Introduction

The global health systems and antimicrobial resistance (AMR) were significantly affected by the COVID-19 pandemic [1] with AMR contributing to 4,95 million deaths in 2019, of which 1,27 million were directly linked to drug resistance [2]. Combating AMR requires a multifaceted approach [3] but COVID-19 has hindered progress [4, 5]. Various factors, including antibiotic use, infection prevention and control, public health policy making, and cross-border spread, influence AMR either positively or negatively [5]. While the impact of COVID-19 on AMR is not fully understood yet, ongoing surveillance is crucial to monitor trends during the pandemic. In Greece, healthcare-acquired infections affect approximately 121,000 patients annually, with a prevalence 50% higher than the European average [6, 7], often due to multidrug-resistant organisms (MDRO) [8]. Since the late 2000s, Greece has been facing an endemic situation in hospitals, specifically carbapenem-resistant Gram-negative bacilli, and the country’s AMR rates rank among the highest in Europe [9]. This study aimed to determine rates of hospital-onset MDRO bloodstream Infection (BSI) and overall Infection/Colonization among inpatients of three tertiary Greek hospitals, providing insights into increased rates during the COVID-19 years and supporting efforts to reduce MDRO rates and enhance patient care quality.

Methods

Study design

This 4-year study, from January 2019 to December 2022, involved three tertiary Greek hospitals: Attikon University Hospital of Athens (no of beds: 730), AHEPA University Hospital of Thessaloniki (no of beds: 680) and Tzaneio General Hospital of Piraeus (no of beds: 454). The first two are major teaching hospitals with medical and postgraduate student programs, while the third is a tertiary hospital offering residency and/or fellowship training programs. All handle complex cases in internal medicine, surgery, and pediatrics, and have intensive care units (ICUs).

Starting from January 2019, we monitored Multidrug-Resistant Organisms (MDROs), following the CDC protocol [10]. The surveillance protocol relied on culture results of clinical samples that were collected from hospitalized adult and pediatric patients. Active surveillance samples were not included and clinical evaluation of cases was not considered. Monitored MDROs included carbapenem-resistant Enterobacterales (CRE), carbapenem-resistant Acinetobacter baumannii (CRAB), carbapenem-resistant Pseudomonas aeruginosa (CRPA), vancomycin-resistant enterococci (VRE), and methicillin-resistant Staphylococcus aureus (MRSA).

All MDRO isolates from any specimen were examined to identify Laboratory-Identified (LabID) Events monthly and by location group (ICU, non-ICU). First MDRO isolates per patient and month were reported as LabID events regardless of specimen source; if a duplicate MDRO isolate was from blood, a LabID event was reported only if it represented a unique blood source (no prior MDRO isolation in the blood from the same patient within 14 days, even across calendar months]. Each LabID event was reported individually.

Surveillance was conducted prospectively with local investigators trained in the CDC protocol. Total patient days obtained from hospital records monthly served as denominator data.

Microbiological methods

Species were identified by MALDI-ToF (Microflex LT system; Bruker Daltonics, Italy) and antimicrobial susceptibility was tested by BD Phoenix 50 (Becton Dickinson Diagnostic Systems, Hunt Valley, MD) or VITEK 2 (bioMérieux, Marcy l’Etoile, France) system. MDROs were defined by CDC protocol [10] and expanded to accommodate Greece’s endemic situation: (i) CRE: Escherichia coli, Klebsiella oxytoca, Klebsiella pneumoniae, Klebsiella aerogenes or Enterobacter spp. resistant to imipenem and meropenem, (ii) CRAB: Acinetobacter baumannii resistant to imipenem and meropenem, (iii) CRPA: Pseudomonas aeruginosa resistant to imipenem and meropenem, (iv) VRE: Enterococcus faecalis or Enterococcus faecium resistant to vancomycin, and (v) MRSA: Staphylococcus aureus resistant to oxacillin/cefoxitin. CRE isolates with elevated MICs for imipenem and meropenem underwent NG-Test CARBA 5 immunochromatographic assay to detect the five most common carbapenemases (KPC, IMP, NDM, VIM, and OXA-48) rapidly and accurately. Breakpoints were interpreted following the European Committee on Antimicrobial Susceptibility Testing guidelines [11].

MDRO data analysis

LabID Events were classified as Community-Onset (CO) if the specimen was collected within the first 3 days after hospital admission, or Hospital-Onset (HO) if collected more than 3 days after admission.

For each MDRO, location group, hospital and year we reported:

  1. i.

    Measures for MDRO Bloodstream Infection:

  • HO unique blood source LabID events (from blood specimens only).

  • MDRO BSI Incidence Density Rate (IDR): number of unique blood source HO LabID events, divided by patient days for the location group x 1,000.

  1. ii.

    Measures for MDRO Healthcare Acquisition:

  • first new HO LabID events per patient (from any specimen, no prior evidence of any infection or colonization for that specific MDRO from a previous LabID Event).

  • Overall MDRO Infection/Colonization IDR: number of first new HO LabID events per patient, divided by patient days for the location group x 1,000.

Pooled MDRO IDR was presented for each location group and year with monthly and run charts for CRE and CRAB bloodstream Infection IDR. Data were categorized into pre-COVID-19 (2019) and COVID-19 (2020–2022) to assess the pandemic’s impact on MDRO incidence. Changes in IDR from pre-COVID-19 to COVID-19 years were analyzed and the ratio of these two rates (incidence rate ratio) with its 95% Confidence Interval and associated P-value [12, 13] was calculated to determine the impact [14].

To examine the prevalence of the most common carbapenemase genes across different years, we used the study population of first new HO CRE LabID Events.

Results and discussion

Table 1 presents the pooled patient days, annual MDRO BSI and annual overall Infection/Colonization IDR for all location groups during the surveillance period. Compared to 2019, pooled patient days decreased in non-ICU but increased in ICU. The annual MDRO rates for each hospital are presented in Supplementary Table 1.

Table 1 Pooled multidrug-resistant organism incidence density rates by year and pathogen and Incidence rate ratios following the COVID-19 pandemic in three tertiary Greek hospitals, 2019–2022
Full size table

Compared to 2019, CRE BSI IDR increased significantly in the wards during all COVID-19 years (112% in 2020, 127% in 2021, and 170% in 2022), while in ICUs the increase was significant only in 2021 (73%). Supplementary Fig. 1 shows a run chart tracking monthly changes in CRE and CRAB BSI IDR from 2019 to 2022. Overall CRE Infection/Colonization IDR aligned consistently with BSI IDR in both location groups. Previous reports showed varied CRE trends [15,16,17] with risk factors including prolonged hospital stays, invasive devices, carbapenems and corticosteroids exposure, and ICU admission [18, 19]. Inappropriate use of antibiotics in COVID-19 patients and poor adherence to infection prevention and control (IPC) practices [5] contribute to CRE emergence and transmission [20]. Interestingly, we observed that the distribution of carbapenemases has changed over time during the COVID-19 era. Among CRE isolates, KPC was the most common carbapenemase (60.9% in 2019, 66.7% in 2022) (Table 2). Metallo-β-lactamase (MBL)-producing isolates decreased from 30.0 to 18.4%, while KPC plus MBL increased (from 9.1 to 13.5%). Despite KPC and NDM being recognized as predominant carbapenemases in COVID-19 patients [21], few studies have assessed longitudinal trends in carbapenemase prevalence during the COVID-19 period, which is crucial for effective treatment strategies and infection control.

Table 2 Carbapenemases/β-lactamases in the group of hospital onset carbapenem-resistant Enterobacterales Laboratory-Identified Events by year in three tertiary Greek hospitals, 2019–2022
Full size table

In the ICU, the CRAB BSI IDR increased significantly (142%) from 4.24 infections/1,000 patient days (pd) in 2019 to 10.26 infections/1,000 pd in 2021 and declined to 5.17 infections/1,000 pd in 2022. Non-ICU faced very significant increases in CRAB BSI IDR, which was 241% higher in 2022 than 2019. CRAB Infection/Colonization IDR paralleled BSI IDR. The global rise in CRAB during the pandemic, reported as increased incidence or prevalence is concerning [16]. A. baumannii is commonly isolated among COVID-19 patients and survives for long periods on surfaces and under dry conditions [22], making prioritized preventive measures crucial. Challenges during the pandemic may compromise these measures, contributing to increased CRAB transmission. Overuse of carbapenems among hospitalized COVID-19 patients, changes in compliance to IPC programs and practices (difficulties in adhering to standard IPC precautions, inappropriate glove use, low health care workers-to-patient ratios, temporary discontinuation of conventional IPC efforts) (5) and other risk factors (prolonged hospitalization, immunosuppressive therapy, invasive devices, ICU admission), could also contribute to this surge in CRAB infections [23].

A significant increase (187%) in CRPA BSI IDR was observed in non-ICU during 2022 compared with 2019 (from 0.04 to 0.13 infections/1,000 pd). The overall Infection/Colonization IDR followed similar trend. In our study, CRPA BSI remained stable, except for a notable increase in 2022 in the wards. Literature on the pandemic’s impact on CRPA shows conflicting results, with some studies reporting a decrease and others an increase [16, 17].

In both ICU and non-ICU settings, VRE BSI IDR in non-ICU increased from 0.03 in 2019 to 0.17 and 1.07 infections/1,000 pd in 2021 and 2022, respectively. In ICU, the IDR rose from 0.44 infections/1,000 pd in 2019 to 1.56 in 2021 and 1.03 in 2022, making a substantial rise of almost 260% in ICU and 380% in non-ICU in 2021. Notably, the COVID-19 pandemic significantly increased VRE BSI and overall infection/colonization in 2021 and 2022, consistently with other studies [16, 17]. Disruption of certain infection control measures (targeted screening, isolation measures, cleaning and disinfection) [24] along with inappropriate use of antibiotics, such as vancomycin and third-generation cephalosporins, may contribute to this rise [25, 26]. Notably, a few studies reported a reduction in VRE incidence, possibly due to improved hygiene practices [16].

Between 2019 and 2022, overall MRSA BSI IDR remained stable in both ICU and non-ICU. For non-ICU, we observed a decrease of 36% in the IDR of MRSA overall infection/colonization in non-ICU settings during 2021 (Table 2). However, more than 50% of studies reported increased MRSA prevalence/incidence during the COVID-19 pandemic [16, 17] showing geographic disparities and variations. Despite the evidence from a meta-analysis linking antibiotic exposure to MRSA [27], Greece saw higher systemic antibacterial use in 2021 compared to 2019 [28]. Pre-pandemic guidelines on early detection and isolation of MRSA positive patients were proposed [29], but compliance appears unchanged, with no unavailable data. Various complex factors are likely to influence MRSA rates.

The current study has both strengths and limitations. This study is the first to establish rates for hospital-onset MDRO BSI and overall Infection/Colonization among inpatients in Greek hospitals providing a perspective on the increased rates observed during the COVID-19 years. Also, it stands out as one of the few studies that specifically examine MDRO rates during the COVID-19 era in two distinct settings: ICU and non-ICU. Additionally, the data from three hospitals, located in the three largest Greek cities (Athens, Piraeus, and Thessaloniki) were merged using the same validated methodology from the CDC, enabling reliable conclusions. However, variability between settings underscores the need for continuous monitoring within each hospital. While data covers geographically distinct regions, a lack of a representative sample hinders national benchmarking. Furthermore, the focus on inpatients limits generalizability to the community. Lastly, despite the absence of clinical evaluation of the patient to distinguish between infection and colonization, the observed trends closely parallel those seen in BSIs suggesting a common trend between these two categories.

In conclusion, compared to 2019, non-ICU settings experienced significant increases in CRE, CRAB, and VRE throughout all COVID-19 years, with CRPA increasing only in 2022 and MRSA decreasing in 2021. In ICUs, CRE increased only in 2021, CRAB in 2020 and 2021, and VRE in both 2021 and 2022. KPC was the predominant carbapenemase among CRE in all years, while MBL decreased, and CRE with both KPC and MBL increased in 2022. The COVID-19 pandemic has led to increased MDRO rates due to disruptions in infection control, reduced contact precautions and surveillance programs, challenges in isolating infected patients and increased antibiotic use among COVID-19 patients. Continuous surveillance, infection control measures and antimicrobial stewardship interventions are recommended to reduce further spread of MDROs within healthcare settings.