A cross-sectional survey was carried out among second and fourth semester MBBS students of MCOMS willing to participate. As the students had completed two and four semesters of integrated basic sciences, they were knowledgeable about common medicines used for self-medication like non-steroidal anti-inflammatory drugs (NSAIDs), anti-histaminic (anti-allergic) agents, antibiotics, etc.

The study was carried out from the last week of July to the first week of August 2015 using a questionnaire (shown as an additional file). The questionnaire consisted of three parts. The first part obtained demographic data of the respondents e.g. age, gender, nationality, semester of study, location of their home town (e.g. urban or rural), and occupation of their parents (father and mother). The second part of the questionnaire contained 40 statements to measure respondents’ knowledge and attitude about self-medication. The students were asked to provide a score (1 to 5) according to their degree of agreement with the statements according to the scale 1= strongly disagree, 2= disagree, 3= neutral, 4= agree and 5= strongly agree. Twenty statements (statement no. 1, 3-6, 9, 11-14, 18, 20, 24, 28, 29, 31, 36, and 38-40) measured knowledge while the remaining 20 statements measured attitude. The third part of the questionnaire included few questions to study the practice of self-medication and source of information used by the respondents while practicing self-medication.

The questionnaire was developed by the authors after an extensive review of literature. The questionnaire was circulated among a group of faculty members at MCOMS, Pokhara for their inputs regarding validity of the statements and their ease of comprehension. The questionnaire was tested for ease of comprehension and readability among six clinical students of the institution. Their data was not included in the final analysis. Cronbach’s alpha was calculated as a measure of internal validity of the questionnaire.

The collected data were entered into and analysed using SPSS (Statistical Package for the Social Sciences) version 20 for windows. To avoid bias, certain statements (statement numbers 8, 11, 17, 19, and 27) were negatively worded and their scores were reversed while calculating the total and subscale scores. The normality of distribution of the total, knowledge, and attitude scores was studied using one sample Kolmogorov-Smirnov test (p-value of 0.05). The average scores and their dispersion were compared among different categories of respondents using appropriate statistical tests. The frequency of prevalence of self-medication among respondents grouped according to demographic characteristics was compared using chi-square test. A p-value of less than 0.05 was taken as statistically significant. Free text comments about self-medication were noted and common ones tabulated.

The study was approved by the Institutional Review Committee (IRC), MCOMS, Pokhara vide notification MEMG/IRC/GA (i). The students were invited to participate and informed about the objectives and importance of the study. It was clarified that participation was voluntary and they could withdraw from the study at any time without giving any reason. They were assured about the confidentiality of their identity and had the opportunity to answer all questions honestly. Written informed consent was obtained from all the respondents prior to administering the questionnaire.

Two hundred and seventy six of the 295 (93.6%) second and fourth semester students participated in the study. More number of respondents were female, of Nepalese nationality, and of age 20 years [Table/Fig-2]. The mean (±SD) age of the respondents was 20.16 (±1.24) years. [Table/Fig-2] shows the demographic characteristics of the respondents.

The Cronbach’s alpha value for the questionnaire was 0.712 indicating a good level of internal consistency. The one sample Kolmogorov-Smirnov test showed the total, knowledge and attitude scores were normally distributed. Hence mean was used as the measure of central tendency and standard deviation as the measure of variation. The mean scores were compared among different subgroups of respondents using appropriate statistical tests. Independent samples t-test was used for dichotomous variables and analysis of variance (ANOVA) for others. A p-value less than 0.05 were taken as statistically significant. Tukey HSD test was used as the Post-hoc test for pair wise comparison if ANOVA showed a significant difference.

The mean (SD) knowledge, attitude, and total scores were 74.54 (6.92), 67.18 (5.68), and 141.73 (10.76), respectively. The maximum values for knowledge and attitudes were 100 and the maximum possible total score was 200.

[Table/Fig-3] shows the knowledge scores according to respondents’ personal characteristics. There was no significant difference in scores according to respondents’ gender, age, and the profession of their parents. However, the differences in mean score according to respondents’ semester of study and nationality were statistically significant. The mean scores of fourth semester students were significantly higher compared to second semester students. With regard to nationality post-hoc test showed that the mean knowledge score of Nepalese respondents (75.4) was significantly (p=0.008) higher compared to Indians (72.0) and the mean knowledge score of other country respondents (76.6) was significantly (p=0.033) higher compared to Indian respondents (72.0).

[Table/Fig-4] shows the attitude scores according to respondents’ personal characteristics. There was no significant difference in the attitude scores according to the respondents’ gender, age, and the profession of their parents. However, the mean score according to respondents’ semester of study, and nationality was statistically significant. Post-hoc test showed that the mean attitude score of Nepalese respondents (68.3) was significantly (p=0.001) higher compared to Indian respondents (65.1). [Table/Fig-5] shows the total scores according to respondents’ personal characteristics. Significant differences in the mean (SD) total scores of the respondents were found (one-way ANOVA and Post-hoc tests) between Nepalese and Indian respondents (p<0.001) and between Indian and other country respondents (p=0.017). The mean (SD) total scores of respondents from Nepal and other countries were 143.7 (11.2) and 144.8 (9.0), respectively, which was significantly higher than the score of Indian respondents i.e. 137.1 (11.0). Similarly, the mean total score of fourth semester students were significantly higher compared to second semester students. However, there was no significant difference in the total score according to respondents’ gender, age, and the profession of their parents.

Two hundred and twenty six (81.9%) of 276 respondents mentioned that they had self-medicated during the one year period preceding the study and 149 respondents (54%) shared that previous experience with the medicine was one of the information sources for self-medication. Similarly, 92 (33.3%), 56 (20.3%), 51 (18.5 %) and 30 (10.9%) respondents shared that communication with their parents, their textbooks, Medical Shop (Pharmacy) and others (including internet), respectively, were used as information sources for self-medication. As the respondent could choose multiple answers in response to this question, the total per cent was more than 100.

Comparison of the reported prevalence of self-medication during the preceding year showed there was no statistically significant difference among respondents by semester of study, gender, age, and profession of the parents. However, there was a significant difference among respondents of different nationalities. The use of self-medication was more common among Sri Lankan respondents compared to Nepalese respondents [Table/Fig-6]. Painkiller (73.2%), antipyretics (68.8%), and antimicrobials (56.2%) were most commonly used medicine for self-medication [Table/Fig-7].

Certain free text comments by respondents about self-medication are mentioned below. Some of the comments were:

The factors promoting self-medication, as shared by the respondents, are as follows:

The students’ overall knowledge and attitude about self-medication in the present study was encouraging. The scores were significantly higher among fourth semester students compared to those in the second semester. Mean knowledge score and total score about self-medication was significantly lower among Indian students compared to students from Nepal and other countries. The mean attitude score for Indian students was lower than Nepalese students.

The overall prevalence of self-medication in our study (81.9%) was comparable to the prevalence (82.3%) reported among medical students (including interns) in Ahmedabad, India [13] and in Serbia (79.9%) [15] and greater than the incidence (55.2%) reported from Egypt [9].

Unlike, the findings from Serbia where prevalence of self-medication was dependent on age and gender [15], in our study the prevalence did not significantly differ according to semester of study, gender, and age. The respondents through their free text comments, shared that if their parent/s is/are medical doctor/s then their accessibility to medicines increases so the chance of self-medication increases. In the present study, the proportion of respondents, whose parent/s is/are medical doctor, practicing self-medication was higher compared to respondents whose parent/s was/were nonmedical personnel. However, the difference was not significant. A higher proportion of Sri Lankan students and students from other countries were practicing self-medication compared to students from Nepal or India. Students from Sri Lanka, Maldives and other countries express difficulty in communicating with people in Pokhara (mainly due to language problems), which may be one of the reasons for higher incidence of self-medication. Almost 50% of students from Maldives had worked as Clinical Assistant before joining the MBBS program at MCOMS which might have also encouraged self-medication (Out of 22 respondents from other countries, 21 were Maldivian and 1 was American). As shared in the free comments by the respondents, the peer pressure of the self-medicating students might have some influence. Further and in-depth study is required.

Studies have reported that the practice of self-medication increases as student progress through medical school [13,16–18]. Banerjee et al., reported significant difference (p<0.001) in proportion of students practicing self-medication 41.67% (first year) Vs 79.31% (Final year) [16]. Similarly, the proportions were 89.4% (first year) vs 94.1% (final year) (p= 0.038) in the study conducted in Slovenia [17] and 52.6% (2 year) vs 73.3% (4 year) (p= 0.02) in study conducted in Bahrain [18]. In this study the self-medication prevalence among fourth semester students was higher compared to the second semester but the difference was not significant. The finding is in congruence with the study done in Nagpur, India where prevalence of self-medication in first year and third year students were 78.0% and 74.7%, respectively and the difference was not significant [12]. Even though, in present study, the knowledge and attitude of self-medication was higher among fourth semester respondents compared to second semester, the prevalence of self-medication was not significantly different.

Lukovic et al., reported that female respondents self-medicate more than males (p<0.001). In our study, proportion of self-medicating female respondents were more compared to male respondents but the difference was not significant [15]. Similar to our study, the study from India also reported no significant association of self-medication with gender [13].

Almost three quarter (73.2%) of the respondents had used painkillers to self-medicate; similarly antipyretics (68.8%), antimicrobials (56.2%) and anti-allergics (54.7%) were commonly used in the study. In another study conducted in Nepal the percentage of respondents using analgesic, antipyretics, antibiotics and anti-allergies for self-medication was 75.8%, 46.8%, 40.3% and 35.5%, respectively [10]. Similarly, in the study conducted among medical students in Bahrain the proportions were analgesic (81.3%), antipyretics (43%), antibiotics (6%) and anti-allergies (13%) [19]. Various other studies have also reported analgesics and anti-pyretics to be the most common medicine used for self-medication [7,13,15].

Antibiotics use for self-medication was 41% in Egypt, 38.9% in Serbia, 31.09% in West Bengal, India, 34% in South India and up to 34.8% in Nagpur, India [7,9,14―16]. In our study, more than half (56.2%) of the respondents reported using antimicrobials. This higher proportion of use of antimicrobials may be of great concern because over use and irrational use of antimicrobials may lead to resistance. Some respondents, through their free text comments, shared their apprehensions regarding use of antimicrobials for self-medication or availability of antimicrobials without a prescription. Few of them suggested for enforcement of strict regulations and severe penalties for defaulters. This shows that the respondents were aware of the problem uncontrolled and irrational use of antimicrobials in the region. Availability of antibiotics without prescription was also reported in a previous study conducted in Pokhara [6].

In a previous study conducted in Nepal among medical and paramedical students, 60.3% of respondents used pharmacist as a source of information which is higher than that reported in our study (18.5%) [10]. In our study, 20.3% of respondents reported to use text book as a source of information for self-medication which is less compared to other studies e.g. 46% [10], 39% [7] and 44.5% [12]. Mehta et al., reported that 34.7% used the same medicine which had been previously prescribed to them when they experienced similar symptoms [10] but in our study 54% of respondents shared that previous experience with the medicine was one of the reasons to self-medicate.

As the questionnaire was self-administered and information about self-medication in the preceding one year was collected, there could be recall bias. Although the respondents were requested to complete the questionnaire independently, discussion between the respondents could not be entirely ruled out. The strength of the study was the high response rate and good sample size.

The study shows high prevalence of self-medication among basic science medical students of the institution. The students’ knowledge and attitude about self-medication is encouraging. High prevalence of self-medication with antibiotics is of concern. Further studies to investigate in detail the cause of high self-medication could be planned. There is a need of a module to educate and aware the students about advantages of responsible self-medication and drawbacks of self-medication. A similar can be carried out among students during the clinical years of study. Similar type of study on self-medication in various other medical colleges in Nepal may provide a clearer picture of self-medication among medical students of Nepal.