Abstract
BACKGROUND: This study aims to assess how Romanian medical students suffer from dry eye disease symptoms, establish the prevalence and severity of dry eye (DE) symptoms and identify potential risk factors. METHODS: An analytical, cross-sectional study was conducted on students from "Carol Davila" University of Medicine, Romania, after the final examination period of July 2022. The OSDI score (Ocular Surface Disease Index©) was applied in an online survey. The study adopted the standards used by other authors, who defined symptomatic DED as an OSDI score greater than 12. The chi-square test was used to establish statistical significance at a cutoff value of p < 0.05. The predictive model was created using linear logistic regression analysis. The goodness of fit of the logistic regression model was assessed using the Hosmer-Lemeshow test. When the severity outcome had a nominal categorical form, multinomial regression analysis with normal subjects as a reference was performed. The distribution of the severe type of symptomatology across sex categories and years of study was analyzed using a nonparametric test (Independent-Samples Kruskal-Wallis Test). RESULTS: A total of 274 answers were received from 81.4% females and 18.6% males with a response rate of 35.58%. The mean age was 22.7 years ± 1.66 with an age range between 20 and 25 years old. Using the OSDI score, we established that the overall prevalence of DE symptoms was 83.6% (95%CI: 79.6%, 88%), with an 85.2% (95%CI: 80.5%, 89.8%) prevalence in females and 76.5% (95%CI: 65%,88%) in males. The severe form of DE was the most prevalent, regardless of the study year or sex. Increased screen time (p-value < 0.05) and non-smokers (p-value < 0.05) were proven risk factors. The predictive model which includes the explanatory variables (sex, contact lens wearers, smoking, oral contraceptives, screen time) proved an 84.7% predictability for symptomatic DE and was able to better predict the dependent variable than the intercept model only (p-value < 0.05). Smoking (p = 0.002) and screen time (p = 0.009) preserved their significance in the multinominal regression as well. CONCLUSIONS: This is the first study to report the epidemiology of DE symptoms among Romanian medical students. OSDI revealed a high prevalence of symptomatic DE in medical students. Screen time, although not the only factor, likely plays a role in exacerbating the disease. This information can be used to inform healthcare policies, establish occupational health guidelines, and implement preventive measures for individuals in similar high-stress academic or professional environments.