This study examines self-reported changes in antibiotic prescribing practices among 222 physicians post-COVID-19, focusing on factors influencing these changes. The sample had a mean age of 47.7 years (SD = 11.1), with 62.2% female participants. Statistical analyses included Mann-Whitney U tests and logistic regression to identify predictors of prescribing changes.

Physicians reporting changes were more likely to prescribe antibiotics due to patient demand (U = 2725, p < 0.001), time constraints (U = 3158, p < 0.01), concerns about complications (U = 2803, p < 0.01), and peer behavior (U = 3190, p < 0.01). Logistic regression revealed that age predicted changes in prescribing behavior (OR = 1.0831, p = 0.029), while professional experience (OR = 0.9274, p = 0.044) and additional specializations (OR = 0.3835, p = 0.023) were inversely related. Residence in villages showed a trend toward increased prescribing (OR = 7.8055, p = 0.097).

These findings underscore the influence of demographic, professional, and geographic factors on antibiotic prescribing practices. Future research should investigate the long-term effects on antibiotic resistance and evaluate targeted interventions to address the identified drivers of prescribing behavior.

antimicrobial resistance, evidence-based medicine, physician behavior

Antibiotic resistance has been a long-standing global health issue, largely driven by the overuse and inappropriate prescribing of antibiotics (Arason and Sigurdsson 2010; Imanpour et al. 2017). Before the COVID-19 pandemic, extensive research highlighted factors influencing antibiotic prescribing, including variations in practices among healthcare providers and across specialties (Arason and Sigurdsson 2010; Imanpour et al. 2017). Prescribing antibiotics is a complex decision that balances immediate patient needs with long-term resistance risks, influenced by factors such as patient socioeconomic status, diagnostic challenges, patient pressure, adherence to clinical guidelines, and regional resistance patterns (Gulliford et al. 2014; Gaarslev et al. 2016; Smieszek et al. 2018).

Physicians’ prescribing habits are shaped by their medical training, awareness of antimicrobial resistance, patient expectations, and diagnostic uncertainties (Thakolkaran et al. 2017). In uncertain diagnoses, some physicians may over-prescribe antibiotics as a precaution (Thakolkaran et al. 2017). Patient pressure, particularly the misconception that antibiotics are needed for viral infections, exacerbates this issue (Macfarlane et al. 1997; Butler et al. 1998; Faber et al. 2010; Thakolkaran et al. 2017; Gulliford et al. 2021), leading to unnecessary antibiotic prescriptions to meet expectations and maintain satisfaction (Shapiro 2002).

Adherence to clinical guidelines typically improves prescribing practices, but not all physicians consistently follow them (Jaunay et al. 2000; Thakolkaran et al. 2017). Education and training in antibiotic stewardship are crucial for improving knowledge and practices (Thakolkaran et al. 2017). Physicians may prescribe antibiotics to mitigate risks, particularly in patients with compromised immune systems (Thakolkaran et al. 2017). The impact of healthcare policies, such as antibiotic stewardship programs, also influences prescribing behaviors, with stricter policies generally leading to more cautious practices (Thakolkaran et al. 2017).

Physician experience and specialty significantly affect antibiotic prescribing. More experienced physicians may be better at avoiding unnecessary prescriptions, particularly for viral infections (Thakolkaran et al. 2017; Abdel et al. 2020). However, factors like work-related stress, patient pressure, and uncertainty can still drive over-prescription (Shapiro 2002; Chandra Deb et al. 2022). Different specialties face varying patterns of antibiotic use, with surgeons often focusing on prophylaxis and pediatricians more cautious due to potential risks in children (Thakolkaran et al. 2017; Abdel et al. 2020).

The COVID-19 pandemic has dramatically impacted healthcare systems and altered antibiotic prescribing practices. Increased demand for healthcare services, along with managing COVID-19 cases and potential co-infections, has led to a surge in antibiotic use (De Oliveira Santos et al. 2022). Inappropriate prescribing, already prevalent in ambulatory care, worsened during the pandemic as physicians faced uncertainty in managing the disease (Imanpour et al. 2017; Chandra Deb et al. 2022). The rise of telemedicine, while beneficial, may have further encouraged over-prescription, as physicians could prescribe antibiotics without a physical examination (Alzahrani et al. 2022). Additionally, the strain on healthcare resources and the prioritization of COVID-19 cases led some physicians to prescribe antibiotics even when not clearly indicated (Anthony 2020; Gunasekeran et al. 2021; Alzahrani et al. 2022), potentially exacerbating antibiotic resistance.

Social media has also played a role in shaping medical practices during the pandemic, as misinformation and online discussions may influence physicians’ prescribing decisions (Goel and Gupta 2020). The World Health Organization and the Centers for Disease Control and Prevention have identified antibiotic resistance as one of the greatest global health threats, calling for stronger stewardship to reduce resistance and preserve antibiotic effectiveness (Ray et al. 2019; Chandra Deb et al. 2022). Effective education, training, and adherence to clinical guidelines are key in improving antibiotic prescribing (Rassi et al. 2021). However, first, it is crucial to examine current prescribing habits and identify gaps to guide improvements.

This study aims to investigate how the COVID-19 pandemic has altered antibiotic prescribing patterns among physicians in Bulgaria, with a particular focus on the demographic and professional factors influencing these changes.

The sample comprised 222 participants with an average age of 47.7 ± 11.1 years, indicating variability within the cohort. Gender distribution showed a predominance of females (62.2%) compared to males (37.8%). Participants reported an average of 18.7 ± 10.7 years of professional experience, reflecting a diverse range of experience levels. The majority of participants worked in regional cities (87.4%), while smaller proportions worked in small cities (10.8%) or villages (1.8%). Ethnicity was predominantly Bulgarian (95.4%), with a minority identifying as other ethnicities (4.6%). In terms of practice settings, approximately one-third were general practitioners (29.7%), while half worked in outpatient units (50.5%), and the remaining portion practiced in hospital units (19.8%) (Table 1).

Table 2 provides a detailed analysis of the factors influencing antibiotic prescribing practices among physicians who reported changes in their practices following COVID-19 compared to those who reported no changes. Significant disparities were observed across various clinical scenarios, indicating nuanced shifts in prescription behaviors post-pandemic. Notably, physicians who reported changes adopted more conservative antibiotic prescribing practices when an antibiogram was conducted, with a lower mean (M) of average frequency of prescriptions (M = 2.19 ± 0.770) compared to those who maintained their habits (M = 2.68 ± 0.893), U = 2551, p < 0.0001. Conversely, physicians who altered their prescribing habits were more likely to prescribe antibiotics when patients expressed a desire for treatment (M = 1.45 ± 0.583) than their counterparts (M = 1.15 ± 0.378), U = 2725, p < 0.001.

Furthermore, physicians who changed their practices were more inclined to prescribe antibiotics due to external factors, such as incentives from pharmaceutical companies (U = 3499, p < 0.05), time constraints (U = 3158, p < 0.01), and concerns about potential complications (U = 2803, p < 0.01). Differences were also noted in prescription behaviors related to specific clinical circumstances, including patient preferences (U = 2725, p < 0.001), professional workload (U = 2870, p < 0.0001), and temporal factors, such as prescribing before holidays or weekends (U = 2870, p < 0.0001).

Additionally, physicians who reported changes in their practices were more likely to prescribe antibiotics due to perceived time constraints (M = 1.26 ± 0.530) compared to those who did not (M = 1.08 ± 0.339), U = 3158, p < 0.01. They also showed a higher tendency to prescribe antibiotics when they had established relationships with patients and opted not to conduct further tests (M = 1.91 ± 0.775) compared to their counterparts (M = 1.60 ± 0.669), U = 2854, p < 0.01. Moreover, concerns about potential complications led physicians who changed their practices to prescribe antibiotics more frequently (M = 2.26 ± 0.846) than those who did not (M = 1.92 ± 0.816), U = 2803, p < 0.01. Lastly, apprehensions regarding the prescribing behavior of colleagues also influenced prescription decisions, as physicians who changed their practices were more likely to prescribe antibiotics for this reason (M = 1.30 ± 0.720) compared to those who did not (M = 1.08 ± 0.350), U = 3190, p < 0.01.

Logistic regression was used to test what factors could possibly contribute to changes in antibiotic prescription after COVID-19. The model fit measures indicate a good fit: deviance of 194, AIC of 214, explaining around 10.2% of the variance (R²), and statistical significance in explaining the outcome variable (McF χ² = 22.1, df = 9, p = 0.009).

Age emerged as a significant factor (OR = 1.0831, p = 0.029), with each additional year associated with an 8.31% increase in the odds of reporting a change in prescription behavior. Conversely, years of professional experience were inversely related to changes in prescription behavior (OR = 0.9274, p = 0.044), with each additional year decreasing the odds by approximately 7.26%. Holding additional specializations significantly decreased the likelihood of reporting changes (OR = 0.3835, p = 0.023), with individuals with additional specializations being approximately 61.5% less likely to report changes. Gender did not have a significant impact on changes in prescription behavior (OR = 0.8552, p = 0.683). Similarly, practitioners in outpatient practice (OR = 0.4660, p = 0.085) and those in hospital practice (OR = 0.5365, p = 0.223) did not demonstrate significant effects either. Residence in villages demonstrated a trend toward significance in its association with changes in prescription behavior (OR = 7.8055, p = 0.097). Additionally, residing in small towns did not have a significant effect on changes in prescription behavior (OR = 0.3778, p = 0.136), suggesting that small-town residency does not substantially influence prescribing practices.

Age played a significant role in changes to prescription behavior (OR = 1.0831, p = 0.029), with older physicians more likely to report changes. In contrast, more years of professional experience were linked to a lower likelihood of reporting changes (OR = 0.9274, p = 0.044). Physicians with additional specializations were also less likely to report changes (OR = 0.3835, p = 0.023). While living in a village showed a trend toward significance (OR = 7.8055, p = 0.097), residing in a small town had no significant impact (OR = 0.3778, p = 0.136) (Table 3).

This study offers insights into how demographic and professional factors have influenced changes in antibiotic prescription patterns in the post-COVID-19 era. Our findings suggest that older age and fewer years of professional experience are linked to a greater likelihood of changes in antibiotic prescription behavior following the pandemic. This trend may arise from older clinicians demonstrating enhanced adaptability in the evolving treatment landscape, while less experienced practitioners may be more willing to modify their practices. Interestingly, physicians with additional specializations appear to experience a protective effect, as they were less likely to report changes in their antibiotic prescribing habits, possibly reflecting greater clinical knowledge and confidence in their decision-making.

Conversely, the observed inverse relationship between years of professional experience and changes in prescribing patterns could be attributed to entrenched prescribing habits and potential resistance to adopting new practices. Supporting this, a Canadian study of family physicians analyzed 5.6 million antibiotic prescriptions from 10,616 doctors and found that age and career stage were significant predictors of antibiotic duration; late-career and older physicians tended to prescribe longer courses compared to their early-career counterparts (Fernandez-Lazaro et al. 2019). Furthermore, Mandelli et al. (2023) report that older physicians generally prescribe fewer but higher doses of antibiotics than their younger colleagues.

Our study found significant differences in antibiotic prescribing practices between physicians who reported changes after COVID-19 and those who did not. The higher reluctance to prescribe antibiotics when an antibiogram was available among physicians who changed their prescribing habits suggests a more cautious and evidence-based approach to antibiotic selection. In contrast, physicians who changed their habits displayed an increased tendency to prescribe antibiotics in response to patient preferences or demands, as well as under the influence of external factors such as pharmaceutical incentives and time constraints. The observed differences in prescription behaviors, particularly related to clinical decision-making factors such as antibiogram availability, patient preferences, and concerns about potential complications, highlight the complex interplay between individual, professional, and contextual factors in shaping antibiotic usage patterns. Antibiograms are valuable tools for guiding empiric antibiotic prescribing, but their utilization by physicians remains suboptimal as there is often a mismatch between antibiogram recommendations and actual prescribing practices (Kaur et al. 2016). Multifaceted interventions incorporating antibiograms have shown positive impacts on antibiotic consumption, prescribing appropriateness, and treatment costs (Khatri et al. 2023). Patient preferences and demands significantly influence antibiotic prescribing by physicians, though the effect varies across settings. In outpatient care, perceived patient demand is a major driver of unnecessary prescribing (Coenen et al. 2006; Kohut et al. 2020). However, direct requests for antibiotics are rare (Mangione-Smith et al. 2001). Physicians tend to prescribe more antibiotics when they perceive patient demand, especially in cases of clinical uncertainty (Miller et al. 1999). Interestingly, in hospital settings, patient requests may decrease antibiotic prescribing (Wang et al. 2022). Effective strategies for managing patient expectations include foreshadowing non-antibiotic outcomes, using persuasion, and offering contingency plans (Mangione-Smith et al. 2001; Stivers and Timmermans 2021). Guidelines that allow for patient preferences may help reconcile evidence-based practice with patient-centered care (Brabers et al. 2017).

On the other hand, factors like gender, practice setting, and use of alternative medicine did not significantly influence changes in prescription patterns of medical doctors in Bulgaria. This is contrary to other studies in which it has been observed that factors such as physicians’ gender, practice type, and location also affect prescribing behavior (Huang et al. 2005; Lin et al. 2010; Fleming-Dutra et al. 2018; McKay et al. 2019). Male physicians are more likely to be high-volume antibiotic prescribers compared to their female counterparts. Fleming-Dutra et al. (2018) found that “male primary care physicians aged 40 to 64 years were more likely to be high-volume antibiotic prescribers than their younger female colleagues.” Similarly, McKay et al. (2019) reported that female physicians prescribed less often than male physicians did. These nuanced findings underscore the need for tailored antibiotic stewardship interventions that account for the diverse demographic and professional characteristics of healthcare providers.

Limitations of this study include the self-reported nature of the data, which may be subject to recall bias or social desirability bias, and the cross-sectional design, which precludes inferences about causal relationships. Ultimately, understanding the impact of these factors can inform the development of targeted strategies to promote responsible antibiotic use, reduce the rise of antimicrobial resistance, and ensure optimal patient outcomes in the evolving post-pandemic landscape. By addressing the specific demographic and professional influences on antibiotic prescribing, healthcare systems can implement more effective and personalized interventions to foster responsible antibiotic use and safeguard public health in the years to come.

As we look to the future, it is vital to draw lessons from the COVID-19 pandemic to shape more effective strategies for the safe and appropriate use of medicines, particularly antibiotics. In the context of shifting antibiotic prescribing behaviors, our findings highlight the complex interplay of individual, professional, and contextual factors that lead to changes in antibiotic usage patterns. Addressing antibiotic overuse requires a holistic, multifaceted approach by healthcare systems and policymakers. Tailored interventions must be developed to address the nuanced demographic and clinical characteristics of prescribers, as well as the evolving needs of patients and healthcare systems. By embracing a broad and evidence-based approach, we can empower clinicians, engage patients, and strengthen antibiotic stewardship to combat the growing threat of antimicrobial resistance and preserve public health.

Conflict of interest

The authors have declared that no competing interests exist.

Ethical statements

The authors declared that no clinical trials were used in the present study.

The authors declared that no experiments on humans or human tissues were performed for the present study.

Informed consent from the humans, donors or donors’ representatives: Participation was voluntary, and all participants provided informed consent before completing the survey. The informed consent have been deposited in the database of Google Forms.

The authors declared that no experiments on animals were performed for the present study.

The authors declared that no commercially available immortalised human and animal cell lines were used in the present study.

Funding

No funding was reported.

Author contributions

K. Zaykova conceived and designed the study, contributed to data analysis, and led the writing of the manuscript. S. P. Nikolova provided significant contributions to data analysis and statistical evaluation, and assisted in drafting the manuscript. R. Pancheva played a key role in the study’s methodology development, data collection, and interpretation of results. A. Serbezova assisted with data acquisition, contributed to the initial draft, and critically reviewed the manuscript for important intellectual content. R. Zlatanova-Velikova reviewed and revised the manuscript, provided insights on the discussion, and ensured the accuracy of the findings presented.

Author ORCIDs

Krassimira Zaykova https://orcid.org/0000-0002-8421-9493

Silviya Nikolova https://orcid.org/0000-0002-7325-0110

Rouzha Pancheva https://orcid.org/0000-0001-8286-3923

Asena Serbezova https://orcid.org/0000-0002-1949-1212

Ralitsa Zlatanova-Velikova https://orcid.org/0000-0003-0849-7341

Data availability

All data that support the findings reported in this study are available from the corresponding author upon reasonable request.

Annex A: Survey

Section 1: Informed Consent

Dear Sir/Madam,

You are invited to participate in a study on antibiotic use in Bulgaria. The increased use of antibiotics in Bulgaria is part of a broader trend observed in many developed and developing countries. However, it remains unclear how aware the public is of antibiotic treatments and what drives the overprescription of antibiotics among physicians in our country. Therefore, the main objective of this study is to determine the attitudes and practices related to antibiotic use among physicians in Bulgaria. The collected information will be used solely for research purposes.

As a participant in this study, you should be aware that: You will have the opportunity to complete the questionnaire only once. Participation in this study is voluntary and anonymous. You must be a physician to take part in the study. Completing the questionnaire takes approximately 5 minutes. There are no identified significant risks or inconveniences associated with your participation in the study.

Data Confidentiality: The collected information will remain confidential. You will not be required to provide your contact details as part of this study. Your identity as a participant in this study will remain confidential.

Contact Information: Should you need additional information, please contact me during working hours: Krasimira Zaykova, phone:…, or via email:… .

Consent:

1. If you agree to participate in the study and are a physician, please confirm your consent by selecting “Agree” and proceed to complete the questionnaire by marking your responses.

Section 2: Practice of Antibiotic Treatment

• Please specify the primary type of patients you work with:
  	Very few (under 25%)/ I have almost no such patients	About 25% of my patients	About 50% of my patients	Over 75% of my patients
  Children (up to 18 years old)
  Adults 18–60 years old
  Adults over 60 years old
• Please indicate your practice of prescribing antibiotics using the rating scale provided below.

1 – I don’t do it

2 – sometimes

3 – often

4 – always

• Question: Reasons for antibiotic prescription
  	1 – I don’t do it	2 – sometimes	3 – often	4 – always
  After antibiogram is made.
  After blood test showing bacterial infection.
  When I judge clinically that it is a bacterial infection.
  Due to the patient’s expressed desire for antibiotic treatment.
  Due to incentives from pharmaceutical companies.
  Due to lack of time.
  When I know the patient well; no further tests.
  I prescribe when I worry about possible complications regardless the results from the clinical analysis.
  I prescribe more often before holidays/weekends.
  I prescribe more often at the end of the work day.
  I prescribe more often at the beginning of the work day.
  I prescribe because I worry that another colleague will do it after me.
• Has your practice of prescribing antibiotics changed compared to before the COVID pandemic?

1 – Yes, I prescribe more often;

2 – Yes, I prescribe less often;

3 – No change;

4 – Cannot determine.

• Does the cost of the antibiotic influence your prescribing decisions?

1 – Yes, I prescribe based on the patient’s financial means;

2 – Yes, I always prescribe the cheaper option;

3 – Yes, I always prescribe the more expensive option;

4 – No, I do not consider the cost;

5 – Cannot determine.

Section 3: Your opinion on the use of antibiotics in the country as a whole

• Scale: 0 – Cannot determine 1 – Strongly disagree 2 – Mostly disagree 3 – Mostly agree 4 – Strongly agree
  	0	1	2	3	4
  Do you agree with the statement that there is a practice of overprescribing antibiotics in Bulgaria?
  Do you believe that the use of antibiotics in animals is related to the development of antimicrobial resistance in the treatment of infections in humans?
• Please specify what you believe are the reasons for the possible inappropriate prescribing of antibiotics that do not align with good medical practice in Bulgaria as a whole.

Scale: 1 – completely disagree 2 – somewhat disagree 3 – somewhat agree 4 – completely agree

Section 4: Demographic characteristics

• You are:

A general practitioner

A general practitioner with an additional specialty

A specialist working in an outpatient unit

A specialist working in a hospital unit

• Your main specialty is:

…

• Do you have a second or additional specialty?

Yes

No

• Please indicate your second or additional specialty:

…

• Please indicate how many years you have been practicing your specialty:

…

• Please indicate if you are applying knowledge from training in:

Homeopathy

Herbal Therapy

Ayurveda

Chinese Medicine

Other

• Your gender is:

Male

Female

Other

• Your age is:

…

• Your ethnicity is:

Bulgarian ethnicity

Turkish ethnicity

Roma ethnicity

I do not wish to answer

Other

• The municipality where you live is:

Regional city

Small town

Village