Knowledge levels and awareness about rational antibiotic use and antimicrobial resistance before and after graduation

Objective Irrational use of antibiotics may adversely affect the treatment outcomes or even lead to increased antimicrobial resistance. We aimed to evaluate the level of theoretical knowledge of rational antibiotic use and awareness about antimicrobial resistance among the senior students at a medical school and the family physicians. Methods This study was cross-sectional research and was carried out between 01 February-30 April 2019 in Malatya province. Two-hundred twenty-ve senior students in the Inonu University Medical School (Group 1) and 230 actively-working family physicians in Malatya primary healthcare services who were found eligible (Group 2). Power analysis was calculated as the minimum of 240 participants when considering a proportion difference of 0.18 between the groups, a type I error of 0.05 and a type II error of 0.20. A p<0.05 value was considered to be statistically signicant. Results Researchers argued that penal sanctions can be more effective by developing strict use policies to raise awareness of resistance to antimicrobials. Group 2 had higher self-condence, and it was also concluded that they forgot their theoretical antibiotic knowledge over time, and they could not follow the current information because of the intensity of their working life. Both groups stated that post-graduation training could be used effectively for reducing antimicrobial resistance. This study attempted difference in theoretical knowledge levels and Sustainable antibiotic training for doctors after graduation will contribute positive resistance of antibiotics.

In this study, we aimed to investigate antimicrobial prescribing habits and to evaluate the level of theoretical knowledge of rational antibiotic use and awareness about antimicrobial resistance among senior students of the medical school and family physicians in the Malatya province in Turkey.

Study Design
This study was carried out between 01 February-30 April 2019 in Malatya, which has approximately 750 thousand populations. This study survey was designed in line with the information obtained from various published research articles (9)(10)(11)(12), and the form was prepared, including seven sections and twenty-eight questions. Section 1 consisted of data that included ages, gender, and professional experiences of groups. Section 2 consisted of four questions containing personal antibiotic use data. Section 3 consisted of three questions involving the decision to start antibiotic treatment. Section 4 consisted of nine questions relating to rational antibiotic use. Section 5 consisted of four questions related to antimicrobial resistance. Section 6 consisted of three questions related to antibiotic training before graduation from medical school. Section 7 consisted of ve questions in the form of 'blanks-lling' regarding theoretical antibiotic knowledge.
The principles and purpose of this survey study were explained to all participants. The approval of all participants was taken orally and a consent form was signed by each participant. The informed, voluntary consent form for all participants in this study was received. Data were collected using face-to-face interviews. The questionnaire started by keeping con dential of the names of the participants in this study. All of the answers were recorded. Then, the collected data were compared according to the groups. We con rm that all methods were carried out in accordance with relevant guidelines and regulations.

Power analysis
Power analysis suggested a minimum of 120 senior students and 120 family physicians (totally minimum 240) when considering a proportion difference of 0.18 between the groups, a type I error of 0.05 and a type II error of 0.20. The number of the participants was calculated as a minimum seven-fold of each question in the questionnaire.

Inclusion and Exclusion Criteria
In this study, 225 senior students (group 1: the SS group) from Inonu University Medical School and 230 activeworking family physicians in Malatya province (group 2: the FP group) were included. Students of other faculties other than the Inonu University Medical School, physicians who worked outside primary care services, specialist physicians from other branches other than family physicians, and physicians who were not active even if they had experience to work as a family physician were excluded from this study.

Data Analysis
The quantitative data were given as mean±standard deviation, and qualitative data were summarized as numbers with percentages. Normal distribution was checked using the Kolmogorov-Smirnov test. Qualitative data were analyzed by the Pearson Chi-Square test as appropriate. Comparison of proportions was made using the Bonferroniadjusted Pearson Chi-Square test. A p<0.05 value was considered to be statistically signi cant. The IBM SPSS Statistics 26.0 software for Windows was used for analyses. A multilayer perceptron (MLP) arti cial neural network model was used to identify the questions that may be effective in classifying the group variable (categories: senior students, family physicians). To build the proposed model, the data set examined was divided into two sets as training and testing. In this study, the questions that may be effective in classifying the group variable, which is accepted as the target variable, were estimated with the proposed model, and their signi cance levels were calculated in descending order.
Q2: In which condition did you feel it necessary to start antibiotics the most? Q3: The way you took before you started antibiotics? Q5: The most important concern before an antibiotic prescription for a patient?
Q6: The most important nding that drives physicians mostly to prescribe? Q7: The most important nding-condition that drives physicians to prescribe?
The distribution of the questions related to the rational use of antibiotics is summarized in Table 2. As presented in Table 2, although the differences between the groups were signi cant concerning Q10-11-14-15 (p<0.05), the differences between the groups were non-signi cant concerning other questions (p>0.05). The distribution of antimicrobial resistance and awareness, developing recommendations, and slogans for decreasing antimicrobial resistance and increasing awareness is given in Table 3. In Table 3, Q19-20-21-23 were signi cantly different between the groups (p<0.05); the remaining questions were not signi cantly different (p>0.05).
Q19: The most frequent reason for penicillin resistance in staphylococci? Q20: How the national awareness of antimicrobials resistance can be increased? Q21: The adequacy of the education provided to you on antibiotics during medical education? Q23: What is your slogan about lowering the rate of antimicrobials resistance? Table 4 indicates the data of the groups on answering the questions on antibiotic information ( ll in the blanks questions). As shown in Table 4, Group 1 answered more questions, and the answers they gave were found to be more accurate. Table 5 demonstrates three-year antibiotic consumption rates by years in primary healthcare services in Malatya province in Turkey. According to the e-prescription information system data of the Malatya Public Health Directorate, the rate of prescribing antibiotics tends to decrease in years in patients applying for primary healthcare services in Malatya ( Table 5).
The importance values of the questions on the prediction of the groups are tabulated in Table 6. Among the questions, age was the most related variable (22.7%) with the groups. The second most important variable after the age variable was determined as a professional experience with a rate of 11.9%. On the other hand, the variable with the lowest signi cance value was obtained as gender with a rate of 0.1%.

Discussion
Increased antibiotics consumption and antimicrobial resistance are still signi cant problems in low socioeconomic countries (12). In our country, Turkey, the efforts have been made to reduce higher antibiotic consumption rates, to increase antimicrobial resistance awareness levels of physicians with training before and after graduation though raising the issue in writing/visual media, increasing the public awareness with public spots, and to prevent the nonprescribed sales of antibiotics in pharmacies since 2014.
One of the studies that was reported by Mahmood et al. (2016) showed that the irrational antimicrobial use resulted in reduced quality of care, increased morbidity and mortality, and increased cases of adverse drug reactions and medication errors (13). Andrajati et al. (14) conducted a study though analyzing 788 oral prescriptions, which were prescribed by 28 doctors for acute pharyngitis and nonspeci c respiratory infections at primary healthcare services, and 392 of them were evaluated for rationality according to local guidelines issued by Indonesian Ministry of Health.
They found that 220 prescriptions for selecting the right antibiotic did not meet the rational antibiotic prescribing criteria. Besides, they concluded that training for rational antibiotic use and physician experience were the factors associated with the rationality of antibiotic prescriptions (14). Since viral agents play roles in the etiology of the common cold, antibiotics have no usage in its treatment; and resting and supportive treatments are adequate (15).
We found that both groups had a su cient level of awareness of this issue.
Practitioners sometimes cannot deal with their health problems in a detailed way because of their busy work schedules and they want to solve the problems in the most practical way (16). It may be argued that group 1 started antibiotics for simpler reasons. It was also possible to argue that the self-con dence of group 2 was higher than group 1 in making this decision, and group 1 was more cautious in this respect. These different approaches might be associated with the lack of self-con dence in the professional experience of the physicians or lack of adequate time rather than the presence of a healthcare employee in the place of residence and family. Given that the group 2 decided not to start antibiotics treatment, although it would necessarily cause more worries for themselves, and that the concerns of the group 2 being less might be explained by having more professional experience. The differences in knowledge, attitudes, and behaviors of physicians who prescribe antibiotics in primary healthcare facilities on deciding to start treatment and the necessity of treatment may also cause the rates of increase antibiotic prescription (17). The quality of a candidate physicians' pre-graduation training and the ability to apply this theoretical training in the master-apprentice or individual practice may signi cantly affect the attitude and behavior when starting antibiotics. Many factors may in uence doctors' decisions, leading them to breach the principles of good clinical practice (18). For example, the fear of possible future complications in their patients or a desire to ful ll patients' expectations. Many physicians fear that they may miss out on the infection in the presence of leukocytosis, and often prefer to prescribe antibiotics to feel safe (19). In our study, the group 2 had more professional experience generally about antibiotics prescription in the presence of leukocytosis and argued the most important condition that forced primary healthcare physicians to write antibiotics is the lack of knowledge of these physicians. The physicians who had less experience believed that the presence of fever was more important, and the suspicion of infection that is not evidenced was the most important factor driving primary healthcare physicians to write antibiotics (20). Today, unfortunately, physicians can allocate less than optimal time to the patients for diagnosis because of an overcrowded patient for primary health care services and are often under pressure to prescribe antibiotics by patients and/or their relatives (21).
Among the most common causes of irrational antibiotic use, there is the expectation of the patient to prescribe an antibiotic from the physician considering "being a good doctor" equivalent to writing a prescription, insu cient examination facilities, and the need for the physician for feel safe by prescribing antibiotics by fearing that they might overlook an infection (22). The expectation of a patient to prescribe an antibiotic from the physician, keeping the equivalent of writing good medicine prescriptions, insu cient examination facilities, and the need for the physician to feel safe by prescribing antibiotics with the fear of distracting a possible infection are among the most common causes of irrational antibiotic use. (23). Another nding is that physicians forget their theoretical knowledge in time after they start active duty; they do not improve themselves, and follow update literature and developments because of their intensive work schedule (24). Group 1 thought that these reasons were more prominent in primary healthcare.
Chemoprophylaxis is an optional procedure that is aimed to keep an infection with a high probability of development (25). The correct de nition in this respect was made by group 1. Although there are guidelines released on surgical prophylaxis, we have witnessed misconduct in many clinics. The level of knowledge on surgical prophylaxis was low in both of our groups, and the possible cause of this might have been that they did not use it too much. Thus, the theoretical knowledge could be forgotten in time.
Antimicrobial resistance can be de ned as the ine ciency of antibiotics in time due to the irrational/excessive use of antibiotics (26). The correct de nition is made by group 1 at higher rates may be explained by that their theoretical knowledge is more recent. The most common reason for penicillin resistance in Staphylococci is the synthesis of a new PBP (Penicillin-Binding Protein) (27). Although nearly half of group 1 was thought in this way, more than half of group 2 believed that beta-lactamase was secreted, which was wrong. This nding may suggest that both groups are inadequate or do not update their knowledge of antimicrobial resistance mechanisms.
The presence of comorbid conditions and diseases must be considered in a patient who is scheduled to receive antibiotics. If an antibiotic drug that is initiated for the treatment of an infection is administered without considering comorbid diseases, it will inevitably lead to undesirable outcomes (28). The ability to answer the questions on comorbid conditions and diseases (e.g., pregnancy and renal failure) and the percentages of accurate answers were higher at a signi cant level in group 1. This may make us consider that group 2 forgets the theoretical knowledge in time and they are more likely to make mistakes. This subject shows the importance of transforming the antibiotic awareness in the basic pharmacology courses to the students in the early stages of school into knowledge that may also be used in the future (29). In this study, the most probable factors for the three most common primary healthcare groups were asked. The ndings showed that the percentages of the rates of answering to questions and providing correct answers were low. Another interesting detail was that some of them in group 2 answered this question by writing the name antibiotics instead of the name of the agent by mistake and trying to answer the question more carelessly.
The opinions and suggestions of the participants in the awareness of antimicrobial resistance were also of interest when planning the questionnaire. Among the suggestions that were included in our questionnaire, group 1 defended the idea that periodic training before and after graduation, developing strict usage policies and applying penal sanctions would be more effective; and the group 2 defended that the issue should be brought to the agenda in written and visual media more frequently, using public spots emphasizing antimicrobial resistance, and preparing brochures and posters would be more effective. A new generation of messages that encourage the rst-choice use of narrow-spectrum antibiotics is needed, re ecting international efforts to preserve broad-spectrum antibiotic classes (30).
We have demonstrated the following conclusions: The levels of theoretical antibiotic knowledge are better in the pregraduation period. After graduation, doctors ' theoretical knowledge of antibiotics is forgotten over time, and they are more likely to make mistakes. This situation has once again revealed the importance of education after graduation.
The penal sanctions could be more effective by developing strict use policies to raise awareness of antimicrobial resistance. It could be said that they could not update their knowledge in this period due to the intensity in working life, and they forgot their theoretical knowledge of antibiotics over time. Especially training after graduation is crucial in reducing the excessive rates of antibiotic usage. This study could draw attention to increased antimicrobial resistance by highlighting the slogan which was put forward by Unal S et al. for the rst time in 2014 at Hacettepe University in Ankara "Either awareness or resistance develops for antibiotics". This slogan may be useful for raising awareness against increased antimicrobial resistance. These results once again highlighted the need for immediate action of training and corrective actions.
The survey model used in this study was adapted from several similar studies that have been carried out before. It was applied to medical school senior students and family physicians to determine the differences between before and after graduation. If there were more participants, perhaps more accurate results could have been obtained, and more accurate analyses might have been performed.

Conclusions
This study may raise awareness to determine the difference in theoretical knowledge levels concerning antibiotics and behavior models of physicians before and after graduation and to reduce higher use rates to lower levels of antibiotics. Sustainable antibiotic training for the doctors after graduation will highly likely to contribute positively to antimicrobial resistance and the rational use of antibiotics. Further studies should be conducted with more participants.

Declarations
Ethical approval and consent to participate This study was approved by the Non-Interventional Ethics Committee at Inonu University (Approval no: 2019/01-8).
Another necessary permission for e-prescription information system data was obtained from the Public Health Unit of Malatya Provincial Healthcare Directorate (2019-1208). The informed, voluntary consent form for all participants in this study was received.

Patient consent for publication
The consent form was received from all participants for publication.

Availability of data and materials
The ndings data are available from the corresponding author upon reasonable request.

Competing interest
The authors have no competing interest  The most vital criterion to start?
The patients' fever Acute phase reactants being high Having a positive culture result The disrupted hemodynamics in the patient    Table 4. Data of the groups on answering the questions on antibiotic information (fill in the blanks). Different letters (a, b) in each row show a statistically significant difference (p<0.05; Bonferroni-corrected Pearson's chisquare test).