that twins had a "different physiology" compared to singletons, this idea was used to build the "incorrect" answer choice for the question on method.
(ii) The questionnaire also explored motivations which might lie behind the individual choice to donate a blood sample for the ITR biobank [7]. Three main ideas, drawn from different ethical and anthropological approaches to interpret human behaviour, were conceptualized as motivation and were used for response categories to question n. 3 (Fig. 1, item n. 3). One idea highlights a moral duty to donation for research where the individual believes he/she should find fulfilment by acting for the benefit of others [8]. A second idea relies on a kind of pragmatic attitude where donation is interpreted as a contribution to research advancements in a balanced relationship between participants and researchers [9]. A third idea focuses on spontaneity of donation as an inner attitude of the human beings not culturally driven [10]. The categorization process was also guided by the results of a series of pre-interviews conducted on a subsample of Euroclot participants. Open responses were avoided to skip possible misinterpretations and errors in categorization.
(iii) The last question (Fig. 1, item n. 4) regarded attention and awareness towards blood donation for biobanking. As all the Euroclot participants had agreed to donate additional blood, we considered the recall to have been asked to donate as a proxy for awareness. "I do not know/remember" was included for "understanding" and "awareness" questions (Fig. 1, items n.1, 2 and 4) to avoid that individuals might guess the "correct answer" introducing a bias; the "I do not know/remember" responses were counted as "non-correct" answers.
Other characteristics recorded in the questionnaire were age, educational level, and twin zygosity (assessed by a standard questionnaire regarding physical similarity of the twins) [11]. A final version of the questionnaire was obtained after it was tested on a core of twins (12 pairs) randomly chosen among the Euroclot participants.
Survey Population
We chose to carry out this survey on the Euroclot study participants because, for this research, a thorough person-to-person interaction was performed during information and consent procedures.
In fact, previous investigations on the effects of different interventions to increase participants' understanding, even if almost exclusively on clinical trials, showed that person-to-person interactions between participant and researcher may be more effective than any other method [12].
The Italian Euroclot participants were 181 subjects (89 twin pairs and one triplet). Recruitment for the Euroclot study was performed in February-April 2006 with the following procedure: 135 individuals (age range: 21–24; mean age: 22.7), previously enrolled in the ITR ("ITR-enrolled"), agreed to participate after having received an informative leaflet by mail and having been subsequently contacted by telephone. Further 46 participants (age range: 21–63; mean age: 43.3), not enrolled in the ITR ("volunteers"), asked to participate to Euroclot as a result of the local press and radio advertising campaigns. The Euroclot study had received approval by the Ethical Review Board of the Istituto Superiore di Sanità (Rome, Italy) and Euroclot examinations were conducted from April to October 2006.
Both the "ITR-enrolled" and "volunteer" subjects were asked about family history of stroke and other cardiovascular diseases: no significant differences emerged between the two groups.
In November 2006, the survey questionnaire was sent by mail to the whole sample of Euroclot participants; responses were received up to June 2007.
Ethical procedures for the Euroclot study
All the information about different aspects of the Euroclot study, i.e. aims, twin methodology, clinical/non clinical value of results, donation of blood samples for ITR Biobank, rights of participants/donors including access and withdrawal were illustrated during a face-to face conversation, before the health examination, by two medical doctors and one bioethics adviser. All subjects were explained they would have no immediate benefit from blood donation for future undetermined research, and that they would not be discriminated in case of refusal or withdrawal. It was also clearly explained that the consent regarded exclusively the storage of samples in the biobank and the recording of related data, and not the future re-use of samples and data. Specifically, in agreement with the recent Italian regulation on "Genetic Data Treatment" [13], as a highly conservative option, we structured the consent on the basis that donors would be re-contacted to authorize any further use of their samples (i.e. a new study-specific consent for any single use of the biological material and data). Legal requirements were accomplished to guarantee privacy and confidentiality according to the Italian Law on Personal Data Treatment [14].
Statistics
Frequency distributions of replies were computed.
Multiple logistic regression models were done to evaluate the independent association of gender, age, education, time elapsed between Euroclot examinations and survey responses, and modality of twin recruitment ("ITR-enrolled" and "volunteer" subjects) with each of the investigated dimensions. Response items were dichotomously recoded: for item 1 and 2 (understanding of Euroclot aims and methods), correct answers vs incorrect answers plus "I don't know/remember"; for item 3 (attitude), "pragmatic" answers (b response) vs other answers; for item 4 (awareness), "yes" answers vs "no" plus "I don't know/remember". Robust estimation of standard errors was used to take account of data clustering on twin pairs.
Pairwise twin concordance [C = nc/(nc+nd), nc = number of concordant pairs and nd = number of discordant pairs] for the "pragmatic" attitude to donation (item 3, b response) was estimated in MZ and DZ pairs separately, and interpreted under the assumptions of the classical twin method [15, 16]: a higher concordance in MZ twins, genetically identical, than in DZ twins, who share on average 50% of their genes, suggests that genetic factors may influence the predisposition to the attitude. To this end, the dichotomous version of the item was used. The triplet was treated as three DZ pairs.
All analyses were performed using Stata software (Release 8, 2003).
