Sociodemographic Determinants of Health Literacy Among University Students of Health Sciences in Kosovo

We decided to invite all students (N=671) from the universities of Prishtina and Gjakova attending studies in one of the following branches: nursing, health management, physiotherapy, or public health.

The survey consisted of self-administration of an anonymous and structured pen-and-paper questionnaire to all students who agreed to participate. Students were invited to fill in the questionnaires during classes by their respective lecturers, without being offered any incentives, while being informed about the aim and procedures of the survey and being expressly guaranteed the anonymity of the survey and the subsequent aggregated data analysis.

The questionnaire included an assessment of the HL and sociodemographic information of participating students. Assessment of HL was based on the internationally standardised European Health Literacy Survey Questionnaire (HLS-EU-Q) (20, 25, 26), which had been already validated in Albanian settings (27). The HLS-EU-Q consists of 47 items measuring students’ ability to access, understand, appraise and apply health information across three health domains (healthcare, disease prevention and health promotion) (25, 26). Each item of the instrument was rated on a 4-level Likert scale: “very easy” (score: 4), “fairly easy” (score: 3), “fairly difficult” (score: 2), and “very difficult” (score: 1).

Standard descriptive statistics methods were used to describe the variables, while binary logistic regression (direct method) was used to assess the association of ‘inadequate/problematic HL’ with the sociodemographic characteristics of study participants. A simple method with a single reference category was used to create dummy variables. Odds ratios (ORs), their respective 95% confidence intervals (95% CIs) and p-values were calculated. A Hosmer-Lemeshow test was used to assess the goodness-of-fit of the multivariable-adjusted logistic regression models (29). SPSS (version 19.0) was used for all the statistical analyses.

The study was approved by the Ethics Committee of the National Institute of Public Health of Kosovo (Decision: 01/1531, date: 28-12-2023) and by the Ethics Committee of the Faculty of Medicine, University of Gjakova (Decision: 006/82, date: 12-01-2024).

Of all the invited students (N=67l), 113 (≈17% of the target population) did not complete the survey, while a further 88 participants (13% of the total) provided incomplete responses, which were excluded from the analysis. There were no significant differences among respondents and non-respondents regarding gender and age distribution.

The study sample included in the statistical analyses consisted of 470 students (405 females, or ≈86% of the overall study sample - Table 1). The final response rate in this study was: 470/671=70%. All other characteristics are shown in Table 1.

The HL-score results showed that the mean value was 40.7±5.0, while the HL-level results showed that out of 470 students, 3 (0.6%) of them exhibited an inadequate, 30 (6.4%) exhibited problematic, 239 (50.9%) exhibited sufficient, and 198 (42.1%) exhibited an excellent HL-level. Overall, only 33 (7.0%) of the students exhibited ‘inadequate/problematic HL’.

In crude/unadjusted binary logistic regression models (Table 2), the results showed a statistically significant association between the observed outcome and gender, marital status and employment status. Additionally, the association between inadequate/problematic HL and university institution was marginally statistically significant.

Upon simultaneous adjustment for all sociodemographic characteristics (Table 3), the positive association of the observed outcome with gender persisted. The odds even slightly increased. The only determinant for which the odds increased in the multivariate analysis, besides gender, was university, although the association was only marginally statistically significant. Conversely, the relationships with marital status and employment were no longer statistically significant and their odds considerably decreased (Table 3).

The model of the multivariate association of inadequate/problematic health literacy with sociodemographic determinants as a whole was statistically significant, as the HL test of goodness-of-fit showed that the model fits the data well (P=0.790).

The main finding of this study consists of a relatively high general HL level among university students in Kosovo, especially those from the University of Gjakova. Hence, only 7% of the students exhibited “inadequate and/or problematic” HL. Overall, female students had a remarkably higher HL level compared to males, a finding which was quite consistent and persisted upon adjustment for a wide range of other sociodemographic characteristics. Conversely, economic level, an indicator of social status, was not a significant determinant of general HL.

The general HL level in our study is a bit higher than a previous report from Albania, which consisted of a similar cross-sectional study applying the same HL instrument in a sample of students pertinent to a range of health sciences (nursing, physiotherapy, midwifery and laboratory technicians) (20). Furthermore, according to this recent report (20), about half of Albanian students had sufficient HL (which is very similar to our estimate), while an excellent HL level was higher than in our study. Hence, on the face of it, the HL levels among university students from Kosovo are higher compared to their counterparts from Albania attending similar study branches (pertinent to health sciences). Apparently, Kosovo has developed a robust curriculum for health sciences that emphasises HL, including practical skills and knowledge about navigating health information (24). Also, in Kosovo, cultural attitudes toward health, well-being and education could play a significant role in promoting HL. Additionally, students in Kosovo may have better access to modern health information technologies, such as e-health platforms or digital libraries, which may engage them more effectively with up-to-date health information, boosting their HL levels. However, our findings on the seemingly high HL levels should be interpreted with caution and replicated in future studies in Kosovo.

The prevalence of adequate (sufficient or excellent) HL in our study was higher than the estimate among Albanian students, as well as among nursing students in Namibia (30). A study including Spanish and French students from different branches reported that only about two in five students had sufficient HL (31). However, nursing students exhibited the best HL levels in that study, which used a different HL measuring scale (31), making it difficult to compare with our study and the previous one from Albania (20). Yet, methodological considerations aside (including especially different measuring HL scales), previous literature on nursing students indicates that around one third of the surveyed students have problematic or limited HL (31, 32). A recent report from Spain (8) indicated that few nursing students had an inadequate HL level, while one third of them had a problematic HL level.

Regarding another aspect, our finding related to a higher HL level among females compared to males is compatible with a previous study from Albania (20) which reported that, in multivariable-adjusted binary logistic regression models, the only factor significantly increasing the likelihood of limited HL was male gender. Seemingly, females tend to engage more with health-related information in Albanian settings, which may be rooted in their traditional societies imposing certain cultural and social roles. From this perspective, females in Albania and Kosovo may assume more responsibility for health management within families, such as caring for children and elderly relatives. This role could lead to a greater focus on acquiring health knowledge and skills, thereby enhancing their HL levels.

Furthermore, the large European HL survey, employing the same measuring instrument used in our study (HLS-EU-Q), has reported a weak yet positive relationship between female gender and HL level (25, 26). However, a recent study from Spain has reported opposite findings, with female students exhibiting higher levels of insufficient HL (8). On the other hand, some other studies conducted on nursing students have not reported significant gender differences in HL levels (32, 33).

In a previous study from Albania, low social and economic status also increased the likelihood of limited HL, but these associations were not statistically significant in multivariable-adjusted models (20). In our study, economic level was not significantly associated with HL levels even in crude (unadjusted) models.

Of note, the higher prevalence of “inadequate and/or problematic” HL among students who were married compared with those who were single evidenced in our study is explained by the strong confounding effect of gender, with male students (exhibiting lower HL levels) reporting a much higher degree of being married than their female counterparts (displaying significantly higher HL levels). As a matter of fact, upon multivariable adjustment for the other sociodemographic factors including gender, there was no evidence of a significant relationship of HL with marital status.

Conversely, our findings on a positive relationship of age with higher HL levels are intuitive and in line with previous reports from Albania and other countries (8, 20, 34–36). As a matter of fact, in studies involving student populations, older age indicates a more advanced academic year, which has been shown to exhibit a direct linear association with higher HL levels (8, 34–36).

This study may have some limitations. The first and probably the most important limitation is that the observed outcome is relatively rare. This posed some technical difficulties in the analysis, especially in determinants that had some rare categories (e.g. male gender). However, important results were obtained for planning potential interventions despite this limitation. Second, there is a possibility of selection bias (due to non-response and/or non-valid information provided by almost 30% of the targeted population). Nevertheless, the achieved response rate still provided valuable results. Next, there is a possibility of information bias (likelihood of differential reporting between various sociodemographic groupings). The next potential limitation refers to the measurement of socioeconomic status, which should have combined various economic and social factors of participating students (such as the level and quality of education attained before attending the current studies in health sciences, income, wealth and social status). Finally, the survey design is cross-sectional which does not allow inferences about causality.

Conversely, our study has important strengths. First, for participation in the survey the entire targeted population was invited, and second, the study instrument used was based on a standardised international instrument, previously validated in Albanian settings (27).

Our study indicates a need to strengthen HL among male students in Kosovo and that the problem seems to be greater at the University of Prishtina. A potential explanation for this regional difference may be the educational environment (curriculum differences, as well as resources and faculty expertise), along with gender dynamics and cultural norms which could influence attitudes towards HL (e.g., different traditional views in these regions may affect how male students perceive health education).

At a broader level, health professionals are essential for patient care at all levels of healthcare services (37, 38). An important component of their work consists of empowering patients and enabling prompt and effective self-care by enhancing their HL skills (37). Therefore, preparation of upcoming health professionals is of paramount importance for addressing the difficulties and needs of patients with inadequate HL levels (37). From this perspective, our study indicates that proper education and training especially of male students from different branches of health sciences is crucial for ensuring they can effectively comprehend and apply health information in their future professional practices, ultimately enhancing patient care and population health outcomes.

Conversely, from a research viewpoint, future studies on HL among health sciences students should explore particularly the impact of curriculum design and cultural factors on HL levels to inform targeted educational interventions for upcoming health professionals.