Results

In total, 27,718 urine cultures with E. coli were reviewed from 2018 to 2021. On average, 6929.5 urine cultures were performed per year. For E. coli, the resistance rates to ceftriaxone were 25.21% in 2018, 26.05% in 2019, 24.84% in 2020 and 24.65% in 2021 (Table 1). The resistance rates to meropenem were 0.08% in 2018, 0.13% in 2019, 0.10% in 2020, and 0.30% in 2021 (Table 1). The average E. coli with ceftriaxone resistance rates were 25.63% and 24.75% before and after the COVID-19 pandemic, respectively (Fig. 2). The average rates of E. coli with CRE were 0.11% and 0.20% before and after the COVID-19 pandemic, respectively (Fig. 3). Between 2020 and 2021, a 200% increase in carbapenem-resistant E. coli occurred.

Table 1. Ceftriaxone Resistance (Possible ESBL) and Meropenem Resistance (CRE) in E. coli Isolates from Urine

Note. ESBL, extended-spectrum β-lactamase; CRE, carbapenem-resistant Enterobacterales.

Fig. 2. Escherichia coli from urine with ceftriaxone resistance and Pseudomonas aeruginosa (PSAR) from bodily fluids with ceftazidime resistance (possible extended-spectrum β-lactamase or ESBL) before and after the COVID-19 pandemic.

Fig. 3. Escherichia coli from urine and Pseudomonas aeruginosa (PSAR) from bodily fluids with carbapenem resistance before and after the COVID-19 pandemic.

In total, 2,111 bodily fluid cultures with PSAR were reviewed from 2018 to 2021. On average, 527.75 cultures were performed per year. For PSAR, the resistance rates to ceftazidime were 11.79% in 2018, 13.48% in 2019, 15.19% in 2020 and 15.13% in 2021 (Table 2). The resistance rates to meropenem were 2.12% in 2018, 2.53% in 2019, 2.33% in 2020, and 5.35% in 2021 (Table 2). The average rates of PSAR with ceftazidime resistance in bodily fluid were 12.64% and 15.16% before and after the COVID-19 pandemic, respectively (Fig. 2). The average rates of PSAR with carbapenem resistance were 2.33% and 3.84% before and after the COVID-19 pandemic, respectively (Fig. 3). Between 2020 and 2021, a 130% increase in CR-PSAR occurred.

Table 2. Phenotypic Resistance Consistent With ESBL (Ceftazidime Resistance) and CR-PSAR (Meropenem Resistance) in PSAR Isolates From Bodily Fluids

Note. ESBL, extended-spectrum β-lactamase; PSAR, Pseudomonas aeruginosa; CR-PSAR carbapenem-resistant PSAR.

In conclusion, we detected high levels of resistance in E. coli isolated from urine samples and rising resistance to carbapenems in PSAR from bodily fluids. On average, resistance to ceftriaxone was present in 25.19% of E. coli isolates from urine each year. These resistance patterns suggest that ESBL is common in the Dominican Republic, at levels comparable to other countries in Latin American. ^{Reference Gales, Castanheira, Jones and Sader12–Reference Guzmán-Blanco, Labarca, Villegas and Gotuzzo14} Carbapenem resistance was rare in both E. coli and PSAR, averaging 0.15% and 3.08% each year, respectively. However, an increase was seen after the COVID-19 pandemic, particularly in PSAR from bodily fluids. With bodily fluids as a source, this increase may be in more complex patients than those from urine samples and raises the concern of a similar effect in hospitalized patients.

In Latin America, the use of azithromycin for COVID-19 was widespread. ^{10,Reference Tapia15} Antibiotic use increased in hospitalized and ambulatory patients alike, despite efforts from professional societies to discourage their use. ^8–10 Reports showing an increase in AMR in Latin America after the pandemic are emerging. ^{Reference Thomas, Corso and Pasterán16} This finding mirrors a rise of AMR in the United States, where nosocomial infections by ESBL increased by 32%, CRE increased by 35% and AMR PSAR increased by 132% in the first year of the pandemic. ¹¹

Excess antibiotic use is a significant problem in LMICs. ^{Reference Muri-Gama, Figueras and Secoli4,Reference Shet, Sundaresan and Forsberg5} In the Dominican Republic, antimicrobials can be purchased without prescriptions and are available in pharmacies and stores. ^{Reference Mena Lora, Rojas-Fermin, Bisono, Almonte and Bleasdale6} Aminopenicillins are commonly used and may contribute to high rates of ESBL. Studies have reported ceftriaxone resistance in >24% of E. coli at a pediatric hospital and 46% in hospitalized adults. ^{Reference Luna, Lora, Roque, Franceschini, Perez and Cabán17,Reference Mena Lora, Rodriguez Abreu, Blanco, de Lara and Bleasdale18} Circulating ESBL genotypes in the Dominican Republic include blaCTX and blaTEM. ^{Reference Roque, de Luna and Mena Lora19} Future ASP interventions must focus on both ambulatory and hospital settings. Treatment guidelines reflecting local susceptibilities reported herein can be an important patient safety and stewardship tool.

A strength of our study was the large number of samples and wide geographic representation of samples in the Dominican Republic. These samples may provide a valuable portrait of susceptibilities in these communities. The study also had several limitations. All cultures performed during the study period were included; thus, patients with recurrent infections may be overrepresented. The large number of isolates in the study may balance this weakness. Inferring resistance based on phenotypic resistance patterns rather than genotypes is the main limitation of our study. Ceftriaxone nonsusceptibility without ESBL genes can occur. MIC accuracy can be affected by bacterial inoculum, antimicrobial dilutions, and differing enzyme expression. ^{Reference Villegas, Esparza and Reyes20} High inoculum can occur in severe infections and may have an impact on MICs. Ceftazidime resistance may indicate ESBL in PSAR. However, genes other than ESBL may also contribute to ceftazidime nonsusceptibility in PSAR. The lack of patient specific information, such as comorbidities or prior antibiotic use, is another weakness in our study. Despite these weaknesses, we have reported susceptibilities using the same methodology used in routine clinical practice and may provide valuable information for empiric antimicrobial selection and the development of local treatment guidelines.