The European Journal of Public Health Advance Access originally published online on March 8, 2006
The European Journal of Public Health 2006 16(5):498-504; doi:10.1093/eurpub/ckl024
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Health-related behaviours |
Third Italian national survey on knowledge, attitudes, and sexual behaviour in relation to HIV/AIDS risk and the role of health education campaigns
Carlo Signorelli1, Cesira Pasquarella1, Rosa Maria Limina2, Edoardo Colzani1, Mila Fanti1, Antonia Cielo1, Massimo Greco3, Chiara Porro de' Somenzi4, Maria Chironna5 and Michele Quarto5
1 Department of Public Health—Hygiene Section, University of Parma, Parma, Italy
2 Department of Experimental and Applied Medicine—Hygiene Section, University of Brescia, Brescia, Italy
3 Local Health Authority n.2 of Perugia, Italy
4 Local Health Authority ‘City of Milan’, Italy
5 Department of Biomedical Science and Human Oncology, Hygiene Section, University of Bari, Bari, Italy
Correspondence: Carlo Signorelli, Department of Public Health—Hygiene Section, Via Volturno 39, 43100 Parma, Italy, tel: +39 0521 903831, fax: +39 0521 903832, e-mail: carlo.signorelli{at}unipr.it
Received January 29, 2005, accepted January 19, 2006
 |
Abstract |
|---|
 Top Abstract MethodsResultsDiscussionReferences |
|---|
Aim: The survey, supported by the National Health Institute (ISS), investigates on sources of information for HIV/AIDS and sexual behaviour of the Italian general population. Methods: The survey was carried out in four different Italian provinces with different geographical, social, and epidemiological patterns of HIV/AIDS: Bari, Milan, Parma, and Perugia. A self-administered questionnaire was used to interview a sample of 2000 people, aged 18–49 years, chosen using a quota-sampling, considering age, sex, and level of education. Results: A total of 1985 persons filled in the questionnaire. Among them, 73.2% reported Ministry of Health as their main source of information and 76.7% TV/radio advertising as their main means of information. Sexually active subjects, experiencing occasional sexual intercourse, were 22.4%; of them 34.6% did not refer a systematic use of condom. Among sexually active people, half (45.7%) reported as irrelevant their risk of infection, 6.9% as high. Being male (OR = 6.175) and having the first sexual intercourse before 18 years (OR = 2.076) were significantly associated with sexual behaviour at risk (males having unsafe sexual intercourses with prostitutes or during partners' menstrual period; both sexes having unsafe occasional sexual intercourses or anal sex), while living in Northern Italy (OR = 0.672) was significantly protective. Subjects with sexual behaviour at risk were about 40%. Conclusions: The need for further education campaigns, and their consequent evaluations, is relevant. Monitoring sexual behaviour of general population, focusing on risk behaviour rather than risk groups, is fundamental nowadays in preventing HIV.
Keywords: AIDS, attitudes, knowledge, sexual behaviour, survey
The introduction of new and more effective therapies to treat AIDS does not decrease the importance of preventive initiatives concerning information and education, especially concerning the risk of sexual transmission of the virus. Preventing the transmission is still the only way in order to stop the diffusion of HIV/AIDS epidemic.
Nowadays, gaining and monitoring knowledge of sexual behaviour of the whole population, and in particular of the most sexually active groups (18–49 years), has become of striking importance since epidemiological data suggest a relative increase of infections due to heterosexual transmission.
In Western European countries, between 1997 and 2002, the number of new HIV cases among subjects infected through heterosexual contact has markedly increased (+116%) while among injecting drug users (IDU) it has gradually decreased (–9%) over the same period.1
In Italy, the proportion of heterosexuals infected by HIV was 11.8% between 1982 and 1984 and then grew up to 39.8% between 2002 and 2003. Heterosexual transmission is estimated to count for 39.8% of total infections among women and for 13.7% among men.2
The majority of the studies carried out so far in Italy evaluated knowledge, information, and the different ways of transmission among specific risk groups of the population.3–7
In Europe, in the nineties, a common methodology in order to compare different data concerning sexual behaviour in different countries (‘Concerted Action’) was adopted, so that comparable information from different cultural and social backgrounds could be obtained.8–10
Unfortunately, studies aiming at the impact evaluation of HIV/AIDS education campaigns and at the analysis of the correlation between information/education initiatives and possible changes in sexual behaviour are still missing in Italy so far.11,12 There is also a lack of studies considering representative samples of the entire Italian population.
These considerations underlined the need for a survey among the sexually active Italian population that could be infected by sexual transmission. The choice of investigating the general population was also due to the need to focus on risk behaviour rather than risk groups.
New techniques for monitoring sexual behaviour are therefore fundamental if the objective is to develop new and more effective strategies of HIV/AIDS prevention for the general population, since HIV is still a major public health burden in Europe.13–16
|
Methods |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
A cross-sectional study was carried out considering four Italian provinces (Milan, Parma, Perugia, and Bari) thought to be representative of the entire country. The four provinces analysed have different incidence of HIV infection as well as geographical and social characteristics.2,17 Data collection was carried out through an anonymous and self-administered questionnaire that was validated by a previous pilot study performed on 100 subjects.
Target population
The target population was represented by people aged 18–49 years. Even though the risk of infection persists in people aged more than 49, the majority of those who adopt sexual behaviour at risk are infected within 49 years of age, according to WHO recommendation.1
Sample choice
The sample size was of 2000 subjects. Considering that the total resident population aged 18–49 years on 1 January 2002, according to ISTAT (Italian Institute of National Statistics) estimates, was 25 877 030 (12 870 987 women and 13 006 043 men); 
7.73 interviews in 100 000 residents were performed.17 Five hundreds subjects were interviewed for each province, even though their population size was slightly different. However, data were weighted during the analysis by each province size.
A quota sampling was performed.18–20 The population was equally divided into three categories including sex (male and female), level of education (up to primary school, high school, and university degree), and age groups (18–28, 29–39, 40–49). Selection of subjects to be enrolled in the study was performed by the investigators carrying out the sampling until they fulfilled each quota for the four recruitment categories (province, age group, sex, and level of education). The subjects participating in the study, assured on the confidentiality of the information given, received little gadgets after completing the questionnaire.
Quota sampling showed the following advantages:
- adequate sample size: the sampling continued till the fulfilling of each quota and therefore possible problems due to a low response rate were avoided;
- reduced length of the study: subjects filled in the questionnaire where and when the recruitment took place; and
- problems of privacy (L. 675/96) avoided: since sensitive data, such as addresses or personal data taken for example from the General Registry Office, were not used.
Sampling places and times
Place of sampling and timing were of strategic importance. The sampling took place where everyone had the same chance to be recruited independently of his/her attitudes possibly related to sexual behaviour: places of local importance such as central avenues during days of high pedestrian transit were chosen either during working or weekend days.
The interviews took place in the same period in the four cities during November–December 2002. Days and times for the interviews were chosen locally.
The questionnaire
The questionnaire was based on 53 multiple choice questions divided into four sections (general information, health education campaigns and HIV/AIDS knowledge, sexual behaviour, and present sexual partners).
The use of a self-administered questionnaire showed advantages and disadvantages. Some of the advantages were anonymity with consequent higher reliability of the answers, free time to read and understand the questions, and possibility to ask the investigator if something was unclear. Moreover, the administration of the questionnaire was standardized, the costs were low, and the filling was relatively quick.
Some possible disadvantages were diffidence at the first investigator approach and filling mistakes such as skipping questions or giving inconsistent answers.
Data analysis
Data analysis was initially performed using the PC platform Stat ViewTM SE+ 1.02 for Macintosh. The population sizes, relevantly different among provinces, according to ISTAT Database (last update 27 March 2002), were 3 642 721 (weight: 1.356 x 10–6) for the province of Milan, 398 350 (weight: 12.552 x 10–6) for the province of Parma, 618 698 for the province of Perugia (weight: 7.936 x 10–6) and 1 530 840 (weight: 3.266 x 10–6) for the province of Bari.21 However, computing both weighted data and not-weighted data yielded very similar results; therefore, not-weighted data were considered in our data analysis.
As a further step the PC platform SPSS 11.5 for Windows was used in order to carry out a logistic regression model, using the Backward Stepwise (Wald) method with a 0.10 cut-off and 
< 0.05, including possible determinants of sexual behaviour. As independent variables various individual factors and possible confounders such as age group, gender, place of living, level of education, sources and means of information about HIV/AIDS, and age at the first sexual intercourse were considered. Sexual behaviour at risk was considered as dependent variable and was defined as experiencing at least one of the following: males with at least one unsafe sexual intercourse with prostitutes or during partners' menstruation period, males and females with at least one unsafe sexual intercourse with occasional partners or anal sexual intercourse. The logistic regression considered only subjects with at least one sexual intercourse in life.
|
Results |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
A proportion of 42.8% did not stop at the recruitment place and were unwilling to listen to the purposes of the study usually because they were in a hurry or afraid of being cheated. A total of 1985 subjects had their questionnaires analysed from the 2000 expected (response rate = 99.25%). In particular, of the expected 500 per province, 500 were correctly completed in Parma, 491 in Perugia, 494 in Milan, and 500 in Bari. Fifteen questionnaires were not considered because they were unreliable, not sufficiently completed, or given back blank to the interviewers.
The most frequent means of information reported (table 1) were TV and radio messages (76.7%). Subjects who got information from newspapers or magazines were 69.4%, while 66.1% were informed mainly by TV or radio programmes and 58.1% by information booklets. On the other hand, the most reported source of information for HIV/AIDS (table 1) was the Ministry of Health with its information campaigns (73.2%).
|
A proportion of 35.7% subjects reported that they have modified their sexual behaviour after the information campaigns by avoiding occasional sexual intercourses (50.9%); always using condom in occasional sexual intercourses (47%); always using condom in all sexual intercourses (37.2%). A proportion of 28.2% changed their sexual behaviour completely, while 7.5% did it but just for a brief period.
By maintaining ‘AIDS is not a matter of concern because now there are therapies’ (table 2) the investigators evaluated subjects opinions about implications of new drugs against HIV and its progression to AIDS. A proportion of 80.9% correctly disagreed, but 7.1% agreed and 12% did not know. The proportion of people answering correctly decreased with age from 83.8% (18- to 28-years-old) to 77.3% (40- to 49-years-old).
|
Other two similar questions were asked in order to assess knowledge of HIV/AIDS and its current therapies (table 2). Most of the people correctly disagreed with the statement that ‘Precautions are not necessary in order to prevent infection’ (79.5%) and correctly agreed instead with the statement that ‘It is better to prevent infection with correct behaviour because with current therapies there is no complete recovery’ (85.6%). However, there were significant differences between the provinces of Parma and Perugia and the provinces of Milan and Bari, quite surprisingly showing a greater proportion of wrong and ‘Don't know’ answers in the two great urban provinces (Milan and Bari).
The overall number of sexually active people in the sample, that is those who experienced at least one sexual intercourse in life, was 1696 and corresponded to a proportion of 86.84%.
The age cohort analysis of the first sexual intercourse (table 3) showed a clear decrease in the mean age of the first sexual intercourse in younger cohorts.
|
The mean age at first sexual intercourse of the oldest age cohort (those born in 1953–1957) was 18.69 gradually decreasing down to 18.26 for the age cohort born in 1963–1967. This decreasing trend was interrupted by the age cohort born in 1968–1972 for which the mean age at the first sexual intercourse was 18.37. In the age cohort born in 1973–1977 (25- to 29-years-old at the moment of the interview) there was again a slight decrease of mean age down to 18.17 (median value 18), and eventually the mean age at the first sexual intercourse decreased suddenly down to 16.91 (median value 17) among the youngest age cohort born in 1978–1984 (18- to 24-years-old at the moment of the interview).
According to the previous mentioned risk categories, a proportion of 39.67% of the overall sample is to be considered at risk for HIV infection. Sexually inactive subjects were also included in this computation in order to get the overall burden of sexual risk behaviour in a population representative of the Italian general population.
Subjects, with a sexual intercourse in life, who had experienced occasional sexual intercourses were 22.4%. This behaviour seemed more frequent among those living in urban areas and less frequent among people living in rural areas. The subjects were asked how many times in a 0–10 scale they did not used condom during previous occasional sexual intercourses and those who answered ‘less than always’ were classified as being at risk for HIV. The crude proportion of subjects who had experienced occasional sex in life and did not always use condom in such occasions was 67.3% among males and 62.1% among females (300 males and 145 females, respectively). Association between this behaviour at risk and province of living was not significant but showed a moderately higher proportion in Central (Perugia, 69.2%) and in Southern (Bari, 65.6%) Italy.
When asked about self-risk perception, 45.7% of the sexually active group reported as irrelevant their own risk of contracting the infection, only 6.9% reported it as high. A proportion of 75.1% of subjects with sexual behaviour at risk reported their risk profile to be low or irrelevant (table 4). A strong significant positive association with actual sexual behaviour at risk was found, showing that people at risk were self-conscious of their own risk profile.
|
According to the data from this survey, 39.3% of the sexually active subjects tested at least once for HIV. Of these, 20.7% tested just once, while 18.7% more than once. The association between sexual risk behaviour and prevalence of HIV-testing was significant, indicating that people at risk were testing more frequently than people with sexual behaviour not at risk. Anyway 58.7% of the subjects at risk have never sought HIV testing (table 4).
From the logistic regression model it is possible to see how being male (OR = 6.175; 95% CI 4.905–7.774), having first sexual intercourse before 18 (OR = 2.076; 95% CI 1.645–2.619), and getting information about AIDS from friends (OR = 1.369; 95% CI 1.065–1.758) and from books (OR = 1.381; 95% CI 1.084–1.760) were associated with a higher frequency of sexual behaviour at risk for HIV infection. Receiving information about HIV/AIDS from the family (OR = 0.728; 95% CI 0.563–0.943) and living in the Northern provinces (Milan and Parma) (OR = 0.672; 95% CI 0.534–0.846) were, instead, significantly protective for sexual behaviours at risk (table 5).
|
Possible biases
Quota sampling cannot lead to a precise estimation of the random error because the probability of the recruitment of each subject is not known. Moreover, the lower the standard deviation within the sample and the larger the sample size, the lower would be the random error. The sample was considered adequate for the objectives of this study.
Interviewers training and standardization of approaching methods were used in order to avoid possible biases caused by the investigator.
Information bias might have been caused by possible pre-existing attitudes towards HIV or surveys in general, especially if the objectives of the study had not been clearly explained in advance. The investigators were aware of this possible bias and tried to write an exhaustive and clear questionnaire explaining carefully to the interviewed all the objectives of the study and the fact that it was absolutely free and anonymous.
|
Discussion |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
The sample remembers the Official national institutions such as the Ministry of Health are remembered as the main source of information for HIV/AIDS; the role of the local health educators such as general practitioners (17%), teachers (31%), and health professionals of the Local Health Authorities (32%) was not considered as relevant. The fact that the most reported sources of information were TV, radio, and magazines might open the discussion on whether these ways of information are really educational or just superficially informative for most of the people. This concern seems confirmed by the fact that even though the level of knowledge about HIV/AIDS was good (
80% of correct answers), at the same time sexual behaviour at risk seems really diffuse in the population (40%) even though strict criteria for sexual risk behaviour were adopted in this study. Previous health education campaigns, other than those of the Ministry of Health, did not seem to be very effective in reducing risk behaviour. Most of the sources and means of information reported by the subjects recruited were not significantly related to their sexual behaviour. Nevertheless, information received by books, friends, and family showed significant association, even though family was the only one associated to a decrease in sexual risk behaviour. Previous information campaigns of the Ministry of Health seem to have positively influenced sexual behaviour, though this association was not significant P < 0.05. These results confirm the concerns about the real efficacy of previous health education campaigns and the lack of reinforcing campaigns. Nevertheless, they underline the need to strengthen and focus them.
As expected, the first sexual intercourse at a very young age is an indicator of a higher probability of behaviour at risk in adult life. If the decreasing trend of the age at the first sexual intercourse is confirmed by the new generations, these data suggest that in the next future AIDS might increasingly involve even very young heterosexuals. The age group 18–24 (born in 1978–1984) showed an average age at the first sexual intercourse lower than 17 years. This fact becomes even more relevant if compared to the age group 25–29 (born in 1973–1977) where the median age at the first sexual intercourse was 18 years (table 3). It is interesting to notice how this decreasing trend changed sensitively among the 30–34 age cohort (born in 1968–1972): this cohort of people, during the eighties, was in fact the first generation dealing with HIV.
Since HIV pattern of diffusion is involving more and more the heterosexuals in Western Europe, one of the most concerning ways of transmission is therefore occasional sex, especially in the youngest age groups, also because it involves a significantly higher number of people in the general population than homosexual intercourses and IDU do. This concern was confirmed by these data, since one-fourth of the interviewed experienced occasional sex at least once (22.4%). Even more worrying is the fact that about two-thirds of those experiencing occasional sex reported that they did not always use condom in such occasions. The crude data suggest a greater burden among males and in the provinces of Perugia and Bari (Central and Southern Italy). These crude trends were confirmed by the logistic regression analysis in which males and people from Central and Southern Italy were found to be generally more at risk than females and people from the North.
Risk perception is believed to be a very important indicator in order to assess general population's concern, and therefore knowledge, of HIV/AIDS, especially if compared with actual sexual behaviour. The results showed a positive association between risk perception and sexual behaviour at risk. However, there was a high proportion (75.1%) of subjects perceiving their own risk as low, even though their sexual behaviour was not always safe. This issue highlights for the need of new and effective education campaigns.
A significant association between those who tested for HIV and subjects with sexual behaviour at risk was found; moreover, 58.7% of people considered at risk in this study never sought HIV testing. As regard HIV test execution, there is a wide agreement about the fact that an increasing number of people at risk voluntary testing for HIV would be a positive result for Public Health since some studies showed that knowing someone's own serologic status can positively influence sexual behaviour; it is still not very clear if it would be positive, or if it is to be encouraged an increase in the proportion of general population testing for HIV.22
The difference between the proportion of people with behaviour at risk in Northern Italy and in Central-southern Italy might prove that there are important cultural differences between these two areas and that there might be a different need for health education campaigns. As was expected, sexual behaviour of males was more at risk than females. However, HIV can be heterosexually transmitted to the usual partner, if one of the two, more often the male, has or had sexual behaviours at risk.
Health education represents a fundamental tool for HIV/AIDS prevention. Even though many health education campaigns have been carried out in Italy at a national and at a local level, there is still a lack of efficacy evaluation of such interventions. As a consequence of the results presented in this study, the importance of further education campaigns focused on both specific groups with behaviour at risk and the general population has to be stressed.
AIDS is no more a disease limited to certain groups of people such as homosexual males, prostitutes, or IDU, but it is gradually spreading among the general population and the heterosexuals. Monitoring by surveys the sexual behaviour of the entire population is therefore of striking importance in order to carry out efficient and well targeted health education campaigns that should involve the whole sexually active population.
The choice of risk categories with such strict and cumulative criteria might be considered a limit to the interpretation of the data collected, but this is consistent with the objective of the study: the aim was to survey the general population with the higher possible sensitivity, even though losing specificity, because, unfortunately, there is no scientific reason not to consider at risk for HIV people who had unsafe sex in their life, even though only once. Actually, the higher is the prevalence of sexual behaviour at risk, the higher will be the incidence of infection among the heterosexuals, and therefore among the general population. In this case even people with a safe sexual behaviour are exposed to a moderate/high risk due to previous behaviour of their partner, that is what is currently happening in developing regions like sub-Saharan Africa and Asia.23
In conclusion additional actions need to be performed in the field of HIV/AIDS prevention, even if each initiative should be monitored and then evaluated and compared with previous ones. If efficacy is the aim, we need to assess it, because we cannot be sure that certain actions are equally effective in different cultural and geographical backgrounds. It is also very important to monitor not only infections and new cases but also sexual behaviour of the whole population, which is now becoming the main target of HIV infection.
Key points
|
|
Acknowledgments |
|---|
The work should be attributed to the Hygiene Section of the Department of Public Health—University of Parma, Via Volturno 39, 43100 Parma, Italy. Source of support: Italian Ministry of Health. This study was supported by National Institute of Health (ISS). The authors thank the Authorities of the municipalities and all those involved in data collection (N. Sodano, D. Bellini, A. M. Papa, S. Goretti, V. Capoccia, B. Diso, M. Marra, L. Mazzilli, A. Milano, R. Squicciarini, L. Vitto, E. Alliata, L. Bassoli, L. Brizzi, B. M. Castelli, D. Di Napoli, E. Falappi, M. Ferrari, P. Moja, G. Perotti, and S. Settembrini). Special thanks to G. Rezza for the support.
|
References |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
1 European Centre for the Epidemiological Monitoring of AIDS. HIV/AIDS Surveillance in Europe. Mid-year Report 2003; No. 69.
2 COA-ISS. Aggiornamento dei casi di AIDS notificati in Italia al 31 dicembre 2003. Not Ist Super Sanità 2004;17 (Suppl 1):1–17.
3 Fara GM, Tarsitani G, Osborn JF, et al. Comportamenti sessuali e rischio AIDS nei giovani italiani. Ottobre 1999. Quaderni dell'Istituto di Igiene dell'Università di Milano, 1999.
4 Renzi C, Signorelli C, Zantedeschi E, Fara GM. Early onset of sexual activity-A possible "marker" for subsequent high risk sexual behaviour in young Italians. J Prev Med Hyg 2001;42:1–14.
5 Signorelli C, Allegretti A, Benigni M, et al. Indagine nazionale sui comportamenti sessuali dei giovani in relazione al rischio AIDS. Ann Ig 1998;10:19.[Medline]
6 La Torre G, Colazingari G, Arzano I, et al. Sexual knowledge, attitudes, and behavior in HIV-seropositive individuals at the AIDS referral centre Latina. Ann Ig 2002;14:465–72.[Medline]
7 Porta D, Peducci CA, Forestiere F, De Luca A. Temporal trend of HIV infection: an update of the HIV surveillance system in Lazio, Italy, 1985–2000. Eur J Public Health 2004;14:156–60.
8 Renzi C, Zantedeschi E, Signorelli C, et al. Sexual behavior and the HIV/risk in the general population: the methodological aspects of a national study within the context of a European Concerted Action. Ann Ig 1999;11:83–94.[Medline]
9 De Luca A, Renzi C, Zantedeschi E, et al. Sexually trasmitted disease among Italian heterosexuals: result from a general population survey within a European Concerted Action. Ann Ig 2001;13:387–92.[Medline]
10 Renzi C, Zantedeschi E, Signorelli C, Osborn JF. Factors associated with HIV testing: results from an Italian general population survey. Prev Med 2001;32:40–8.[CrossRef][Web of Science][Medline]
11 Signorelli C, Zantedeschi E, Allegretti A, et al. L'AIDS ed i mezzi di informazione. Ann Ig 1998;10:182.
12 Signorelli C, Renzi C, Zantedeschi E, Bossi A. Prevention focused on sexual behavior. Ann Ist Super Sanità 2000;36:441–3.[Medline]
13 Anderson JE, Stall R. How many people are at risk for HIV in the United States? The need for behavioral surveys of at risk populations. J AIDS 2002;29:104–5.
14 Herlitz CA, Steel JL. A decade of HIV/AIDS prevention in Sweden: changes in attitudes associated with HIV and sexual risk behaviour from 1987 to 1997. AIDS 2000;14:881–90.[CrossRef][Web of Science][Medline]
15 Holtzman D, Bland SD, Lansky A, Mack KAZ. HIV-related behaviour and perceptions among adults in 25 states: 19976 behavioural risk factor surveillance system. Am J Public Health 2001;91:1882–8.
16 Anderson JE. Condom use and HIV risk among US adults. Am J Public Health 2003;93:912–4.
17 Available at: http://demo.istat.it/pop2002/index.html. Accessed on 15 October 2002.
18 Hansen MH, Hurwitz WN, Madow WG. Sample Survey Methods and Theory, Volume 1, Methods and Applications. New York: John Wiley & Sons, Inc. 1993.
19 Signorelli C. I questionari nella ricerca epidemiologica. Roma: SEU, 1998.
20 Signorelli C, Osborn JF. Comparison of different approaches for a survey of the general population. Ann Ig 1998;10:67–74.[Medline]
21 Available at: http://dawinci.istat.it/daWinci/jsp/prTavola.jsp. Accessed on 15 October 2002.
22 Renzi C, Zantedeschi E, Signorelli C, the NEM Group. Voluntary HIV testing in Europe. Scand J Public Health 2004;32:102–10.[CrossRef][Web of Science][Medline]
23 Mwaluko G, Urassa M, Isingo R, et al. Trends in HIV and sexual behaviour in a longitudinal study in a rural population in Tanzania, 1994–2000 AIDS 2003;17:2645–51.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  L de la Fuente, M Suarez, M J Belza, F Vallejo, M Garcia, R Alvarez, J Castilla, A Rodes, and The Health and Sexual Behaviour Survey Group Human immunodeficiency virus testing uptake and risk behaviours in Spain J Epidemiol Community Health, July 1, 2009; 63(7): 552 - 558. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||