Extended Abstract
Introduction
Antibiotic resistance has become a global public health threat in recent years. The World Health Organization has reported increasing antimicrobial resistance and loss of effectiveness among many commonly used antibiotics. Empirical antibiotic therapy is a cornerstone of clinical management for severe infections, particularly when rapid and accurate identification of the causative pathogen is not possible. However, incorrect or inappropriate empirical antibiotic prescribing may lead to treatment failure, adverse effects, prolonged hospitalization, and, most importantly, the spread of antimicrobial resistance. Correct and timely selection of empirical therapy can be decisive in improving patient outcomes, whereas inappropriate selection is directly associated with increased mortality and healthcare costs. Antibiotic therapy based on available information may be empirical or targeted. Empirical therapy is the initial regimen used in the absence of definitive pathogen identification and antibiogram results and is generally initiated based on clinical manifestations or available guidelines. Targeted therapy is selected after pathogen identification and review of antibiogram results. Once laboratory results become available after empirical therapy has begun, treatment should be modified or discontinued if inappropriate. Deciding whether to initiate antibiotics is a daily clinical challenge influenced by clinical manifestations, comorbidities, disease severity, local epidemiology, and the risk of delayed treatment. This decision balances the true likelihood of infection against potential harms of antibiotic use, including Clostridium difficile colitis, kidney injury, disruption of the gut microbiome, and the spread of antimicrobial resistance.
Concordance between empirical therapy and antibiogram results varies across healthcare centers and depends on factors such as microorganism prevalence, local resistance patterns, and adherence to clinical guidelines. This study aimed to evaluate concordance of empirical antibiotic therapy with final prescriptions and antibiogram results.
Methods
This descriptive study was conducted during 2021 by census sampling of records from 152 hospitalized patients receiving antibiotics (80 men and 72 women), with a mean age of 54.41±20.91 years and an age range of 15-75 years, who had been admitted to the infectious diseases and surgery wards of Shahid Sayyad Shirazi Teaching Hospital in Gorgan, Iran during 2017-2019.
Extracted demographic and clinical information included age, sex, reason for and duration of hospitalization, type and indication for surgery, underlying disease, antibiotic type, dose, number, and route of administration, type of specimens sent for microbial culture, culture and antibiogram results, clinical manifestations, laboratory findings, duration of administration of each antibiotic, and the final antibiotic prescribed after antibiogram results.
Every antibiotic prescribed at admission was considered empirical therapy. A change in medication for any reason, including an inadequate clinical response or antibiogram results, was considered final therapy. The appropriateness of empirical antibiotic prescriptions was assessed by comparing extracted information with authoritative infectious-disease guidelines.
Data were analyzed using SPSS-18. Quantitative variables were described using mean and standard deviation, and qualitative variables were described by frequency distribution as numbers and percentages.
Results
Overall, 26 types of empirical antibiotics were prescribed. Ceftriaxone (16.89%), clindamycin (12.50%), and ciprofloxacin (12.06%) were the most frequent.
More than 50% of empirical prescriptions were made by internists and 28.67% by infectious disease specialists. Nearly all antibiotics except clindamycin were primarily prescribed by internists; clindamycin was prescribed exclusively by infectious disease specialists. Gentamicin was never prescribed by infectious disease specialists. Final antibiotic prescriptions were made by infectious disease specialists in 71.05% of cases and by internists in 17.76%. Infectious disease specialists had the highest prescription rates for all final antibiotics because consultation was requested after an inadequate response to empirical therapy. Amikacin, Targocid, and doxycycline were prescribed as final treatment only by infectious disease specialists. Ciprofloxacin (14.04%), clindamycin (13.20%), and ceftriaxone (12.92%) were the most frequent final prescriptions.
Of 454 empirical antibiotic prescriptions, 343 (75.55%) were confirmed in final treatment, whereas 111 (24.45%) were inappropriate and were discontinued or replaced with another antibiotic.
Doxycycline (90%), Tavanex (47.89%), and ciprofloxacin (87.2%) had the greatest concordance with final treatment, whereas gentamicin (83.33%) and ceftriaxone (40.25%) had the greatest discordance.
Comparison of empirical therapy with antibiograms showed that disks for 27 antibiotics were used overall, most frequently amikacin (68.4%), ceftriaxone (64.5%), and gentamicin (56.6%).
Of 55 patients empirically receiving ciprofloxacin, 8 underwent antibiogram testing with a ciprofloxacin disk. The microorganism was susceptible in 62.5% of cases, indicating concordance with empirical prescribing. Ciprofloxacin was used as final treatment in 50 patients.
Of 57 patients empirically receiving clindamycin, 24 were tested, and resistance was found in 79.2%; empirical prescribing was therefore inappropriate in most cases. Clindamycin was used as final treatment in 47 cases.
Of 77 patients empirically receiving ceftriaxone, 51 underwent antibiogram testing with this disk; resistance was found in 66.7%, indicating that ceftriaxone was not an appropriate initial treatment. It was prescribed as final therapy in 46 cases.
An imipenem disk was used in 80 antibiograms. Of the 7 patients empirically receiving imipenem, 4 underwent antibiogram testing, and 100% were susceptible.
Of 6 patients empirically receiving gentamicin, only one was tested and was resistant. Gentamicin was not prescribed as final treatment for any patient, although its disk was used in 86 antibiograms of clinical specimens.
For some empirically prescribed antibiotics, such as ceftazidime (30 cases), cefazolin (9 cases), and ampicillin (9 cases), no antibiogram using the corresponding disk was performed for recipients. Nevertheless, ceftazidime, cefazolin, and ampicillin disks were used in 12, 5, and 3 antibiograms, respectively, for other patients.
Antibiotics used in antibiograms included Cotrimoxaxole (81 specimens), cephalexin (76 specimens), norfloxacin (71 specimens), cefotaxime (55 specimens), carbenicillin (49 specimens), nalidixic acid (37 specimens),
co-amoxiclav (30 specimens), as well as neomycin, cephalothin, ofloxacin, and amoxicillin. These agents were either not prescribed empirically or as final treatment or accounted for less than one percent of all prescriptions. Of 454 empirical prescriptions, an antibiogram using the same antibiotic was performed in only 115 cases (25.22%).

Conclusion
The selected empirical therapy was concordant with clinical and microbiological findings in 75% of patients. However, discordance was observed between the antibiotics used in antibiograms and those prescribed clinically, which may reduce the effectiveness of microbiological testing and increase treatment costs.
Ethical Statement
This study was approved by the Research Ethics Committees of Golestan University of Medical Sciences (IR.GOUMS.REC.1399.145).
Funding
This article was derived from the dissertation of Ms. Helia Borujerdi Karimi for the degree of Doctor of Medicine at Golestan University of Medical Sciences and was funded by the Vice Chancellery for Research and Technology of Golestan University of Medical Sciences (No. 111419).
Conflicts of Interest
No conflicts of interest.
Authors' Contributions
Roghieh Golsha (M.D): Project administration and design, Interpretation of the results and Approval of the final manuscript.
Hesamaddin Shirzad-Aski (M.D): Project administration and design, Project execution and Data analysis.
Erfan Rezaieshirazi (M.D): Project execution and Data collection.
Helia Borujerdi Karimi: Project execution and Data collection.
Samira Eshghinia (Ph.D): Interpretation of the results, Drafting of the initial manuscript and Approval of the final manuscript.
Key Message: Culture-based antibiotic prescribing will result in complete treatment concordance.