Spatial patterns analysis and hotspots of HIV over 20 years using geographic information system. A case study of Kermanshah, West Iran

Introduction

Human immunodeficiency virus (HIV) is an epidemic disease causing serious challenges worldwide, threatening the general welfare of humans in a way that has turned into one of the major public health problems in the world [1, 2]. Due to the rapid transmission of epidemic diseases and the resulting potential effects on societies, the early detection of such epidemics is the top priority of public health. The world statistics published in 2014 showed that there were 34 million people infected with HIV worldwide, of whom 4.8 million people were based in Asia [3], and according to the statistics provided by the Ministry of Health of Iran, more than 27 thousand HIV-stricken patients were identified by 2014, of whom 89.3% were male and 10.7% were female [4].
The number of new cases of the disease is still growing, varying from one country/region to another [5]. The spatial analysis of contagious diseases, as one of the effective etio­logical methods, is given utmost importance and priority, together with growing evidence indicating that the dynamic spatial data are key elements in monitoring and predicting the spread of contagious diseases [6-8]. Despite the significant geographic variations in HIV, the analysis of the spatial distribution of this disease has been paid less attention [2, 9].However, like other contagious diseases, understanding the spatial variations in the prevalence of HIV and its stimulants in socio-spatial contexts causes the preventive interventions to be carried out with more appropriate planning [10-12], and to this end, the geographic analytic technique of the geographic information system (GIS) is a powerful tool [10, 13, 14] whose main feature is the possibility of understanding geographic patterns in space and time [15]. GIS is a powerful source for the health of communities, with capabilities like the integration of data from different sources towards generating new information and intrinsic visualization applications, which can leave lasting effects on the lives of people through creative problem-solving [10, 12].
Many studies have been carried out worldwide on the epidemiology and spread of HIV using GIS [2, 10, 12-14, 16-20]. But given the importance of geography in health, unfortunately, the number of geographically explicit studies on HIV is still so limited that, according to investigations so far, only one case study has been performed on the spatial distribution of HIV in Iran [21]. The present work is among the first studies done in this regard in Iran, and given that there are significant weaknesses in the implementation of the existing policies, the performance of organizations and the design of new policies against HIV [22], it seems that some basic steps can be taken towards HIV prevention and the principles and requirements of decision-making, planning and the required actions by all agencies, particularly the health care system and all engaged sectors, can be better formulated through the agency of sufficient knowledge and understanding of the epidemiological status and spatial trends of HIV throughout Iran. Furthermore, the results of the present study can provide useful insights into better understanding of the dynamics of HIV in Iran and other countries, especially the developing ones, and help the knowledge about this disease be globally deepened. Therefore, considering the above-mentioned material and the lack of accurate and scientific data related to patients suffering from HIV in Iran (Kermanshah Province), the present work was intended to provide the Iranian health managers and policymakers with some scientific insights through analyzing the spatial trends of HIV over 20 years.

Material and methods

The research method used in the present study was descriptive, analytic and cross-sectional, and the statistical population consisted of all patients diagnosed with HIV in Kermanshah Province. Furthermore, the data on the HIV-stricken patients were collected from the files recorded in the Vice Chancellery of Kermanshah University of Medical Sciences, Department of Contagious Diseases. Moreover, the resulting demographic and background information included gender, education, probable routes of transmission of the disease, patients’ survival status and leading causes of death. To obtain the descriptive statistics, graphical models and spatial trends of the disease, the GIS Software was used. First, all patients’ data were collected from their recorded files in the Department of Contagious Diseases of Kermanshah University of Medical Sciences, then the collected data were changed into a database (GEO database) in the environment of Arc/GIS 10.2.2 Software, and each case of the disease was entered as a point on the map of Kermanshah according to the patients’ addresses. To identify the spatial patterns of HIV in Kermanshah Province, the graphic-based statistical models were used in the environment of GIS. The tests used in GIS included mean center, standard distance, and Moran index (equations 1, 2, and 3):

Equation 1:

Equation 2:

Equation 3:

Results

Nearly 3318 inhabitants based in Kermanshah were reported to be infected with HIV during 1994-2013, of whom 92% were male and 8% were female. Moreover, intravenous drug users had the highest percentage of being stricken with HIV (85%), which was 13% in the case of being stricken through sexual intercourse. Of all 3318 infected with HIV, 1127 patients died of HIV (45.94%) and addiction (through overdose) (21.09%) (see Table 1).
Studying the trends of HIV over the period under study in Kermanshah Province showed that the majority of the recorded cases were related to 1997 and 2001-2005 (Figure 1A). However, during the past few years, women showed upward trends (Figure 1B). Furthermore, the number of dead with HIV was more than the living ones over the period 2006-2007, and 2013 (Figure 1C). Also, studying the causes of the infected patients’ death demonstrated that the leading cause of their death was HIV per se, and it was only in 1997 that suicide was reported as the most common cause of death (Figure 1D). Intravenous drug injection had the highest degree of transmission compared with other routes. Over the past 20 years, intravenous drug injection and sexual intercourse had downward and upward trends, respectively, which were reported the same in 2013 (Figure 1E). Comparing the age groups infected with HIV showed that the 35 to 44 age group had the highest percentage over 1995-1997, and the 25 to 34 age group made up the majority infected with HIV over 1998-2013 (Figure 1F).
Studying 14 townships based in Kermanshah Province demonstrated that Kermanshah County had the largest number of HIV-stricken patients (88.5%), and the majority of reports of the disease were within 2001-2007 (see Table 2). Moreover, studying the prevalence of this disease in Kermanshah Province during the first period (1994-2000) indicated the presence of four townships with more than 16 cases of the disease. Furthermore, in the second period (2001-2007), the number of townships increased to seven, and in the third period (2007-2013), this volume increased to 10 townships (Figure 2).
To study the spatial trends of HIV, the mean center and standard deviation oval were used. The standard deviation oval indicated that HIV was dispersed and increased over the period 1994-2013, in a way that the spread of the disease had a west-east direction in the first period (1994-2000). In the second period (2001-2007) the spread of the disease drew northeast-southwest, and in the third period (2007-2013) it progressed in the four directions of North, South, West, and East. It is noteworthy that the mean centers of all three periods were in Kermanshah County (Figure 3A). Furthermore, clustering the townships of Kermanshah Province showed the presence of three hot spots in the vicinity of each other, i.e. Kermanshah, Harsin, and Islam Abad (Figure 3B).

Discussion

The present study aimed to investigate the spatial distribution and trends of HIV in Kermanshah Province over 20 years, i.e., 1994-2013, and it is the first study looking into the spatial distribution of HIV in western Iran with an emphasis on Kermanshah Province. The results demonstrated that about 11% of all HIV-stricken patients in Iran are based in Kermanshah Province, and the distributions of the disease were focused in three townships in the form of clusters.
The results showed that 3318 patients infected with HIV were recorded in Kermanshah Province over 1994-2013, which was about 11% of all HIV-stricken patients in Iran (170 cases per 100,000 people) while 2.5% of the total population of Iran is based in Kermanshah Province. To explain this finding, it can be expressed that one of the aspects of HIV has its roots in unemployment and poverty [23], and this province also has high statistics of unemployment and poverty [24, 25]. Therefore, it is possible that the main reasons for the proliferation of HIV in Kermanshah Province are unemployment and poverty. According to a study performed by Ghanbarnejad et al., Kermanshah Province (796 cases) held around nine times more HIV-stricken patients than Hormozgan Province held (88 cases) within 2007-2011. This difference may be due to the miscalculated statistics related to the HIV-stricken patients in Hormozgan Province and the better performance of the monitoring system in Kermanshah Province. However, comparing the statistics related to the HIV-stricken patients in northern Taiwan (1264 cases) [19] with the size of the infected population in Kermanshah (2515) within 1997-2008 demonstrated that Kermanshah held twice as many HIV-stricken patients than northern Taiwan. Therefore, these statistics are of great significance for health managers and policymakers of Kermanshah Province and Iran.
Studying the trend of HIV in Kermanshah Province over 20 years indicated that the majority of cases were registered in 1997. The period 2005-2001 held the second position in this regard (Figure 1A). Also, the trend of HIV in terms of gender over 20 years showed that the male population made up the highest percentage of patients infected with HIV (92%), forming the dominant population in this case. The national statistics also indicate that there are more men (89%) than women in terms of HIV infection [4]. It is noteworthy that during the period under study, the female population had an upward trend; for example, in 1997, there was only one woman (0.19%) infected with HIV compared with 508 men, whereas in 2013, this proportion changed to 21 (30.88%) women versus 47 men (Figure 1B). Since most Iranian women are young and at childbearing ages [26], they are more susceptible to HIV and the serious consequences of this disease, resulting in irreversible damage to females. By comparison with other countries, for example Kenya, the percentage of women infected with HIV is higher than that of men (55%) [27], which is possibly due to virus transmission routes. In terms of patients’ survival, about 34% died, and comparing its 20-year trend showed that the number of patients living with HIV was more than the dead ones over the period under study, except for 2006, 2007 and 2013 (Figure 1C). This situation might have been affected by the use of antiviral drugs during the past few years [28]. Of the 1127 people who died, 45.94% of deaths were because of HIV, 21.09% because of addiction (through overdose) and 4.45% because of suicide. These statistics show that HIV has been the leading cause of mortality of patients.Also, suicide was the most common cause of death in 1997 (Figure 1D). It seems that the existence of this pattern in the early years was a sign of people’s lack of awareness and fear of HIV, which gradually changed into dying of HIV due to carrying out needs assessments in the country over the next few years [29, 30] and training in this regard. Additionally, the findings showed that intravenous drug injection (85%) and sexual intercourse (13%) held the highest percentages in terms of infecting patients with HIV, and it is noteworthy that intravenous drug injection and sexual intercourse had downward and upward trends in recent years, respectively, whose statistics were reported the same in 2013 (Figure 1E). The reason for this change may be the third wave of AIDS (sexual contact) in Iran. The pattern of AIDS transmission in Iran, which was mainly through addiction and sharing syringes, has now changed to sexual transmission. The disease, which has undergone three waves in Iran, became initially epidemic through blood products and then through sharing syringes in prisons and rehabilitation centers, but the third wave is now due to sexual contacts [31, 32]. To explain in more detail, it can be claimed that the pattern of drug use in Iran has changed in recent years; i.e., the rate of intravenous drug injection has decreased and the use of other drugs such as ecstasy, heroin and glass has become more prevalent [33]. Furthermore, HIV transmission through sexual contacts has increased in recent years [34].
Comparing the different age groups infected with HIV showed that the 35-44 and 25-34 age groups had the highest percentages over 1995-1997 and 1998-2013, respectively (Figure 1F), i.e., the age group pattern in Kermanshah Province has changed from middle age to youth, corresponding with the national statistics [4]. Studying the trend of HIV in Kermanshah Province using GIS demonstrated that the rate of this disease has had upward trends over 20 years and the majority of HIV-stricken people (88.5%) were based in Kermanshah (284 cases per 100 000 population) (Figure 2), which is a natural trend like in other countries, but the remarkable point that should be taken into consideration here is that this increase in the HIV transmission rate was the result of the poor health status [35]. Because Kermanshah Province was among the semi-developed provinces of Iran in terms of health indexes [36], any increases in HIV rates will be a serious threat. A comparative study of the number of HIV-stricken patients reported in Kermanshah Province during the three aforementioned periods revealed that most reports were related to 2001-2007 (Table 1). It seems that the monitoring systems have well diagnosed HIV, and the adoption of preventive policies, including the distribution of free syringes to addicts after 2007, might have affected the transmission of this disease.
Additionally, studying the trend of patients infected with this disease indicated that there were no reported statistics in the first (1994-2000) and second (2001-2007) periods in some townships such as Ravansar and Paveh (see Table 2). To further explain, it can be claimed that these townships are possessed of some traditional and ethnic characteristics, influencing the status of the disease.
To further study the spatial trends of HIV, the mean center and the standard deviation oval were used in the environment of GIS. The results of the standard deviation oval indicated that the spread of HIV was growing over 1994-2013, and the spread of the disease had a West-East direction in the first period, i.e., 1994-2000, which was justifiable considering the location of Kermanshah, whereas this spread drew Northeast-Southwest over the second period, i.e., 2001-2007. It can be claimed that this spatial trend is due to an increase of the disease across the townships based in the Northwest (Sahneh and Songhor) and Southwest (Ghasr-e Shirin, Gilan-e Gharb, and Sarpol-e Zahab) of the province and the spread of the disease towards the four directions of North, South, West and East, which may be caused by the spread of HIV across the townships. It is noteworthy that the mean centers of all three periods were in Kermanshah County (Figure 2). Cluster analysis is of prominent significance in epidemiology (with the aim of identifying the accumulation of disease cases and studying the clusters that are statistically significant).
According to the results, the spatial distribution of HIV in Kermanshah Province is not by prevalence, and there is evidence indicating that significant clusters have been observed in this regard. As can be seen in Figure 3, the hot spots of HIV are in Kermanshah, Harsin, and Islam Abad Townships. Over 65% of the total population of the whole Kermanshah Province resides in these three townships, so they are of greater significance, especially Kermanshah County, the capital city of the province. It seems that lack of access to health care centers [37, 38] and the presence of high levels of immigration across Kermanshah [38] are playing leading roles in changing this County into a hot spot for HIV-stricken patients. The results of the present study were consistent with the previous studies in terms of the presence of clusters and hot spots of HIV. What the present work shares with other studies conducted worldwide, such as Brazil, Taiwan, America (Rhode Island), Africa, Russia and Iran, is the use of modern systems to detect and identify the hot spots and certain patterns of the disease [7, 19, 20, 39-45]. It is noteworthy that no modern systems have been comprehensively used yet for the diagnosis of HIV in Iran. Given the identification of the clusters of the disease across the province, one of the purposeful measures that the health policymakers can take is the continuation of studying the spatial trend of HIV across Kermanshah and Iran using GIS. The present study could identify the main hot spots of AIDS, high risk groups and transmission routes of this disease in Kermanshah Province. Moreover, as an effective study, the present research can help managers with future decisions and plans regarding the main hot spots of AIDS, screening at main regions of the province, defining new programs for the women’s group, and paying more attention to ways of transmitting AIDS (through sex) in the province to ultimately identify more cases of the disease and prevent its growth throughout the province.
The primary limitation of the present study was the patients’ incomplete and lost addresses, which made the research team exclude them from the study (missing data = 4%), which represents the fact that the data relating to such patients should be recorded precisely. On the other hand, one of the strengths of the present study was the examination of various courses of HIV’s spatial expansion, and additionally the presented data could help with the purposeful prevention of HIV in the studies performed by the Public Health Headquarters.

Conclusions

The present study aimed to analyze the epidemiology and spatial trends of HIV using GIS, by which the significant geographical areas were identified in terms of the prevalence of HIV. In the past 20 years, the dominant population infected with HIV was male. However, women have had upward trends in recent years in terms of being infected with HIV. Intravenous drug injection and sexual intercourse had downward and upward trends, respectively, which were reported the same in 2013. Over the past 20 years, some changes have been observed in age groups, i.e., the majority of infected people fall into the 25-34 age group and not the 35-44 one. The prevalence of HIV in Kermanshah Province increased, and the majority of HIV-stricken people were based in Kermanshah County in a way that the spread of the disease had a west-east direction in the first period (1994-2000), in the second period (2001-2007) the spread of the disease drew northeast-southwest, and in the third period (2007-2013) it progressed in the four directions of North, South, West, and East. Furthermore, the main hot spots were based in Kermanshah, Harsin, and Islam Abad Townships. Given the weakness of the information systems and the lack of adequate data and statistics in developing countries, these results can be used by managers for planning the health practices, intervention programs in specific areas of the disease (HIV) and defining the local management plans in the area under study.

Conflict of interest

The authors declare no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

References

1. Hariri S, McKenna MT. Epidemiology of human immunodeficiency virus in the United States. Clin Microbiol Rev 2007; 20: 478-488.
2. Rangel MC, Gavin L, Reed C, Fowler MG, Lee LM. Epidemiology of HIV and AIDS among adolescents and young adults in the United States. J Adolescent Health 2006; 39: 156-163.
3. World Health Organization. HIV/AIDS 2015; 2014. Available at: http://www.who.int/mediacentre/factsheets/fs360/en/.
4. Ministry of Health and Medical Education. Report of registered cases of HIV/AIDS in Iran; 2014. Available at: http://aids.behdasht.gov.ir/index.aspx?siteid=328&pageid=50511 (Accessed: 9.11.2015).
5. UNAIDS report on the global AIDS epidemic; 2012; Available at: http://www.unaids.org/sites/default/files/media_asset/20121120_UNAIDS_Global_Report_2012_with_annexes_en_1.pdf.
6. El-Sadr WM, Holmes CB, Mugyenyi P, Thirumurthy H, Eller­brock T,Ferris R, et al. Scale-up of HIV treatment through PEPFAR: a historic public health achievement.J Acquir Immune Defic Syndr 2012; 60 (Suppl 3): S96-S104.
7. Hacker MA, Leite I, Friedman SR, Carrijo RG, Bastos FI, et al. Poverty, bridging between injecting drug users and the general population, and “interiorization” may explain the spread of HIV in southern Brazil. Health Place 2009; 15: 514-519.
8. Pfeiffer D, Robinson TP, Stevenson M, Stevens KB, Rogers DJ, Clements ACA. Spatial Analysis in Epi­demiology. Oxford University Press, Oxford 2008.
9. Centers for Disease Control and Prevention. Revised Guidelines for HIV Counseling, Testing, and Referral and Revised Recommendations for HIV Screening of Pregnant Women. MMWR 2001; 50: 1-74. Available at: https://www.cdc.gov/mmwr/PDF/rr/rr5019.pdf.
10. Tanser F, Bärnighausen T, Cooke GS, Newell ML. Localized spatial clustering of HIV infections in a widely disseminated rural South African epidemic. Int J Epidemiol 2009; 38: 1008-1016.
11. Cromley E, McLafferty S. GIS and public health. Guilford Press, New York 2002.
12. Peng ZH, Cheng YJ, Reilly KH, Wang L, Qin QQ, Ding ZW, et al. Spatial distribution of HIV/AIDS in Yunnan province, People’s Republic of China. Geospat Health 2011; 5: 177-182.
13. Kalipeni E, Zulu L. Using GIS to model and forecast HIV/AIDS rates in Africa, 1986–2010. Professional Geographer 2008; 60: 33-53.
14. Messina JP, Emch M, Muwonga J, Mwandagalirwa K, Edidi SB, Mama N, et al. Spatial and socio-behavioral patterns of HIV prevalence in the Democratic Republic of Congo. Soc Sci Med 2010; 71: 1428-1435.
15. Cromley E, McLafferty S. Public participation GIS and community health. GIS and public health. Guilford Press, New York 2012; 411-422.
16. Bautista CT, Sateren WB, Sanchez JL, Singer DE, Scott P. Geographic mapping of HIV infection among civilian applicants for United States military service. Health Place 2008; 14: 608-615.
17. Jeefoo P. Spatial patterns analysis and hotspots of HIV/AIDS in Phayao Province, Thailand. Arch Des Sci 2012; 65: 37-50.
18. Martinez AN, Mobley LR, Lorvick J, Novak SP, Lopez AM, Kral AH. Spatial analysis of HIV positive injection drug users in San Francisco, 1987 to 2005. Int J Environ Res Public Health 2014; 11: 3937-3955.
19. Yang AC, Wen TH, Shih CC, Fang CT. Differentiating geographic patterns of human immunodeficiency virus (HIV) infection with different risk factors in northern Taiwan: 1997-2008. Applied Geo­graphy 2011; 31: 519-524.
20. Chan PA, Reitsma MB, DeLong A, Boucek B, Nunn A, Salemi M, et al. Phylogenetic and geospatial evaluation of HIV-1 subtype diver­sity at the largest HIV center in Rhode Island. Infect Genet Evol 2014; 28: 358-366.
21. Ghanbarnezhad A, Roustazadeh A, Alizadeh A, Abbastabar H, Nazar­nezhad M, Mohseni S. Spatial distribution of TB and HIV co-infection in South of Iran. AYER 2015; 4: 177-183.
22. Moradi G, Mohraz M, Gouya MM, Dejman M, Seyedalinaghi S, Khoshravesh S, et al. The view of point and perspective of policy makers about HIV/AIDS control policies in Iran – a qualitative study. J Sch Public Health Instit Public Health Res 2014; 12: 11-22.
23. Siam S. Survey of women’s knowledge about AIDS in Rasht 2006. J Guilan Univ Med Sci 2008; 17: 59-67.
24. Rostaei S, Ahadnejad Reveshty M, Asghari Zamani A, Zangenh A. Assessment of urban poverty spatial distribution in Kermanshah City. Social Welfare Quarterly 2012; 12: 77-101.
25. Mosayebi S, Taghdisi A. Spatial analysis and grading the employment indexes in Iran’s provinces (using the factor and cluster analysis techniques). Town and Country Planning 2014; 5: 361-382.
26. Ashrafinia F, Janani L, Khajeh Kazemi R, Dastoorpour M. The relationship between fear of AIDS with childbearing age women knowledge and attitude toward AIDS. Razi J Med Sci 2014; 20: 76-84.
27. Madise NJ, Ziraba AK, Inungu J, Khamadi SA, Ezeh A, Zulu EM, et al. Are slum dwellers at heightened risk of HIV infection than other urban residents? Evidence from population-based HIV prevalence surveys in Kenya. Health Place 2012; 18: 1144-1152.
28. Taiwo B, Li X, Palella F, Jacobson LP, Margolick JB, Detels R, et al. Higher risk of AIDS or death in patients with lower CD4 cell counts after virally suppressive HAART. HIV Med 2009; 10: 657-660.
29. Montazeri A. AIDS knowledge and attitudes in Iran: results from a population-based survey in Tehran. Patient Educ Couns 2005; 57: 199-203.
30. Salehi L, Salehi F, Shakibazadeh E. Education-based needs assessment: a step toward effective prevention of AIDS. J Qazvin Univ Med Sci 2009; 13: 74-78.
31. Azizi H, Saboory E, Ghaderi S. The study of prostitute women’s knowledge about ways of HIV transmission in Tehran in 1390. Journal of Urmia Nursing and Midwifery Faculty 2014; 11: 806-813.
32. Meshkati M, Hajari A, Mostajeran M, Nosratnejad S. Knowledge and attitudes of married women referred to health centers affiliated to Isfahan University of Medical Sciences, Iran, about AIDS and related factors. J Mazandaran Univ Med Sci 2014; 23: 100-107.
33. Amirpour M, Ghorbany M. Investigate the changing patterns of drug use from low risk to high risk. Social Welfare Quarterly 2013; 13: 201-228.
34. Mahmoudi M, Dehdari T, Shojaeezadeh D, Abbasian L. Coping with stress strategies in HIV-infected Iranian patients. J Assoc Nurses AIDS Care 2015; 26: 464-471.
35. Oster E. HIV and sexual behavior change: why not Africa? J Health Econ 2012; 31: 35-49.
36. Tahari MMH, Babaei MH, Morovati SA. Investigation and ranking of Iranian provinces in terms of access to health sector indicators. Health Information Management 2012; 9: 356-369.
37. Reshadat S, Saedi S, Zangeneh A, Ghasemi SR, Gilan NR, Karbasi A, et al. Spatial accessibility of the population to urban health centres in Kermanshah, Islamic Republic of Iran: a geographic information systems analysis. Eastern Mediterranean Health Journal 2015; 25: 389-395.
38. Iman M, Moradi GM. The examination of the relation between acculturation stress and mental health: Kermanshah as a case study. J Appl Soc 2009; 20: 147-170.
39. Cuadros DF, Awad SF, Abu-Raddad LJ. Mapping HIV clustering: a strategy for identifying populations at high risk of HIV infection in sub-Saharan Africa. Int J Health Geographics 2013; 12: 28.
40. Zulu LC, Kalipeni E, Johannes E. Analyzing spatial clustering and the spatiotemporal nature and trends of HIV/AIDS prevalence using GIS: the case of Malawi, 1994-2010. BMC Infect Dis 2014; 14: 285.
41. Ghosh J, Wadhwa V, Kalipeni E. Vulnerability to HIV/AIDS among women of reproductive age in the slums of Delhi and Hyderabad, India. Soc Sci Med 2009; 68: 638-642.
42. Palladino C, Bellón JM, Perez-Hoyos S, Resino R, Guillen S, Garcia D, et al. Spatial pattern of HIV-1 mother-to-child-transmission in Madrid (Spain) from 1980 till now: demographic and socioeconomic factors. AIDS 2008; 22: 2199-2205.
43. Tatem AJ, Hemelaar J, Gray RR, Salemi M. Spatial accessibility and the spread of HIV-1 subtypes and recombinants. AIDS 2012; 26: 2351-2360.
44. Heimer R, Barbour R, Shaboltas AV, Hoffman IF, Kozlovb AP. Spatial distribution of HIV prevalence and incidence among injection drugs users in St Petersburg: implications for HIV transmission. AIDS (London, England) 2008; 22: 123-130.
45. Karimi M, Niknami S. Self-efficacy and perceived benefits/barriers on the AIDS preventive behaviors. J Kermanshah Univ Med Sci 2011; 15: 384-392.