Knowledge‐Behavior Paradox ‐ Highest Knowledge Contributes to Positive Attitude Toward People Living With HIV, But Not to Condom Use: A Cross‐Sectional Study Among Medical Students in Belgrade (Serbia) Authors: Zoran Milosavljević, Tatjana Gazibara, Danijela Simić, Jelena Dotlić Published: 2026 ABSTRACT ABSTRACTBackground and AimIn Serbia, the highest incidence rates of sexually transmitted infections (STI) are recorded among young adults. Students enter medical studies with limited knowledge about STIs and HIV because there is an absence of reproductive health education in schools. The study aim was to evaluate the association of HIV and STI‐related knowledge with attitudes toward people living with HIV and condom use among medical students.MethodsThis cross‐sectional study was conducted in the spring of 2023. Medical students in the first and the fifth year at the Belgrade University completed an anonymous online questionnaire about knowledge of HIV/STIs, attitudes toward people living with HIV (PLWH) and condom use. Data were assessed using the adjusted linear, quantile, and logistic regression models.ResultsOverall, 402 students responded (229 in the first and 173 in the fifth year). The total knowledge score was higher among older students (median 12 vs. 9 out of 14, p = 0.001), but no difference in attitudes toward PLWH was observed (12 vs. 12 out of 19, p = 0.128). Knowledge score only at the 90th percentile was associated with stronger positive attitudes toward PLWH. No difference in condom use at last sexual intercourse was observed (62% in first year vs. 54% in fifth year, p = 0.184). Being female, not being partnered, not practicing anal sex, having some type of sex education, and not binge drinking were associated with condom use at last sexual intercourse.ConclusionOnly students who had the highest level of knowledge on STIs and HIV were more likely to foster positive attitudes toward PLWH. However, neither high knowledge nor positive attitudes increased condom use. Condom use remains suboptimal among medical students. It is recommended that some forms of informal sex education be introduced to secondary and higher education institutions.Keywords: attitudes, condom, HIV, knowledge, medical students Background The incidence of sexually transmitted infections (STIs) and human immunodeficiency virus (HIV) in Europe is higher in young adults compared to other age groups (European Center for Disease Prevention and Control ‐ ECDC [1, 2, 3, 4, 5]). In Serbia in 2023, 39% of the newly reported gonorrhea cases (incidence 6.60/100,000) were registered in people aged 20–29 years [6, 7]. This age group also contributed to the highest incidence of newly reported cases of chlamydia (32.28/100,000), syphilis (14.17/100,000), and HIV (6.46/100,000). Moreover, HIV and syphilis have been steadily increasing among men who have sex with men (MSM), and those aged 20–29 years accounted for 26% of all newly diagnosed HIV in 2023 [6, 7]. Bearing in mind the rise in HIV over the past decade among young people in Serbia, the issue of knowledge and behavior (i.e., sexual intercourse without condom) in this population group has become increasingly relevant. Over the past decade, the efforts of non‐governmental organizations to address HIV and STIs transmission were few and far between. As a result, it seems that there is not enough awareness about STIs among young people. Moreover, there is a lack of comprehensive sex education in primary and secondary schools, which further perpetuates insufficient condom use, lack of understanding of STIs, and increased attack rates among youth. In a report from 2021, Serbia together with Georgia, Bulgaria and Russian Federation were countries in the European region that did not report sexuality education [8], despite the evidence that this type of education is directly linked to the reduction in unwanted pregnancies among adolescents [9]. Although it is believed that sex education contributes to earlier sexual debut, this was not supported by the evidence [9]. Because of the lack of sexuality and reproductive education in Serbian schools, health care providers as well as policy makers are unaware of the STI knowledge level among young people, which directly affects the response to information needs. These factors combined are likely the main drivers of the unfavorable epidemiological situation regarding HIV and STIs in Serbia. To meet the standards of international HIV prevention programs, such as the Joint United Nations Program on HIV/Acquired Immunodeficiency Syndrome (UNAIDS) 95‐95‐95 by 2030 [10], there is a need for adequate sex education led by the experts. In this way, they can meaningfully contribute to the improvement of STI and HIV prevention programs as well as to the development of positive attitudes and destigmatization of people living with HIV (PLWH), not only in the community, but also in health institutions. Medical students represent the future taskforce in the prevention and control of diseases, including STIs and HIV. Thus, their knowledge and attitudes toward PLWH will play a major role in how the general public perceives STIs, HIV, and PLWH. However, evidence suggests that sexual health knowledge is insufficient among medical students worldwide [11, 12, 13, 14, 15]. A study from Serbia reported a higher knowledge of STIs among medical compared to non‐medical students, but insufficient condom use was identified in both student groups [16]. Knowledge on STIs has been previously studied in a linear manner. Analytical approach to knowledge relative to attitudes toward PLWH has seldom included the quantile regression models [17, 18]. By using the quantile regression approach, it is possible to better explain the interconnectedness of knowledge with attitudes toward PLWH. Finally, such data can offer valuable evidence on how to enhance education interventions to destigmatize PLWH. It is expected that over the course of undergraduate medical training students increase their knowledge and shape positive attitudes toward PLWH, but also consistently use condom. While previous studies compared older and younger medical students in terms of knowledge, attitudes, and practices about HIV [19], there is still a gap in understanding how knowledge can help to optimize preventive behaviors in conditions where no reproductive health education exist in the school system. Empirical data, such as these, are necessary to understand information needs among young people and ways to approach the issues surrounding STIs and HIV (e.g., through formal or informal education) to overcome stigma regarding sexual intercourse and condom use. Further efforts to destigmatize PLWH are needed both among medical students and the society at large to promote equality and justice. The purpose of this study was to compare STI‐related knowledge, attitudes toward people living with HIV, and condom use among students in the first and the fifth study year and to evaluate the association of HIV and STI‐related knowledge with attitudes toward PLWH and condom use among medical students. Methods In the preparation of this study, no AI or AI‐assisted tools have been used to facilitate research methodology, data collection, or analysis, or literature review processes. Study Setting This research was conducted among medical students at the University of Belgrade, the largest Faculty of Medicine in Serbia. Faculty of Medicine, University of Belgrade is one of five medical schools in the country (along with the University of Niš, University of Novi Sad, University of Kragujevac, and University of Priština in Kosovska Mitrovica). The University of Belgrade has approximately 90,000 students. As it is the largest institution of higher education in Serbia, it has more than 60% of enrollees outside of Belgrade. The University also accepts foreign students who study in the Serbian as well as in the English program [20]. Faculty of Medicine is one of the largest schools of the University of Belgrade. It has approximately 3000 students. To enter the Faculty, successful candidates must have a secondary education diploma and pass the entrance exam in chemistry and biology. The undergraduate studies last for 6 years. The first three study years focus mainly on pre‐clinical subjects (anatomy, histology, biochemistry, physiology, pathology, pharmacology). The last 3 years are focused on clinical subjects (internal medicine, surgery, gynecology, neurology, psychiatry, dermatology, ophthalmology, etc.). Exams consist of a written, practical, and oral part after which candidates receive the final grade, which ranges from 6 (pass) to 10 (excellent). If candidates do not pass the exam, they have to re‐take it in the following exam sessions. Study Population We invited all first and fifth year students of the Medical School of the University of Belgrade to participate in this research. There were 538 students enrolled in the first study year, and 489 were enrolled in the fifth study year (total population of 1,027 students). Assuming a prevalence of condom use at last sexual intercourse of 50% [21], with 5% precision, a design effect of 1, a 95% confidence level, we estimated a sample size of 400 participants for the survey (OpenEpi software version 3.1.). Participation in this research was voluntary and anonymous. Students consented to participation before filling in the questionnaire. Consent form was placed on the first page of the questionnaire when students willing to participate in the study attempted to access the questionnaire via an online link. Without checking the consent form, participants could not access the content of the questionnaire. Approval to conduct the study was granted by the Ethics Committee of the Faculty of Medicine University of Belgrade (approval no. 17/1‐41 issued on January 12, 2023). Study Design A cross‐sectional study was conducted during the spring semester, from March 1 to May 1, 2023. The cross‐sectional study is especially convenient to study knowledge, attitudes and behaviors, because it offers a snapshot of the circumstances at one point in time. This is suitable for studies on STIs [22, 23, 24]. Cross‐sectional studies are also convenient to perform repeated research at different points in time to compare the situation and evaluate the effectiveness of potential educational interventions. Cross‐sectional studies are relatively inexpensive and can be conducted in a short amount of time. However, the disadvantage is that the exposures and outcomes are recorded at the same time, so causal inference is limited [25]. Data Collection Participants completed an anonymous online questionnaire. The questionnaire was distributed through the web link associated with the online learning platform on the Faculty of Medicine University of Belgrade, which enabled respondents to be anonymous. Only the principal investigator of this study (Z. M.) had access to the questionnaire link and did data handling using a password. The questionnaire was developed purposefully for this particular study. The study was conducted within the framework of the Mediterranean and Black Sea Program for Intervention Epidemiology Training (MediPIET) program, where the principal investigator Z. M. was enrolled. The MediPIET program was run by the ECDC to improve capacity in field Epidemiology in Europe and the Mediterranean region. Three epidemiologists and one gynecologist comprised the expert team who developed the questionnaire and were responsible for the accuracy of questionnaire items. Members of the team had previous experience with HIV questionnaire development which served as the base to generate questionnaire items [21, 26, 27]. As responses to questionnaire items in knowledge, attitude, and behaviors part of the questionnaire were not uniform, it was not possible to conduct psychometric testing except face validity check. Questionnaire Structure The first section included demographic characteristics: gender, age, place of residence before university, current living arrangement, relationship status, sexual orientation, and peer education on sexual/reproductive health. The second section was related to general knowledge about HIV and other STIs. Six items about HIV included knowledge of population groups at higher risk of HIV, modes of transmission, window period, antiretroviral therapy, and pre‐exposure prophylaxis‐PrEP). Eight items about STIs included knowledge about HPV‐associated cancer, age of vaccination against HPV, infertility‐associated with different STIs, transmission and clinical presentation of syphilis, antimicrobial resistance and gonorrhea, and testing of STIs. Each correct answer was awarded one point. We calculated the knowledge score as the sum of points on each knowledge‐related question. The HIV knowledge score ranged from 0 to 6, the STI knowledge score ranged from 0 to 8, and the total combined HIV and STI knowledge score ranged from 0 to 14. The third section covered items about students’ attitudes toward PLWH. Nine items were related to issues such as having PLWH as friends, colleagues at work, promiscuity of PLWH, sharing patients’ HIV status with colleagues, examining PLWH in separate spaces, dividing patients based on their HIV status, living in the same household, and recommending post‐exposure prophylaxis—PEP when practicing unsafe sex. Each answer which supported a positive attitude was awarded 2 points, undecided responses were awarded 1 point, and negative answers were allocated 0 points. For the item “living in the same household,” we used the following grading: yes, any PLWH—3 points; yes, but family PLWH only—2 points; undecided—1 point; no—0 points. We summed the points for each response. The attitude score ranged from 0 to 19. The fourth section included the information about taking the HIV test, number of sexual partners, practicing anal sex, condom use at last sexual intercourse, having sex under influence of alcohol, binge drinking (defined as the intake of 5 or more drinks on one occasion), having STI symptoms in the past 12 months, and action after having had sex with a potential PLWH (using PrEP, using PEP, taking the HIV test, or none of these). Data Analysis As per the Kolmogorov–Smirnov test, continuous variables (age, STI knowledge score, HIV knowledge score, total STI‐HIV knowledge score, attitude score) were not normally distributed. Because of this, they were presented as medians with corresponding interquartile ranges (IQR). The Mann–Whitney test was used to identify differences between these variables. Also, we used the χ 2 (two‐sided) and the χ 2 linear‐by‐linear association to test the differences between categorical variables for 2 and > 2 categories, respectively. In case there were < 5 observations per cell, we used Fisher's exact test (two‐sided). We calculated the prevalence ratio (PR) by dividing the percentage of correct answers in the fifth and in the first year according to each question. To identify factors associated with a higher knowledge score, we used three multiple linear regression models. The dependent variable was the HIV knowledge score in the first model, the STI knowledge score in the second model, and the total HIV and STI combined knowledge score in the third model. The independent variables were all other items collected by the anonymous questionnaire: gender, age, sexual orientation, current living arrangement, size of the city before university, being partnered, practicing anal sex, having had sexual education, wanting additional sexual education, prior HIV testing, number of sexual partners in the past 12 months, STI symptoms in the past 12 months, deciding to use PEP after unprotected sex with potential PLWH and frequency of binge drinking. Furthermore, in another multiple linear regression model, which sought to identify factors associated with stronger positive attitudes toward PLWH, the dependent variable was the total attitude score. The independent variables were all those tested in the knowledge model plus the total HIV and STI knowledge score. To better understand the association between STI/HIV knowledge level and attitudes toward PLWH, the quantile regression model was applied. This model explored whether knowledge at any level was associated with positive attitudes toward PLWH, as previously observed [17]. The dependent variable in the model was the total attitude score, the independent variable was the knowledge score at 30th, 60th and 90th percentile of the knowledge score distribution. At each of the three selected percentiles of knowledge score, the covariates accounted for demographic and behavioral characteristics of study participants described in the previous linear regression model. To identify factors associated with condom use at last sexual intercourse (dichotomous outcome), we calculated odds ratios (ORs) and 95% confidence intervals (95% CIs), using multiple logistic regression. The independent variables were all those included in the attitude model plus the attitude score toward PLWH. Probability level of p < 0.05 was statistically significant. Data were analyzed in the Statistical Package for Social Sciences (SPSS), version 27.0. Results This cross‐sectional study conducted from March 1, to May 1, 2023 presents the results on knowledge, attitudes and behaviors of medical students in the first and the fifth study year. Out of 1027 students in both study years, 39.1% (402/1027) participated in the study overall. 42.5% of all students (229/538) in the first year and 35.4% of all students (173/489) in the fifth year. The median age of the participants was 20 (maximum age range 18–30) years. Median age among first year students was 19 (maximum age range 18–29) years for and median age for fifth year students was 23 (maximum age range 22–30). Of all participants, 292 (72.6%, 292/402), were female; 364 (91%, 364/402) reported being heterosexual (Table 1); 170 (42%, 170/402) lived with their parents and 195 (49%, 195/402) lived in cities with more than 100,000 inhabitants before coming to study at the university. Overall, 276 (69%, 276/402) were not in a relationship and 248 (62%, 248/402) had some form of sex education in high school; 300 (75%, 300/402) expressed willingness to receive additional sex education (Table 1). Table 1Demographic and behavioral characteristics of the study sample.Total study sampleVariable categoriesTotal sample N = 402First year N = 229Fifth year N = 173 p n % n % n % SexMale10927.16427.94526.00.734Female29372.916572.112874.0Age* 20 (19–23)19 (18–19)23 (23–24)0.001Sexual orientationHeterosexual36490.520790.415790.80.667Bisexual225.5125.2105.8Gay143.583.563.5Other20.520.900Current living arrangementWith parents17042.29441.07644.00.954Housemate(s)8019.95122.02917.0With partner205.063.0148.0Dormitory5914.73816.52112.0Alone7318.24017.53319.0Size of town before Uni> 100.00019548.511148.58448.50.30910.000–100.00014636.37633.27040.5< 10.0006115.24218.31911.0PartneredYes12631.35524.07141.00.001Had SexEdYes24861.715065.59856.6%0.071Wants SexEdYes30074.617576.412572.20.342Prior HIV testingNever36190.022397.413980.30.001More than a year215.331.31810.5In the past year194.731.3169.2No. of sex partners in past 12 monthsNone14836.811851.53017.30.001119147.58938.910259.02–55313.2198.33419.76–1082.020.963.5> 1020.510.410.6Participants who reported sex in the past 12 monthsTotal sample N = 254First year N = 111Fifth year N = 143VariableCategories n % n % n % p Anal sexNever21183.19686.511580.40.185Sometimes3614.21311.72316.1Always72.821.853.5Condom last sexYes14657.56962.27753.80.184Sex under alcohYes8633.94036.04632.20.518Binge drinkingYes16565.07769.48861.50.194Frequency of binge drinking08734.33329.75437.80.2061218.398.1128.423011.81513.51510.53166.387.285.64–5259.887.21711.9> 57529.53834.23725.9STI in past yearyes103.910.996.30.046Get PEP after unprotected sexyes14557.14540.510069.90.001Abbreviations: alcoh, alcohol; PEP, post‐exposure prophylaxis; PLWH, people living with HIV; SexEd, sexual and reproductive education; STI, sexually transmitted infections; Uni‐University.*Median (interquartile range). Demographic and behavioral characteristics of the study sample. Abbreviations: alcoh, alcohol; PEP, post‐exposure prophylaxis; PLWH, people living with HIV; SexEd, sexual and reproductive education; STI, sexually transmitted infections; Uni‐University. Median (interquartile range). Knowledge Students in the first study year had a median HIV knowledge score of 4 (IQR 3–6), while students in the fifth study year had a median of 5 (IQR 4–6) out of a maximum of 6 (p = 0.001). Similarly, younger students had a median STIs knowledge score of 5 (IQR 4–7) and older students had 7 (IQR 6–8) out of a maximum of 8 (p = 0.001). The total knowledge score was higher among the fifth study year students (median 12 vs. 9 out of 14, p = 0.001). In the first study year, females scored higher on STI knowledge compared to males (median 4 vs. 5, p = 0.017). This pattern was also observed in the fifth study year (median 6 vs. 7, p = 0.034). Students in the fifth year were more likely to answer correctly all questions related to HIV knowledge. The proportion of fifth‐year students who were aware of the use of PrEP in HIV prevention was almost three times higher compared to first‐year students (PR = 2.92; 95% CI: 2.7–3.1, p = 0.001). Also, fifth‐year students were more likely to answer correctly to all HIV knowledge‐related questions except for the modes of HIV transmission (PR = 1.0, 95% CI: 0.9–1.1, p = 0.890) (Table 2). Table 2Proportion of correct responses to knowledge items and of positive attitudes toward people living with HIV (N = 402).ItemsStudy yearFifth to first year prevalence ratio p First yearFifth year n N tot % n N tot % HIV knowledge 1. Key populations at high risk for HIV20322988.617017398.31.1 (1.0–1.2)0.0012. HIV transmission and body fluids22622998.717117398.81.0 (0.9–1.1)0.8903. Concept of “window” period13122957.215917391.91.6 (1.4–1.7)0.0014. No HIV transmission if viral load undetectable5822925.310317359.52.35 (2.1–2.5)0.0015. Antiretroviral meds adherence18522980.816117393.11.15 (1.0–1.2)0.0016. Biomedical HIV prevention (PrEP, PEP)6222927.113717379.22.92 (2.7–3.1)0.001 STI knowledge 1. HPV and malignant tumors19722986.016717396.51.12 (0.9–1.2)0.0012. HPV vaccination in Serbia free of charge19622985.616617396.01.12 (0.9–1.2)0.0013. Candidiasis not causing infertility10522945.914117381.51.77 (1.6–1.9)0.0014. Chlamydia infection as asymptomatic infection11022948.015217387.91.83 (1.6–2.0)0.0015. Three stages of syphilis12622955.017017398.31.78 (1.6–2.0)0.0016. Highest contagiousness of secondary syphilis6222927.16217335.81.32 (1.1–1.5)0.0607. Gonococcal antimicrobial resistance12122952.811617367.11.27 (1.1–1.5)0.0048. Asymptomatic STIs and testing19122983.415817391.31.09 (1.0–1.2)0.022 Attitudes toward people living with HIV 1. Be friends with PLWH, yes16622972.514317382.71.14 (1.0–1.2)0.0172. Working with PLWH colleague, yes17022974.213517378.01.05 (1.0–1.2)0.3793. All PLWH have multiple partners, no20422989.116117393.11.04 (1.0–1.2)0.1754. Sharing patients’ HIV status, no10122944.17417342.80.97 (1.0–1.2)0.7905. Examining PLWH in separate spaces, no20122994.816717396.51.01 (0.9–1.1)0.3986. Discomfort when examining PLWH, no17522976.412217370.50.92 (0.8–1.0)0.1837. Separate patients based on HIV status, no13022956.813217376.31.34 (1.2–1.4)0.0018a. Living in the same household with PLWH, yes anyone7622933.25617332.40.97 (0.8–1.0)0.8638b. Living in the same household with PLWH, yes family only10322945.08017346.21.02 (0.9–1.1)0.8019. PrEP recommendation for unsafe sex, yes13722959.89817356.60.94 (0.8–1.0)0.522Abbreviations: CI, confidence interval; n, number of students who answered correctly/number of students who had positive opinion; N tot, total number of students; PLWH, people living with HIV; PrEP, pre‐exposure prophylaxis; STI, sexually transmitted infections; %, percentage of correct responses. Proportion of correct responses to knowledge items and of positive attitudes toward people living with HIV (N = 402). Abbreviations: CI, confidence interval; n, number of students who answered correctly/number of students who had positive opinion; N tot, total number of students; PLWH, people living with HIV; PrEP, pre‐exposure prophylaxis; STI, sexually transmitted infections; %, percentage of correct responses. Regarding other STIs, the proportion of fifth‐year students who were aware of chlamydiosis as an asymptomatic infection was almost two times higher (PR = 1.83; 95% CI: 1.6–2.0, p = 0.001) compared with the first‐year students (Table 2). The lowest proportions of correct answers in both study years were found for the highest contagiousness of syphilis in the second stage. Linear Regression Models Students who were older (B = 0.17, 95% CI: 0.13, 0.22, p = 0.001), who had taken an HIV test before (B = 0.33, 95% CI: 0.06, 0.59, p = 0.015) and who would use PEP after unsafe sex (B = 0.48, 95% CI: 0.26, 0.71, p = 0.001) were more likely to score higher on HIV knowledge (Table 3). Table 3Factors associated with higher HIV knowledge score, STI knowledge score, and total knowledge score (N = 402).VariablesKnowledge scores as outcomesHIV knowledge scoreSTI knowledge scoreTotal knowledge score (HIV + STI) B 95% CI p B 95% CI p B 95% CI p Sex0.02−0.24, 0.280.8830.640.33, 0.960.0010.650.17, 1.160.008Age0.170.13, 0.220.0010.180.13, 0.240.0010.360.28, 0.450.001Sex orientation−0.12−0.54, 0.310.589−0.27−0.79, 0.250.305−0.37−1.19, 0.410.343Living arrangement−0.04−0.21, 0.130.659−0.16−0.37, 0.440.121−0.23−0.57, 0.120.216City size before Uni0.09−0.08, 0.260.3110.09−0.12, 0.300.3950.18−0.14, 0.500.278Being partnered0.17−0.07, 0.420.1650.27−0.03, 0.580.0780.45−0.02, 0.920.061Anal sex0.15−0.19, 0.500.396−0.04−0.47, 0.380.8370.10−0.55, 0.750.753Prior SexEd0.19−0.04, 0.420.1080.320.04, 0.610.0260.510.07, 0.950.022Wants SexEd−0.10−0.36, 0.150.432−0.40−0.72, −0.080.012−0.49−0.97, −0.010.047Prior HIV test0.330.06, 0.590.0150.410.09, 0.740.0130.740.24, 1.240.004No. of partners last year0.10−0.06, 0.270.2210.280.08, 0.500.0070.390.07, 0.710.016STI last year−0.23−0.89, 0.420.4930.28−0.52, 1.080.4960.05−1.19, 1.290.936PEP use after unsafe sex0.480.26, 0.710.0010.440.16, 0.730.0020.930.50, 1.360.001Binge drinking0.03−0.02, 0.080.3000.02−0.05, 0.080.6180.05−0.05, 0.150.385Abbreviations: B, linear regression coefficient; CI, confidence interval; mo, months; PEP, post‐exposure prophylaxis; Sex Ed, sexual education; STI, sexually transmitted infections; Uni, university. Factors associated with higher HIV knowledge score, STI knowledge score, and total knowledge score (N = 402). Abbreviations: B, linear regression coefficient; CI, confidence interval; mo, months; PEP, post‐exposure prophylaxis; Sex Ed, sexual education; STI, sexually transmitted infections; Uni, university. Students who were female (B = 0.64, 95% CI: 0.33, 0.96, p = 0.001), older (B = 0.18, 95% CI: 0.13, 0.24, p = 0.001), had prior sex education (B = 0.32, 95% CI: 0.04, 0.61, p = 0.026), did not want additional sex education (B = −0.40, 95% CI: −0.71, −0.08, p = 0.012), who had taken an HIV test before (B = 0.41, 95% CI: 0.09, 0.74, p = 0.013), who had more sexual partners in the past 12 months (B = 0.28, 95% CI: 0.08, 0.50, p = 0.007) or who would use PEP after unsafe sex (B = 0.44, 95% CI: 0.16, 0.73, p = 0.002) were more likely to score higher on STIs knowledge. These factors were also associated with a higher total HIV and STIs knowledge score (Table 3). Attitudes Students in the first study year had a median attitude score of 12 (IQR: 11–13) just as the students in the fifth study year (median of 12, IQR: 10–14) out of a maximum of 19 (p = 0.516). No differences in attitude score were observed between males and females in the first study year (12 vs. 12, p = 0.128). However, a difference in attitude score was found between males and females in the fifth study year, where males had a stronger positive score compared with females (13 vs. 12, p = 0.047). A difference in attitude score was also observed between heterosexual versus other students (12 vs. 14, p = 0.001). When answers of students about attitudes toward PLWH were observed, most attitudes did not differ between the first and the fifth study year. However, students in the fifth year were 14% (PR = 1.14, 95% CI: 1.0–1.2, p = 0.017) more likely to be friends with PLWH and 34% (PR = 1.34, 95% CI: 1.2–1.4, p = 0.001) times more likely to separate PLWH patients from others in health care institutions compared with the first‐year students (Table 2). Linear Regression Model In the multiple linear regression model, students who reported themselves as non‐heterosexual (B = 1.12, 95% CI: 0.29, 1.95, p = 0.008) and those who would use PEP after unsafe sex (B = 0.52, 95% CI: 0.07, 0.98, p = 0.025) were more likely to have stronger positive attitudes toward PLWH (Table 4). Table 4Factors associated with a stronger positive attitude toward people living with HIV (N = 402).VariableCategory B 95% Cl p SexMale vs. female0.18−0.33 to 0.700.490Age−0.05−0.15 to 0.050.313Sexual orientationHeterosexual vs. other1.120.29 to 1.950.008Living arrangementWith parents vs. other−0.08−0.42 to 0.250.619City size before Uni−0.01−0.34 to 0.330.978Being partneredYes vs. no0.09−0.40 to 0.580.718Practicing anal sexYes vs. no0.38−0.30 to 1.050.272Prior Sex EdYes vs. no0.22−0.233 to 0.690.336Wanting Sex EdYes vs. no0.10−0.40 to 0.600.695Prior HIV testingYes vs. no0.41−0.11 to 0.940.121No. of sex partners in past 12 months−0.13−0.50 to 0.200.429STI in past 12 monthsYes vs. no0.18−1.10 to 1.500.787PEP after unsafe sexYes vs. no0.520.07 to 0.980.025Total HIV and STI knowledge score0.048−0.05 to 0.150.363Frequency of binge drinking0.05−0.05 to 0.160.332Abbreviations: B, linear regression coefficient; CI, confidence interval; PEP, post‐exposure prophylaxis; STI, sexually transmitted infections. Factors associated with a stronger positive attitude toward people living with HIV (N = 402). Abbreviations: B, linear regression coefficient; CI, confidence interval; PEP, post‐exposure prophylaxis; STI, sexually transmitted infections. Quantile Regression Models As there was an absence of the association between knowledge score and attitudes toward PLWH, the quantile regression was conducted to understand whether any level of knowledge is associated with positive attitude toward PLWH. The results of the multiple quantile regression model are presented in Table 5. The quantile regression model suggested that knowledge level of STIs and HIV at 30th percentile and 60th percentile was not associated with a positive attitude. However, knowledge of STIs and HIV at the 90th percentile was associated with a stronger positive attitude toward PLWH (B = 0.12, 95% CI: 0.03–0.20), p = 0.001). This means that only the highest level of knowledge allows for the development of a positive attitude toward PLWH among students. Table 5Results of the adjusted quantile regression models: Association of knowledge score at 30th, 60th, and 90th percentile with positive attitudes toward people living with HIV (N = 402).PredictorOutcome: Stronger positive attitude toward people living with HIVKnowledge score on STIs and HIV 30th percentile B (95% CI) 0.11 (−0.10, 0.23), p = 0.173 60th percentile B (95% CI) 0.01 (−0.12, 0.13), p = 0.247 90th percentile B (95% CI) 0.12 (0.03, 0.20), p = 0.001 Note: Each model examining specific percentile of knowledge (30th, 60th, and 90th) was adjusted for sex, age, sexual orientation, current living arrangement, size of the city before University, being partnered, practicing anal sex, prior sex education, wanting additional sex education, prior HIV testing, number of sexual partners in the past 12 months, STI symptoms in the past 12 months, post‐exposure prophylaxis use after unsafe sex, frequency of binge drinking.Abbreviations: B, quantile regression coefficient; CI, confidence interval; STI, sexually transmitted infections. Results of the adjusted quantile regression models: Association of knowledge score at 30th, 60th, and 90th percentile with positive attitudes toward people living with HIV (N = 402). 30th percentile B (95% CI) 60th percentile B (95% CI) 90th percentile B (95% CI) Note: Each model examining specific percentile of knowledge (30th, 60th, and 90th) was adjusted for sex, age, sexual orientation, current living arrangement, size of the city before University, being partnered, practicing anal sex, prior sex education, wanting additional sex education, prior HIV testing, number of sexual partners in the past 12 months, STI symptoms in the past 12 months, post‐exposure prophylaxis use after unsafe sex, frequency of binge drinking. Abbreviations: B, quantile regression coefficient; CI, confidence interval; STI, sexually transmitted infections. Practices Of all students, 10% (40/402) had ever taken an HIV test. Of those who engaged in sexual activity in the past 12 months, 75.2% (191/254) had one partner and 16.9% (43/254) practiced occasionally or always anal sex. Of 254 students who reported sexual activity in the past 12 months, 146 (57.5%, 146/254) reported condom use at last sexual intercourse. Among them, no significant difference in condom use was observed between the first‐ and the fifth‐year students (62.2% [69/111] vs. 53.8% [77/143], p = 0.184). Similarly, we observed no difference in condom use between males and females in the first and the fifth year. Of sexually active students in the past 12 months, 86 (33.8%, 86/254) reported having sex under the influence of alcohol. Binge drinking was reported by 65% (165/254) of students. Ten participants (3.9%, 10/254) reported symptoms of an STI in the past 12 months. About one‐half (57.1%, 145/254) of students reported that they would use PEP after unprotected sex (Table 1). In the multiple logistic regression model, being female (odds ratio [OR] = 1.63, 95% CI: 1.22–1.83, p = 0.010), not being partnered (OR = 0.55, 95% CI: 0.30–0.99, p = 0.047), not practicing anal sex (OR = 0.23, 95% CI: 0.10–0.53, p = 0.001), having sex education (OR = 1.73, 95% CI: 1.01–3.11, p = 0.045) and avoiding binge drinking (OR = 0.81, 95% CI: 0.68–0.95, p = 0.010) were independent factors associated with condom use at last sexual intercourse (Table 6). Table 6Factors associated with condom use during the last sexual intercourse in a subsample of students who reported having sexual intercourse in the past 12 months (N = 254).VariableCategoriesOR95% CI p SexMale vs. female1.631.22–1.830.010Age1.010.90–1.120.841Sexual orientationHeterosexual vs. other1.880.53–6.580.323Living arrangementWith parents vs. other1.010.65–1.560.957Size of the city before University1.210.76–1.910.406Being partneredYes vs. no0.550.30–0.990.047Practicing anal sexYes vs. no0.230.10–0.530.001Prior sexual educationYes vs. no1.731.01– 3.110.045Wanting additional sexual educationYes vs. no1.620.84–3.150.149Total HIV and STI knowledge score0.960.83–1.100.581Attitude score toward PLWH1.040.92–1.190.464Prior HIV testing1.150.62–2.130.642No. of sex partners in past 12 months1.210.65–2.230.534Having sex under the influence of alcoholYes vs. no0.620.31–1.240.184Frequency of binge drinking 5 or more units of alcohol in the last 12 months0.810.68–0.950.010STI symptoms in past 12 months0.830.17–3.840.812Abbreviations: CI, confidence interval; OR, odds ratio; PEP, post‐exposure prophylaxis; STI, sexually transmitted infections. Factors associated with condom use during the last sexual intercourse in a subsample of students who reported having sexual intercourse in the past 12 months (N = 254). Abbreviations: CI, confidence interval; OR, odds ratio; PEP, post‐exposure prophylaxis; STI, sexually transmitted infections. Discussion The study aimed to explore STI‐related knowledge, attitudes toward people living with HIV, and condom use among students in the first and the fifth study year, and to evaluate the association of HIV and STI‐related knowledge with attitudes toward PLWH and condom use among medical students. The results suggested that medical students from the Belgrade University had rather good knowledge of STIs and HIV, with older students having consistently better knowledge than younger students. Attitudes toward PLWH were generally positive, but there were few participants who reported negative opinions. While median attitude score was the same among younger and older students, some differences in attitudes may be explained by medical training. Condom use was consistently low among medical students in Serbia as previously seen in other universities [21]. This study is one of the rare ones to employ the quantile regression, which allowed for a detailed observation on how knowledge level shapes attitudes toward PLWH. As a result, it was evident that only high STI/HIV knowledge level predisposed positive attitudes toward PLWHš, and having any type of sex education increased condom use. These findings call for the introduction of systematic, tailored education for youth to increase knowledge of STIs and HIV to foster positive attitudes toward PLWH and consistent condom use. A higher knowledge on HIV and STIs among medical students in Belgrade was associated not only with their age, but also with their gender and personal experiences related to HIV testing or PEP use after unsafe sex. Thus, students’ knowledge level directly stemmed from prior engagement in preventive practices. Female students, those with more sexual partners, and those who previously did take the HIV test were more knowledgeable about STIs in general. Consistent with our study findings, a similar study among medical school students in Belgrade reported that females had a better knowledge score on STIs compared with males [26]. Our study suggested that there was no difference in knowledge on HIV transmission between the first and the fifth‐year students, which contrasts the findings from previous studies [19]. Previous studies concluded that sexual and reproductive education is necessary before university schooling, that is, in high schools and even in primary schools, are expectedly to improve knowledge and condom use compliance in university students [19, 26, 28]. The results of our study support this notion, because as seen in the quantile regression, only students with the highest level of knowledge on STIs/HIV have the competence to understand the issues surrounding PLWH and, therefore, transform attitudes into positive relationship with PLWH. Furthermore, non‐formal education, such as peer education about reproductive health and sex, increases the likelihood of condom use, as seen among our medical students. That is to say that education is pivotal for attitudes and practices regarding STIs and HIV. The fifth‐year students had overall positive attitudes toward PLWH, especially when the issue of unequal treatment of PLWH in medical institutions and befriending PLWH was addressed. This is in line with some other studies from Serbia, which identified that older, fifth‐year students had stronger positive attitudes toward PLWH [19, 27]. These findings are encouraging, suggesting that the reduction of institutional stigma in medical settings is an achievable goal [29]. Having had an adequate medical training in school, students could be a good resource to destigmatize medical environment, which can be achieved through knowledge building, solidarity, and leadership by own example. The present study also suggested that students who are gay or bisexual are more prone to express positive attitudes toward PLWH. The students from sexual minority groups and those who were aware of HIV post‐exposure prophylaxis after unsafe sex are expectedly more sensitized to the issues that PLWH encounter. One of the explanations could be that they have probably met more PLWH than heterosexual students in their everyday life since the majority of PLWH in Serbia are coming from the social group of men who have sex with men, particularly over the last decade. Those students could also be a good resource for peer education of their fellow colleagues. Finally, it should be acknowledged again that the highest level of knowledge on STIs and HIV is the only knowledge level among our medical students that could be meaningful in efforts to foster a positive attitude toward PLWH. More than one‐half of students used alcohol when having sex, which is a relatively high proportion, with binge drinking being common in this population [21]. It is well known that increased alcohol intake facilitates sexual activity, which consequently increases the risk of STIs and HIV. For example, a large cohort study in Denmark [30] reported that there is a dose‐response relationship between the frequency of binge drinking episodes and risk of HPV infection, that is, the risk of having genital warts increases with the increasing frequency of binge drinking in the past month. Nevertheless, the present study indicated that single female students who did not binge drink were more likely to use condoms as a conventional means of protection against STIs. This is in line with a study from Sweden [12], which concluded that use of condoms among 18–30 years‐old adolescent males was not associated with increased risk perception while that of the adolescent females was. Our study also provided evidence that there was no difference in condom use between students who are older and in the fifth year compared with younger or first‐year students, and no difference in condom use between males and females. This has been observed previously in the student population [21] and remains a public issue among youth in Serbia to date, despite its wide availability in grocery stores, drugstores, and pharmacies as a pack of 3–12 condoms. Users need to pay for condoms out‐of‐pocket, from 3 Euros to 12 Euros per pack, depending on the size of the pack, so the costs may be an underlying issue as to why students do not use condoms often. Empirical data suggest that students often perceive steady relationship as a reason not to use condom [21], or they perceive low risk of contracting STIs [31]. As no systematic school‐based reproductive health education exists in Serbian primary and secondary schools, the low condom use is not, therefore, surprising. Nevertheless, young people keep searching for information about STIs and sex on the Internet [32, 33]. These findings call for continuous work in the public health domain to increase condom use not only for STI prevention, but also in the prevention of unwanted pregnancies [34, 35] and potential subsidies for condoms for young people or distribution of free condoms around student dormitories. However, the finding that having peer education or any type of informal sex education is associated with condom use highlights the importance of structured education about sexual and reproductive health. Bearing in mind this result, having education targeting youth is indispensable to foster reproductive and public health. The vast majority of medical students (90%) have never had an HIV test. This finding mirrors the results of other studies from Europe [36]. Our results could be explained by self‐perceived low risk from catching an STI [31, 37]. Despite the fact that in Belgrade there are several centers where people can take an HIV/STI test, utilization rates of those services among medical students were insufficient. Continuous work with students on and off lectures, workshops, and peer‐education on the relevance of testing, counseling, and new forms of prevention is of paramount importance, to increase compliance with HIV and STI testing. When comparing the results of our study to other studies conducted in Europe, there are many similarities between populations. For example, a recent study in northern Italy suggested that around one‐half of medical students had partial knowledge about HIV and that stigma still persisted [38]. In line with the results of our study, previous sex education was associated with a higher knowledge level among Italian medical students [38]. In Greece, medical students showed a high knowledge level of STIs [39], but among medical students in Croatia it has been observed that participants did not have comprehensive knowledge of HIV transmission which affected their attitudes toward PLWH [40]. Overestimation of risk for catching HIV and misconceptions about HIV has been reported among Polish [41, 42] and Finnish medical students [43] as well. However, positive attitude and low discrimination of PLWH was also observed among medical students in Bosnia and Herzegovina [44]. Medical students in other populations, such as Portugal, also have a low prevalence of condom use despite the knowledge of HIV [45]. Therefore, across Europe, there seems to be a continuing need for the STI and HIV‐related knowledge boost coupled with the promotion of condom use. Limitations and Strengths This study had some limitations. First, this study only included students from the largest medical school in Serbia. Therefore, our results might not be generalizable to all medical students in the country. Second, studies on sexual behavior are inherently open to information bias, and responses recorded in this study might not entirely represent the objective real‐life circumstances. The questionnaire was developed for the purpose of this study following similar research conducted previously in Serbia. However, no specific underlying theoretical foundation was used. Thirdly, the estimated associations were based on a cross‐sectional study design, which limits causal inference. All population‐based surveys are limited by time and the length of the questionnaire, so it is possible that unobserved confounding might occur. This research comprised around 40% of the target population of medical students at the University of Belgrade. Thus, our results could be generalized to medical students at the Faculty. Furthermore, the strength of this study is a relatively objective snapshot of the circumstances regarding STI knowledge and behaviors that are comparable to reports from other universities in Serbia. This means that our results could be used as the baseline for optimization of STI education not only among medical students, but also among students in other disciplines and secondary school students as well. We had the opportunity to observe nuanced differences between younger and older students, but also identify clear evidence that only the highest knowledge represents the real knowledge on STI and HIV. This suggests how complex knowledge‐behaviors axis is from a prevention and public health standpoint. As a result, STI prevention should be multisectorial and include not only public health professionals, epidemiologists, and physicians, but also social scientists. Implications for Policy and Practice Based on the results of this study, only the highest knowledge has the potential to transform opinions into positive attitudes toward PLWH. Despite this, even the highest knowledge was not associated with consistent condom use in the past year. However, informal reproductive health education was found as the contributor to condom use. Bearing these results in mind, it is recommended that some forms of informal peer education or ad hoc workshops/seminars/lectures are introduced into formal higher education. Also, these efforts should include adolescents in secondary education institutions, preferably before sexual debut. Informal education should embrace practical demonstration on condom use, to enhance confidence in condom use skills among youth. As STI prevention is multidimensional, expertise of social science scholars should be integrated into these programs to improve young people's skills in condom negotiation. At the institutional and community levels continuous efforts are needed to destigmatize PLWH, sexual practices, and condom use itself. Conclusions Only the highest level of knowledge was associated with positive attitudes toward PLWH. Students who belonged to sexual minorities had stronger positive attitudes toward PLWH, and they could be a good resource for peer education and sensitization of other students. The prevalence of condom use was suboptimal among medical students, with only a half of them reporting condom use. Being female, having had sex education, not being partnered, not practicing anal sex, and not binge drinking were associated with condom use. Strategies to increase condom use and enhance condom negotiation among women along with the avoidance of binge drinking are necessary to optimize the prevention of STIs spread. Prioritization of healthy behaviors is essential to foster mental and physical health among young adults. Activities to increase knowledge on STIs and HIV are fundamental to improve attitudes toward PLWH and enhance condom use. Author Contributions Z.M. contributed to study design, data collection, analysis, and interpretation of results, and drafted the manuscript. T.G., D.S., and J.D. contributed to study design, analysis, and interpretation of results, and critically reviewed the intellectual content of the manuscript. All authors have read and approved the final version of the manuscript. Z.M. had full access to all of the data in this study and takes complete responsibility for the integrity of the data and the accuracy of the data analysis. Conflicts of Interest The authors declare no conflicts of interest. Declaration of Generative AI and AI‐Assisted Technologies in the Manuscript Preparation Process In the manuscript preparation process, no AI or AI‐assisted technologies have been used. Transparency Statement Z.M., the first author, affirms that this manuscript is an honest, accurate, and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned and/or registered have been explained. Acknowledgments Z.M. is a fellow of the Mediterranean and Black Sea Program for Intervention Epidemiology Training (MediPIET) program, supported financially by the European Center for Disease Prevention and Control (ECDC). The views and opinions expressed herein do not state or reflect those of ECDC. ECDC is not responsible for the data and information collation and analysis, and cannot be held liable for conclusions or opinions drawn. In the manuscript preparation process no AI or AI‐assisted technologies have been used for manuscript drafting, editing, translation, or formatting. Article processing charges or other similar costs applied by the publisher to provide for open access publication will be met by the European Center for Disease Prevention and Control (ECDC) as part of the Mediterranean and Black Sea Program for Intervention Epidemiology Training (MediPIET) program (funded by the European Union), according to the Centre's policy). The ECDC had no involvement in study design, data collection, data analysis, interpretation of data, writing of this manuscript, and decision to submit this manuscript for publication. Data Availability Statement The data set underlying this study is available on request to the corresponding author. References Associated Data Data Availability Statement The data set underlying this study is available on request to the corresponding author.