OUP user menu

How Italian health consumers feel about clinical research: a questionnaire survey

Paola Mosconi, Paola Poli, Antonio Giolo, Giovanni Apolone
DOI: http://dx.doi.org/10.1093/eurpub/cki017 372-379 First published online: 13 July 2005

Abstract

Background: Because little information is available about the public's awareness of and attitudes toward clinical research, we planned a survey on a convenience sample of health consumers. Methods: A cross-sectional national survey was carried out using a questionnaire with seven multiple choice questions and two scenarios. A convenience sample of 2000 individuals aged 18 years and older was interviewed in nine out of 21 Italian regions. Results: Sixty-nine per cent reported they were aware of the existence of clinical research and 45% were aware of ethics committees; 29% and 49% indicated they would agree to participate in a prevention study or therapeutic randomized clinical trial, respectively. These percentages decreased when we asked about giving permission for a younger relative. Participants' awareness, opinions and attitudes varied significantly according to socio-demographic and geographical variables. People who were aware of clinical research tended to have a more open attitude toward participation [preventive study: odds ratio (OR) 1.6, 95% confidence interval (CI) 1.3–2.0, P = 0.001; therapeutic study: OR 1.3, 95% CI 1.1–1.6, P = 0.003), even after adjustment for confounding factors. A comparison with an independent and random sample of Italian citizens documented a difference in relevant case-mix factors and a different profile in terms of awareness, opinion and attitudes, which held after statistical adjustment. Conclusions: Health consumers are generally not very aware of clinical research, and their attitudes towards participation seem to be related to the level of awareness. A few variables, such as age, sex, schooling and area of residence, are related to the study questions. These findings may help with the implementation of educational interventions, and underline the need to create meaningful partnerships between health professionals and consumer associations.

  • clinical research
  • national survey
  • patients' attitude
  • patients' participation

Medical progress is largely based on clinical research, and with the emergence of evidence-based medicine great emphasis is placed on findings from clinical trials, as a basis for health-care decisions at either individual or policy level.1,2 The emphasis on the use of randomization to achieve unbiased evidence makes a vital distinction between clinical practice and clinical research, calling for a correct perception of the rationale underlying any experiments involving human subjects, and emphasizes the role and participation of patients in, and their attitude towards, clinical trials.3,4 With the emphasis laid on preventive medicine, large groups from the general population may will be enrolled in clinical trials, and general practitioners (GPs) will play an important role in pharmacological and non-pharmacological prevention studies. The doctor–patient interaction and communication will become more complex and the potential conflict between medical ethics and human rights more acute.

Notwithstanding, there is very little systematic evidence about the public's awareness of these issues.5,6 Most studies have looked at patients with specific conditions, and misconceptions about clinical trials are frequent; the involvement of health consumers in trials is viewed as favourable by researchers but is rarely actually implemented; the level of participation in clinical trials is low, and is often further reduced by factors not related to patients' awareness or willingness.710 Given the lack of information from citizens and health consumers, who are presumably healthier and more risk-adverse than patients, the Mario Negri Institute for Pharmacological Research (www.marionegri.it) and a voluntary association for the defence of consumer rights, ‘Movimento dei Cittadini’ (http://www.space.tin.it/salute/giorighe), started working together in 1999 to stimulate a debate about participation in clinical research. Two surveys were carried out on two samples of Italian health consumers and citizens between 2000 and 2001. Here we report the results of the analysis carried out on the sample of health consumers and discuss the implications of these results by comparing the present findings with those obtained from a sample of citizens that were published elsewhere.11

Methods

The data presented are part of a wider project that has been implemented over a 2-year period in three steps. First, a questionnaire was developed and empirically tested; then a random sample of Italian citizens was identified and interviewed at home; and finally, a convenience sample of health consumers, namely people seeking care at GPs' offices or in outpatient clinics, was interviewed.

Overall, more than 3000 individuals were involved and interviewed throughout Italy. Elsewhere we have reported results from the random sample of Italian citizens.11

Questionnaire

A preliminary version of the questionnaire was assembled using information from the literature. It was then reviewed by experts not directly involved in the study design, tested in a small independent sample of health consumers, then administered twice to a small group of students. Debriefing interviews to test face validity and formally evaluate reliability showed which were the most valid and reliable questions. The final version was divided into two sections, the first regarding awareness and opinions about general issues pertaining to clinical research and the second regarding the attitude towards participation in clinical trials. Briefly, the first section asked two questions about the awareness of clinical trials and ethics committees (‘Have you ever heard about clinical research?’, ‘Have you ever heard about ethics committees?’), two about the basic guarantees required to enter a clinical trial (‘Who should be responsible for the quality of a clinical study?’, ‘What information is essential for participating in a clinical trial?’), two about the proponent (‘Who should be the proposing physician?’, ‘Would you participate in a clinical trial led by your general practitioner?’) and one about why such studies are carried out (‘Why should a physician propose a clinical trial?’). The second section included two scenarios, one describing a prevention study that involves giving a new influenza vaccine to healthy people, and the other a randomized clinical trial testing the efficacy of a new drug for early acute myocardial infarction (see table 4 for details). People were asked to indicate whether they would agree to enter such trials, then were asked to make a choice on behalf of relatives younger than 18 years old. A cover letter was included explaining the aim of the survey and introducing the standard methods used to develop and test new drugs, together with a section asking for standard socio-demographic data such as age, sex, education, area of residence and self-reported health conditions.

Sample and interview

A non-probabilistic convenience sample of people seeking care at a GPs' offices or outpatient clinics was interviewed by trained interviewers. On the basis of official figures from the last census of the Italian population, a total sample of at least 2000 individuals was designed to reproduce the expected proportions of individuals aged 18 years and older according with area of residence (North, Centre or South of the country), sex and age.

In nine (out of 21) Italian regions (four in the North, two in the Centre and three in the South), interviewers were instructed to identify and recruit a fixed quota of eligible cases. They first identified eligible cases in GPs' offices or outpatient clinics, then introduced the survey and asked for participation; people who accepted were asked to complete the questionnaire. Self-administration was recommended, but interviewers were allowed to read or explain questions if necessary. No information was collected about the number of eligible people actually contacted but not included in the survey.

Statistical analysis

All questionnaires collected were checked and validated by the Movimento dei Cittadini coordinating centre in Verona, then analysed at the IRFMN at Milan.

Associations between potential determinants and individuals' answers were estimated using odds ratios (OR) and the corresponding 95% confidence intervals (95% CI). Stratified analysis was used to control for the effect of each of the potential confounders and, when appropriate, Mantel–Haenszel overall and linear χ2-tests for associations were used.12 The frequency of the responses varied because of some missing or non-valid answers, and because in some cases more than one answer was possible.

Results

Sample characteristics

A total of 2000 people were interviewed between May and June 2000 and included in the analysis. When compared with the random sample of Italian citizens, health consumers appeared to be somewhat older, more educated, more frequently located in Northern Italy and with more self-reported health conditions (table 1). These differences are explained by the intrinsic characteristics of the samples (people seeking care versus healthy citizens) and by the methods of recruitment adopted (convenience versus random sampling), as reported fully in Methods, and in a previously published paper.11 The implications of these differences are fully discussed in the Discussion. Most cases (69%) required no help to fill in the questionnaire, and questions were read in 19% and explained in 12% of cases.

View this table:
Table 1

Samples characteristics and comparison with the random sample of Italian citizens

Convenience sample of health consumers (n = 2000)Random sample of Italian citizens (n = 955)χ2, P-value
Sex
    Male48.150.10.9, 0.33
Age
    Mean age (years)46.943.216.1, 0.001
    18–24 years10.28.7
    25–44 years37.241.6
    45–64 years31.234.1
    >65 years21.415.6
Education
    No certificate0.94.155.7, 0.001
    Certificate of primary school12.720.3
    Certificate of secondary school35.834.7
    General certificate of education40.334.0
    Degree10.66.7
Geographical location
    North46.337.8 (45.7)*24.2, 0.001
    Centre19.325.9 (21.2)
    South34.436.3 (33.1)
Self-reported health conditions
    None47.858.6 (56.2)30.1, 0.001
    At least one52.241.4 (43.8)
  • * In the brackets, weighted value in order to reproduce a distribution consistent with that of the Italian population

Awareness and opinions about issues related to clinical trials

Two-thirds of the participants reported they were aware of the existence of clinical research, and 45% of ethics committees. As to which institutions should be in charge of clinical research, most indicated research institutes or university hospital, while only a few indicated pharmaceutical companies or private hospitals.

As to the type of information they would consider essential before entering a clinical study, most indicated the availability of information on the risk–benefit ratio and the presence of a physician as reference during the study; a few considered it essential to know the composition of the ethics committees and required full disclosure about the study sponsor. When asked to indicate the motivations driving physicians to propose a clinical research, most indicated the progress of science and medicine and the patient's recovery or the promotion of citizens' health. Very few named academic or economic reasons. As the proposing physician, most indicated the hospital physician, but when explicitly asked whether they would enter a clinical study led by a GP, 40% said they would (table 2).

View this table:
Table 2

Respondents' awareness and opinions about clinical research, ethics committees and GPs' role

n%
Have you ever heard about clinical research? (1992)a
    Yes137969
    Never43322
    I don't know1809
Have you ever heard about ethics committees? (1960)
    Yes89145
    Never85844
    I don't know21111
Who should be responsible for the quality of a clinical research? (4545)*
    Research institute113625
    University/academic hospital86519
    Ministry of Health75116
    Non-academic hospital64814
    Ethics committee49811
    GP48110
    Pharmaceutical companies1193
    Private hospitals471
What is the essential information for participating in a clinical trial? (4977)*
    Information on advantages and disadvantages158332
    Presence of a physician as reference97720
    Specific insurance57912
    Proposal endorsed by attending physician50510
    Brochures explaining the trial4709
    Informed consent4188
    Knowledge of composition of the ethics committee2725
    Knowledge of the sponsor of the study1733
Why should a physician propose a clinical trial? (4619)*
    Progress of science and medicine142331
    Patients' recovery94620
    Citizens' health promotion87119
    Benefit of pharmaceutical companies57912
    Academic reasons4419
    Economic reasons3598
Who should be the proposing physician? (1965)
    Hospital physician136870
    GP48224
    Other physician1156
Would you participate in a clinical trial led by your GP? (1971)
    Yes79440
    Never57929
    I don't know59830
  • a: Numbers of valid answers are given in parentheses

  • The frequency of responses to individual questions varies because of some missing data, and because in some cases more than one answer was possible (*)

Participants' awareness and opinions varied according to socio-demographic and geographical variables. Table 3 shows the relation between selected respondents' answers and the main variables: the probability of being aware of clinical research and ethics committees appeared to be related to age, schooling and area of residence: elderly, less educated people and those living in Southern Italy were somewhat less aware of these issues. On the question about the role of GPs, young, more educated people, those living in Southern Italy and those reporting no major health problems were less likely to indicate GPs as potential leaders of clinical studies.

View this table:
Table 3

Relation between respondents' awareness and opinions and a selected list of variables [OR are estimated using the group in bold type as reference category (OR 1)]

OR95% CI
Awareness of clinical research
    Sex
        Male versus female1.0490.867–1.270
    Age*
        <65 versus >65 years0.6350.504–0.801
    Schooling*
        <5 years1
        5–8 years1.3380.979–1.830
        >8 years2.7202.073–3.569
    Area of residence*
        North1
        Centre1.0450.799–1.366
        South0.6490.525–0.801
    Self-reported health conditions
        None versus at least one0.9370.774–1.134
Awareness of ethics committees
    Sex
        Male versus female0.8990.752–1.075
    Age*
        <65 versus >65 years0.6750.536–0.851
    Schooling*
        <5 years1
        5–8 years1.1280.808–1.574
        >8 years2.2601.701–3.003
    Area of residence
        North1
        Centre1.2881.012–1.640
        South0.9430.772–1.152
    Self-reported health conditions
        None versus at least one0.9460.791–1.130
Opinion on GPs' role
    Sex
        Male versus female0.8490.709–1.017
    Age*
        <65 versus >65 years1.4851.185–1.862
    Schooling*
        <5 years1
        5–8 years0.5654.412–0.774
        >8 years0.5430.413–0.713
    Area of residence*
        North1
        Centre1.1890.935–1.513
        South0.6190.503–0.761
    Self-reported health conditions*
        None versus at least one1.6461.373–1.974
  • Schooling: for this analysis education was classified in three levels: <5 years, 5–8 years and >8 years

  • * P-value from Mantel–Haenszel χ2-test, P < 0.05

Attitude towards participation in clinical trials

The attitude toward participation in clinical trials was not favourable for the prevention trial (table 4): 49% indicated they would agree to participate in the therapeutic randomized clinical trial, but only 29% agreed to the vaccination trial. When asked to state whether they would permit the participation of a younger relative in the same studies, the proportion of people willing was lower, particularly for the vaccination study, with only 5% saying they would agree. A few variables were associated with the likehood of a positive attitude towards participation (table 5): female, less educated people and those living in Southern Italy took a less positive view of the prevention trial. A different pattern was observed in the therapeutic scenario. People living in Southern Italy were more likely to participate personally and were also more likely to allow the participation of a younger relative.

View this table:
Table 4

Respondents' attitudes toward participation in clinical trials

The first scenario was phrased as follows:
Outbreaks of influenza occur every winter and have a huge impacts on society and the individual. One way to reduce the impact of the illness is to consider vaccination.
Some researchers are studying a new vaccine which must now be tested on a group of healthy voluntaries. The results of this phase allow will help show whether this intervention is a safe and effective method meriting further investigation.
Would you personally participate in a clinical trial as a healthy volunteer? (1976)a
    Yes29%
    Never43%
    I don't know28%
Would you give your consent for a minor to participate in a clinical trial as a healthy volunteer? (1977)
    Yes5%
    Never78%
    I don't know17%
The second scenario was phrased as follows:
When there is a doubt that a new treatment seems better than the standard treatment, a ‘randomized clinical trial’ is needed. In this case the choice of treatment (the new or the standard) is based on a random allocation, meaning that neither the physician nor the patient can choose, and the treatment is allocated by chance.
About 15 years ago, the mortality of people with early acute myocardial infarction was 13%. Preliminary results with a new drug, streptokinase, suggested that this mortality could be reduced. A group of researchers and cardiologists, after careful analysis of all available information, decided to organize a ‘randomized clinical trial’ to test the efficacy of streptokinase. Patients were randomly allocated to receive the new drug or the standard treatment: 11 806 patients in 176 coronary care units were enrolled over 17 months and followed over time. At the end of the study mortality was 20% lower in the group of patients receiving streptokinase. Now this treatment is routinely administered for acute myocardial infarction.
Would you personally participate in a clinical trial for a new drug? (1962)
    Yes46%
    Never28%
    I don't know26%
Would you give your consent for a minor to participate in a clinical trial for a new drug? (1965)
    Yes23%
    Never46%
    I don't know30%
  • a: Number of valid answers given in parentheses

View this table:
Table 5

Relation between respondent's attitudes towards participation in clinical trials and a selected list of variables [OR are estimated using the group in bold type as reference category (OR 1)]

OR95% CI
Vaccination trial, adults
    Sex*
        Male versus female0.6420.528–0.781
    Age
        <65 versus >65 years0.9050.704–1.164
    Schooling*
        <5 years1
        5–8 years1.2940.900–1.860
        >8 years1.4841.081–2.037
    Area of residence*
        North1
        Centre1.0950.847–1.416
        South0.7420.593–0.928
    Self-reported health conditions
        None versus At least one1.0760.885–1.308
Vaccination trial, minors
    Sex
        Male versus female0.7200.487–1.064
    Age
        <65 versus >65 years0.6810.391–1.186
    Schooling
        <5 years1
        5–8 years0.8970.450–1.787
        >8 years0.9700.537–1.749
    Area of residence
        North1
        Centre1.0830.638–1.838
        South1.1260.728–1.743
    Self-reported health conditions
        None versus at least one0.7900.535–1.166
Randomized therapeutic trial, adults
    Sex*
        Male versus female0.8020.671–0.958
    Age
        <65 versus >65 years0.8200.653–1.031
    Schooling
        <5 years1
        5–8 years1.1980.873–1.645
        >8 years1.2230.928–1.612
    Area of residence*
        North1
        Centre1.0260.804–1.310
        South1.5441.265–1.885
    Self-reported health conditions
        None versus at least one0.9600.804–1.147
Randomized therapeutic trial, minors
    Sex
        Male versus female0.8680.704–1.070
    Age
        <65 versus >65 years0.4980.367–0.675
    Schooling
        <5 years1
        5–8 years1.4650.998–2.152
        >8 years1.2960.920–1.825
    Area of residence*
       North1
        Centre0.9360.683–1.257
        South1.7611.401–2.215
    Self-reported health conditions
        None versus at least one0.8540.693–1.052
  • Schooling: for this analysis education was classified in three levels: <5 years, 5–8 years and >8 years

  • * P-value from Mantel–Haenszel χ2-test, P < 0.05

Subgroup analysis and comparison with Italian citizens

Sex, age and schooling were related to participants' answers and to their area of residence. Thus, to control for the imbalance due to socio-demographic factors, we carried out a stratified analysis to evaluate the strength and statistical significance of the associations between area of residence and individuals' answers, controlling for other confounders. As the North and Centre gave similar estimates, for the purpose of the present analysis we classified this variable as follows: North and Centre versus Southern Italy. Even after adjustment for sex, age and schooling (variables describing the comorbidity status did not change the results), people living in Southern Italy were significantly less likely to be aware of clinical research-related issues (OR 0.63, 95% CI 0.52–0.78), and were less favourable towards participating in clinical studies where a GP was the proponent or principal investigator (OR 0.61, 95% CI 0.49–0.75). They were also less likely to enter a vaccination trial (OR 0.72, 95% CI 0.58–0.89), but more willing to take personally part in a therapeutic trial (OR 1.53, 95% CI 1.27–1.85 ) and to give consent for a minor to participate (OR 1.75, 95% CI 1.41–2.17).

When we compared this sample of health consumers with the citizens,11 significant differences emerged both in terms of potential determinants (table 1) and for all but one of the seven questions considered (table 6). In order to control for the unbalance due to some confounders (education, age, comorbidities and geographical location) we performed a stratified analysis: health consumers from the convenience sample appeared to be more willing to participate in a clinical research lead by GP and to personally participate in a clinical research as a healthy volunteer, whereas citizens from the random sample were more likely to participate in, or to give consent for a minor to participate in, clinic research for a new drug.

View this table:
Table 6

Awareness, opinions and attitude of convenience sample of health consumers versus random sample of Italian citizen (Italian citizens, OR 1)

Unadjusted estimatesAdjusted estimatesa
ORP-valueOR95% CIP-value
Have you ever heard about clinical research?1.230.0121.030.87–1.230.70
Have you ever heard about ethics committees?1.160.071.020.86–1.190.84
Would you participate at a clinical research lead by your GP?21.240.0081.211.02–1.420.025
Would you personally participate in a clinical research as health voluntary (for example in a study testing new influenza vaccine)?1.690.0011.571.29–1.900.001
Would you give your consent for a minor to participate in a clinical research as health voluntary (for example, in a study testing new influenza vaccine)?1.180.3551.230.84–1.760.277
Would you personally participate in a clinical research for a new drug (for example, in a randomized study testing a new drug for acute myocardial infarction)?0.760.010.720.61–0.840.001
Would you give your consent for a minor to participate in a clinical research for a new drug (for example, in a randomized study testing a new drug for acute myocardial infarction)?0.810.0210.790.66–0.950.01
  • a: Adjusted for education, self-reported health conditions and geographical location

Discussion

Community awareness about the need for clinical research and the manner in which it is conducted is important, as there appears to be a relation between knowledge/awareness and willingness to participate.3 Most research on the nature of this association has been conducted in English-speaking countries, on samples of specific diseases8,9,13,14 or in samples of people from individual metropolitan areas.15,16 Empirical data from national surveys are very rare indeed, the only example we found in the recent literature being focused on cancer-related issues.10

Healthy individuals and patients do not have the same perception and attitudes, healthy people being more risk-adverse than patients, and thus have a less favourable attitude towards research.17 In addition, awareness and attitude tend to differ across a country.6 We therefore carried out two surveys using the same questionnaire to establish the Italian public's awareness about clinical research and to identify any points that might help the implementation of specific interventions. The previous study on a random sample of 1000 Italian citizens suggested that society at large does not have a favourable attitude towards clinical research (slightly more than half said that they would participate in a therapeutic clinical trial and only about 20% indicated they would participate in a vaccination trial); a few socio-demographic and geographical variables were associated with both the awareness and the attitude toward participation, with people living in Southern Italy having a different view of clinical research.6,11

The present survey on a convenience sample of people seeking health care at GP or outclinic levels shows that, in general, this non-random sample of Italian health consumers is quite different from the sample of citizens in terms of case-mix, but has limited awareness of issues pertaining to clinical research. It also had scant enthusiasm about participation in clinical trials, particularly for the prevention trial that required the participation of healthy minors. Some important determinants appeared to influence people's awareness, opinions and attitude, such as age, education and area of residence. Differences according to area of residence remained after adjustment for socio-demographic factors and must be kept in mind when planning interventions to change awareness or nationwide clinical studies. In addition, people who were more aware of the existence of clinical research were somewhat more likely to indicate they would enter the clinical trials outlined in the two scenarios, even after taking into account the confounding effect of the main variables. Finally, after adjusting for age, education, self-reported health conditions and geographical location, health consumers showed significative differences in all but one the seven questions considered.

This survey has some limitations. First, the study design and the sampling strategy gave rise to oversampling of health consumers with a higher than expected level of schooling. This was probably due to a selection bias introduced by the recruitment strategy: interviewers in GPs' offices or outpatient clinics contacted all the consecutive people who matched the characteristics required (i.e. age, sex and area of residence) until they reached the number needed, and self-selection presumably meant that more educated people were more willing to participate. This might have introduced a distortion, with an overestimate of the level of awareness reported by the sample present findings that might not be generalized to health consumers as a whole. Secondly, in 12% of cases interviewers helped complete the questionnaire, and might just have influenced the answers, particularly since elderly or less educated people were those actually requiring more assistance. Finally, the specific scenarios we used to assess peoples' attitudes could have influenced the results, which therefore cannot be considered general indicators of people's attitudes. In other words, the use of other examples could have produced different results.

Despite these limitations, we believe that these factors did not seriously impair the results, for the following reasons. First, if a selection bias did distort the results it probably did not change the direction of the associations found, as results in the companion survey in a random sample of Italian citizens results were very similar.11 Secondly, we carried out further analyses to test whether people receiving help answered in systematically different ways: we checked the distribution of the help across the main categories, and then introduced this variable into the model. The findings did not change at all.

Finally, although we were not able to find other studies against which to compare the validity and generalizability of our results, as those available were published several years ago or were carried out in groups with specific diseases, a few more recent papers support our findings or give additional and complementary evidence about the need to involve public representatives or health consumers in main aspect of clinical research. Ellis,18 in a review of the literature, evaluated the issues pertinent to physician and patient participation and identified strategies to inform patients better about clinical trials. The author reported several patients' and physicians' characteristics that influence participation in randomized clinical trials, such as doctors' attitudes to randomization, logistic difficulties, ethical concerns, etc., and suggested that patients' understanding about clinical trials may play an important role. They identified at least three randomized clinical trials testing this hypothesis in cancer patients, and all confirmed that providing patients with of more information patients preserved their independence, improved recall and understanding of information, but had little or no impact on participation.17,19,20

Ellis et al.9 also carried out a cross-sectional survey on a small sample of outpatients attending a medical oncology clinic, to assess the understanding of and attitudes towards randomized clinical trials. Knowledge was not good: only one-third would consider participating in a randomized study and patients were more likely to participate if a study was endorsed by an independent cancer information service. They concluded that new strategies are needed to educate the public and patients about randomized clinical trials, and that consumers should be involved in the design and conduct of these trials.

Hanley et al.7 administered a structured questionnaire to a sample of 103 coordinating centre involved in clinical trials in the UK to assess the extent to which consumers were involved. Of the 62 eligible responders, 23 reported that consumers had already been involved in their work and 17 said they had planned to involve consumers. Responders were mostly positive about consumers participation, and most commented that input from consumer had helped refine research questions, improve the quality of information and make the trial more relevant to the needs of patients.

Key points

  • Few information is available about the consumer's awareness and attitude toward clinical research.

  • A standardized questionnaire (seven questions on knowledge and two scenario on attitude to trial participation) was used on a convenience sample of 2000 Italian health consumers.

  • Consumers are generally not very aware regarding clinical research, and their attitude to participate is related to the level of awareness.

  • New strategies are needed to educate the health consumers and patients about clinical research, and trial participation.

  • It is important to create a meaningful partnership between health professionals, consumers, and their associations to implement educational interventions.

Acknowledgments

We thank A. De Toni, C. Hanau, M. Niero, A. Orrù, P.L. Persi, L. Pomodoro, G. Recchia and G. Righetti for their contribution to developing and testing the questionnaire. We thank the interviewers who conducted the survey, in particular: A. Orrù (Lombardia), G. Righetti (Veneto), E. Cuccaro (Friuli), M. Foti (Emilia Romagna), O. Testa (Marche), L. Pilla (Lazio), I. Silvano (Campania), L. Lezzi (Puglia) and R. Calamita (Sicilia). We also thank all the people who responded to the survey. This work was partially supported by an unrestricted research grant from GlaxoSmithKline SpA, Verona, Italy. G.A. served as a paid consultant to GlaxoSmithKline SpA during the period the study was carried out.

References

View Abstract