This cross-sectional community based study was conducted in Minia city, Minia governorate, one of the Upper Egypt governorates, located 240 km south of Cairo, Egypt, during the period from September to November, 2016.

The required sample size was estimated based on the following conditions: assuming that the expected proportion of the population who practiced self- medication in Egypt (P) = 50%; tolerated error/margin of error (d) = 0.05; confidence interval (CI) = 95%. The following formula was used [n = p * (1 − p) * (Zα/d)2]. The value of Z is found in statistical tables which contain the area under the normal curve. Accordingly, the sample size was estimated and additions of 10% of the sample were added to guard against non-respondents rate [10] .

Inclusion criteria: Finally, a total 422 adult respondents were included in the study. Persons who were unable to answer the questions or give incomplete response due to some barriers were excluded from the study. Respondents were randomly selected using a multi-stage random sampling technique. Minia governorate was found to be divided into nine districts from which Minia district was chosen randomly (first stage), then Minia city which represent the urban area in the district was chosen (second stage) which was divided into four blocks (North, East, South and West) from which the second and third blocks were chosen randomly (third stage).

The households were selected by systematic random sample by visiting every third apartment building in a randomly-selected direction asking for adults (over 18 years old). Our team met about 460, eligible persons, from whom 422 agreed to be interviewed and participate in the study (response rate 91.7%).

Data were collected using a structured questionnaire, filled by interviewing each individual by researchers, consisted of five sections, the first section contained questions regarding socio-demographic information such as sex, age, educational level and employment status; also they asked whether they had any chronic illness. In addition, participants were asked whether or not they have health insurance, and if they have ever practiced self-medication in general, and in the past month in particular.

The second section of the questionnaire consisted of questions related to their perception about medications, knowledge about the use of the purchased drug; and practices related to the product purchased.

In the third section of the questionnaire the respondents were requested to report on the sources of medication used for self-treatment and sources of information about such medication, also it focused on the health conditions that respondents would self-treat; this section also investigated reasons for self- medication.

The fourth section covered another area which included antibiotic usage patterns, the respondents were asked to indicate how often they practiced self- medication with antibiotics and how to decide which types of antibiotics were suitable for their medical conditions and how they adjust the proper course of antibiotics. The final section contained questions about respondents’ beliefs and attitude concerning antibiotics self-use. This questionnaire had been tested on a small number of eligible persons as a pilot study to test the reliability of the questions and the time needed to conduct an interview. Then, proper corrections and adjustment had been fulfilled.

All the procedures of this study were reviewed and approved by the Institution Review Board of the Faculty of Medicine, Minia University (Approval No.: 17-037). Prior to data collection, informed consents were obtained from all participants after supplying comprehensive information about the nature and the objectives of the study.

The Statistical Program Statistical package of social science (SPSS) for Windows version 20 was used for data entry and analysis. Quantitative data were presented by mean and standard deviation, while qualitative data were presented by frequency distribution and compared by Chi-Square test. Risk ratios were estimated by calculating odds ratios (OR), and a multivariate logistic regression analysis was performed. The lowest accepted level of significance was ≤0.05.

Likewise findings from several studies [13] [17] [18] , our study revealed that drugs purchased from private pharmacies were the most commonly used sources of self-medication, reported by the majority of self-medicated individuals (86.7%), and the use of leftover medicine was also prevalent and reported by about 80% of respondents. Easy accessibility of medicines from the pharmacy without prescription could explain the high percentage of purchasing from private pharmacies as a major source of the practice of self-medication. Also, keeping medicine at home is an important concern lead to increase the possibility of self-medication and mistakes in proper consumption.

Regarding the source of information about the drugs used for self-medication, the commonest source of information was the pharmacists, reported by about 92% of respondents. This was followed by respondents’ experiences or knowledge from previous episodes (84.7%). The overall sources of information show that self-medication practices among adults in this study are not influenced by advertisements or the Internet; they were the least common source of information, reported by only 6.5% of respondents. which was similar to a study conducted in Saudi Arabia [18] , and revealed that 74% of participant had the source of information from pharmacist followed by respondents’ experiences (50.8%) and 16.4% from Internet and advertisements. This, however, is in contrast to a study conducted by Chui et al. [19] that showed that more than half of the participants never consulted a pharmacist on how to manage minor disorders.

The current study revealed that the commonest reason for self-medication was that the illness was perceived as minor (59.7%) followed by previous experience with the treatment (40%). About one-third of self-medicated respondents (29.8%) indicated that they did so because they lacked the time to visit formal health care facilities. One-quarter of self-medicated participants indicated that the cost of consultations with the doctor was a reason for self-medication. Above findings comparable to the results of a study performed by Sallam et al., 2009 (11) in Alexandria who found that the commonest cause was the illness perceived as minor (44.5%), followed by previous experience with the treatment (31%) and 24% indicated for other reasons. Another study conducted by Swetha R and Usha R revealed that long wait at clinics (31.6%), mild nature of the illness (27.56%) and financial problems (17.35%) were most common reasons for adopting self-medication among Indian population [20] .

The most common perceived illnesses for self-medication were common cold (90.6%), followed by headache (71.1%) then cough by (69.5%). This is similar to a finding by Noori [21] which showed that common cold has a high percent of the symptoms that led to self-medication. However, a study conducted in Sri Siddhartha, India showed that cough, headache and fever were common symptoms for which participants practiced self-medication [20] .

Self-medication with antibiotics was reported by 260 of 422 respondents (61.6%). This is in agreement with Khan et al., 2011 who found that 69% of the studied participant used antibiotics without a prescription [22] . Some other studies on self-medication with antibiotics have reported prevalence rates of 74% in Sudan [23] , 46% in Jordan [24] and 78% in Greece [25] .

Among the 260 respondents who used to take unprescribed antibiotics, 47.3% had used antibiotics more than once during the past month. About 44.2% had previous experience in using such antibiotics, 42.3% of the participants know the proper dose of antibiotic by asking the pharmacist. These findings were comparable to the results of a study performed by Widayati et al., 2011 [26] who found that 54% decide the type of antibiotic based on previous prescription and 52% know their information from the pharmacist.

Unfortunately, only 36% of the respondents were aware that inappropriate use of antibiotics leads to antibiotic resistance. Contradictory, Noori [21] reported 59.6% in Kufa, Iraq was aware of the antibiotic resistance due to irrational use of antibiotics. More than half (59.7%) believed that antibiotics should be used until relief of symptoms. This was high compared with a study carried out in Lithuania [27] , where 10% said that antibiotics should be used until relief of symptoms.

The current study revealed that (73%) of the sample thought that antibiotics are effective in the treatment of both bacterial and viral infection. Pavydė et al. [27] showed that Almost half of the respondents incorrectly identified antibiotics as being effective either in treating viral (26.0%) or mixed (bacterial and viral) infections (21.7%).

About two-thirds of the respondents (63.5%) knew that antibiotics could cause adverse drug reactions. This was low compared with a study carried out in Lithuania [27] , where the majority of the respondents (92.9%) knew that fact.

This result highlights the associated risk between self-medication practice and increasing age (>40 years, (OR = 2.28; 95% CI 1.12 - 4.67) and professionally employed participant (OR = 3.44; 95% CI 1.40 - 9.09). This is similar to a finding by Moraes et al. [28] which showed that increasing age (OR = 1.24), professional employment (OR = 1.21) were significant predictors of self-medication. Also, we revealed that the bigger the family size the more use of self-medication (OR = 2.04; 95% CI 1.84 - 2.33). This is in accordance with Widayati et al., 2011 (26) who found that increasing the family size (>4 members) is an important risk factor for self-use of medicine (OR = 1.2).

However we found no significant association between being covered by health insurance and practicing self-medication, this was in line with study conducted by Sarahroodi et al. (2012) who reported no statistically significant difference between self-medication and having or not having medical insurance [29] . On contrary, Widayati et al. (2011) [26] found that Indonesian who had perceived access to health care (having insurance) were 1.47 times more likely to self- medicate than those who reported no access to health care (OR = 1.47).