Long-Term Trends in Antibiotic Use in a Low-Resource Medical Setting: Amman, Jordan (2010–2023)

Abstract

Abstract Background Antimicrobial resistance (AMR) contributes to substantial morbidity and mortality. The misuse and overuse of antibiotics are major drivers of AMR. Lack of point-of-care testing to discern viral from bacterial infections complicates appropriate management in low-resource settings. This study examines antibiotic prescription practices in Amman, Jordan. Methods We conducted three prospective viral surveillance studies at the largest public hospital in Amman Jordan (2010–2013, 2020, and 2023). Our target population for this study was children <2 years old hospitalized with fever or respiratory symptoms. Through parental interviews and chart abstractions, we collected demographic and clinical data, including antibiotic use, as well as results of any blood, urine, and cerebrospinal fluid cultures performed at providers’ discretion. We classified antibiotics according to the 2023 World Health Organization AWaRe (Access, Watch, Reserve) antibiotic classification system, focusing on the utilization of Watch group antibiotics, which are recommended when first-line Access group antibiotics are ineffective or inappropriate. Research personnel also collected nasal or throat swabs from enrolled children, and these samples were tested in a research laboratory for common respiratory viruses using RT-PCR. Providers were unaware of viral testing results. We described categorical variables using absolute and relative frequencies and summarized continuous variables using the median and interquartile range (IQR). Results Of 4,950 children enrolled, we included 4,703. The median age was 3.5 months (IQR, 1.6–8.3), and 59.4% (n=2,793) were male. Most children were enrolled from the general ward (n=4,312 [91.9%]), with the remaining 380 (8.1%) being enrolled from the pediatric intensive care unit. Only 8 children (0.2%) among those eligible for seasonal influenza vaccination (n=1,624) had been vaccinated at presentation. The five most common admission diagnoses were bronchopneumonia (n=1,255 [26.7%]), rule-out sepsis (n=1,217 [25.9%]), bronchiolitis (n=866 [18.4%]), pneumonia (n=606 [12.9%]), and paroxysmal coughing or pertussis-like cough (n=273 [5.8%]). Blood, urine, or CSF samples were collected from 2,556 children (54.5%), 355 (13.9%) of whom tested positive for at least one bacterial species (Figure 1A). During hospitalization, 4,356 children (92.6%) received at least one antibiotic. Of those, 4,227 (97.0%) received at least one antibiotic from the Watch group (Figure 1B). The three most common Access group antibiotics prescribed were ampicillin (n=830 [17.7%]), gentamicin (n=96 [2.0%]), and amikacin (n=85 [1.8%]), whereas the three most common Watch group antibiotics prescribed were ceftriaxone (n=1,956 [41.7%]), cefuroxime (n=1,628 [34.7%]), and azithromycin (n=754 [16.1%]) Overall, 3,898 children (82.9%) were positive for at least one respiratory virus (Figure 1C). Providers ordered cultures most often for children 0–2 months old (n=1,575 [73.5%]) and those with an admission diagnosis of rule-out sepsis (n=1,161 [95.4%]). Despite seasonal variation, prescription and testing practices were consistent across study years. Conclusion Inappropriate antibiotic use remains a significant healthcare challenge in Jordan. Incorporating point-of-care viral testing for suspected respiratory infections and improving diagnostic testing for bacterial infections is crucial. Preventive strategies including influenza vaccination and the use of emerging therapies for RSV or other pathogens associated with high morbidity could indirectly limit inappropriate antibiotic use. Development of context-specific management guidelines and antimicrobial stewardship programs are urgently required to guide prescription practices.