*Article* **Effect of the Pandemic Outbreak on ICU-Associated Infections and Antibiotic Prescription Trends in Non-COVID19 Acute Respiratory Failure Patients**

**Enrico Bussolati 1,†, Rosario Cultrera 1,2,†, Alessandra Quaranta 3, Valentina Cricca 3, Elisabetta Marangoni 3, Riccardo La Rosa 1, Sara Bertacchini 3, Alessandra Bellonzi 3, Riccardo Ragazzi 1,3, Carlo Alberto Volta 1,3, Savino Spadaro 1,3 and Gaetano Scaramuzzo 1,3,\***


**Abstract:** Background: The COVID-19 pandemic had a relevant impact on the organization of intensive care units (ICU) and may have reduced the overall compliance with healthcare-associated infections (HAIs) prevention programs. Invasively ventilated patients are at high risk of ICUassociated infection, but there is little evidence regarding the impact of the pandemic on their occurrence in non-COVID-19 patients. Moreover, little is known of antibiotic prescription trends in the ICU during the first wave of the pandemic. The purpose of this investigation is to assess the incidence, characteristics, and risk factors for ICU-associated HAIs in a population of invasively ventilated patients affected by non-COVID-19 acute respiratory failure (ARF) admitted to the ICU in the first wave of the COVID-19 pandemic, and to evaluate the ICU antimicrobial prescription strategies. Moreover, we compared HAIs and antibiotic use to a cohort of ARF patients admitted to the ICU the year before the pandemic during the same period. Methods: this is a retrospective, single-centered cohort study conducted at S. Anna University Hospital (Ferrara, Italy). We enrolled patients admitted to the ICU for acute respiratory failure requiring invasive mechanical ventilation (MV) between February and April 2020 (intra-pandemic group, IP) and February and April 2019 (before the pandemic group, PP). We excluded patients admitted to the ICU for COVID-19 pneumonia. We recorded patients' baseline characteristics, ICU-associated procedures and devices. Moreover, we evaluated antimicrobial therapy and classified it as prophylactic, empirical or target therapy, according to the evidence of infection at the time of prescription and to the presence of a positive culture sample. We compared the results of the two groups (PP and IP) to assess differences between the two years. Results: One hundred and twenty-eight patients were screened for inclusion and 83 patients were analyzed, 45 and 38 in the PP and I group, respectively. We found a comparable incidence of HAIs (62.2% vs. 65.8%, *p* = 0.74) and multidrug-resistant (MDR) isolations (44.4% vs. 36.8% *p*= 0.48) in the two groups. The year of ICU admission was not independently associated with an increased risk of developing HAIs (OR = 0.35, 95% CI 0.16–1.92, *p* = 0.55). The approach to antimicrobial therapy was characterized by a significant reduction in total antimicrobial use (21.4 ± 18.7 vs. 11.6 ± 9.4 days, *p* = 0.003), especially of target therapy, in the IP group. Conclusions: ICU admission for non-COVID-19 ARF during the first wave of the SARS-CoV-2 pandemic was not associated with an increased risk of ICU-associated HAIs. Nevertheless, ICU prescription of antimicrobial therapy changed and significantly decreased during the pandemic.

**Keywords:** healthcare-associated infections; multidrug resistance; COVID-19; acute respiratory failure; mechanical ventilation; antimicrobial therapy; SARS-CoV-2 pandemic1

**Citation:** Bussolati, E.; Cultrera, R.; Quaranta, A.; Cricca, V.; Marangoni, E.; La Rosa, R.; Bertacchini, S.; Bellonzi, A.; Ragazzi, R.; Volta, C.A.; et al. Effect of the Pandemic Outbreak on ICU-Associated Infections and Antibiotic Prescription Trends in Non-COVID19 Acute Respiratory Failure Patients. *J. Clin. Med.* **2022**, *11*, 7080. https:// doi.org/10.3390/jcm11237080

Academic Editor: Michael Dreher

Received: 8 November 2022 Accepted: 25 November 2022 Published: 29 November 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

#### **1. Introduction**

Healthcare-associated infections (HAIs) and HAIs-related septic shock are the leading causes of death in noncardiac intensive care units (ICUs) and, despite advances in modern intensive care, their incidence is still rising [1]. Several factors are associated with the increase in HAIs, such as patients' comorbidities, increased use of invasive devices, longlasting antibiotic therapies and frequent contact with healthcare personnel caring for multiple patients [2–4].

The SARS-CoV-2 pandemic outbreak had an enormous impact on worldwide health, causing over 533 million confirmed cases and over 6.3 million deaths worldwide by 12 June 2022 (according to the WHO Coronavirus disease situation report). Up to 25% of infected patients were admitted to an ICU, 80% of them requiring invasive mechanical ventilation (MV) [5,6]. The magnitude of the coronavirus disease 2019 (COVID-19) pandemic required the reorganization of healthcare facilities, concerning both the increase of ICUs beds and the improvement in human and material resources. These "new" ICUs were characterized by the extensive use of personal protective equipment (PPE), increased workload and by the presence of healthcare professionals deployed from other areas [7]. All these reasons may have reduced the overall compliance with HAI prevention programs, independently of COVID-19 infection [8,9].

Although the incidence of HAIs in the COVID-19 population has been extensively studied [10–15], the indirect effect of the pandemic on the occurrence of HAIs in non-COVID-19 acute respiratory failure patients is still unknown. An association between hospitalization during the pandemic and HAIs was found in patients admitted to the neurology ward and stroke units [16], but the impact of the pandemic on HAIs in ICU non-COVID-19 ARF patients remains unknown.

Moreover, despite few data demonstrating an overall reduction in antibiotic use in outreach patients, little is known regarding the ICU antimicrobial prescription trends during the first wave of the COVID-19 pandemic.

We therefore hypothesized that the pandemic could have had indirect effects on ICU antimicrobial prescription trends and on the incidence and characteristics of ICU-associated HAIs, especially in the first wave of the pandemic.

To test this hypothesis, we assessed the incidence, characteristics and risk factors for HAIs and the ICU antimicrobial management of patients admitted to the ICU for non-COVID-19 acute respiratory failure requiring invasive mechanical ventilation during the first wave of the COVID-19 pandemic (February–April 2020). Furthermore, we compared this group to patients admitted to the same ICU during the same period in the year before the pandemic (February–April 2019).

#### **2. Materials and Methods**

#### *2.1. Study Population and Protocol*

This is a retrospective, single-center, observational cohort study of patients admitted to the ICU of the S. Anna University Hospital (Ferrara, Italy) over a period of 3 months (February, March and April) of two consecutive years, before (2019) and during the first wave (2020) of the COVID-19 pandemic. The first wave of the pandemic was defined as the time from the first detected case (31 January 2020) to the start of reopening after the national lockdown (26 April 2020). The study was approved by the institutional ethics board of Area Vasta Emilia Centrale, site in IRCCS Azienda Ospedaliera—Universitaria di Bologna, Policlinico S. Orsola-Malpighi (Protocol number 235/2022/Oss/AOUFe), and informed consent was collected or waived if collection was not possible according to the local regulations.

### *2.2. Inclusions and Exclusions Criteria*

All consecutive patients admitted to the ICU during the study period were screened for inclusion. The inclusion criteria were: age 18–90 years; invasive mechanical ventilation; ICU admission for acute respiratory failure requiring invasive mechanical ventilation; and availability of a digital clinical record with detailed information on therapy and devices used during ICU stay. Exclusion criteria were: incomplete or incorrect records; unavailability of cultural samples data during ICU stay; presence of positive cultural isolations on admission and ICU admission for COVID-19-related acute respiratory failure.

### *2.3. Study Protocol and Definitions*

For all patients admitted to the ICU and meeting inclusion criteria, data about demographics (i.e., age, sex, height, weight), comorbidities, ICU entrance diagnosis, medication before ICU admission, the Simplified Acute Physiology Score (SAPS) II, which is an index of disease severity [17], and duration of hospital stay before ICU admission were collected.

We collected data on ventilatory features (duration of invasive and non-invasive ventilation, oxygen therapy, tracheostomy and eventually prone positioning), invasive device features (central venous line, midline and arterial line), and presence and duration of laparostomy. Ventilatory free days (VFDs) were calculated as previously described [18]. As concerns antimicrobial therapy, we defined it as prophylactic, empiric or target according to the evidence of infection when the antimicrobial treatment was started. Specifically, we defined as (1) prophylactic any antimicrobial therapy prescribed in the absence of any sign and symptom of infection (e.g., fever, leukocytosis, increase of PCR/procalcitonin); as (2) empiric any therapy initiated without any positive cultural isolation in presence of signs and/or symptoms of infection; and as (3) target any therapy started after positive cultural isolation.

We also defined days on antimicrobial therapy as the number of days on antimicrobial treatment, independently of how many drugs were prescribed at the same time. Total antimicrobial use was defined as the cumulative sum of days on therapy for all antimicrobials during ICU stay, as previously defined by Campbell et al. [19]. Outcomes regarding length of ICU stay, mortality, microbiologic isolations (bloodstream, respiratory tract and urinary tract cultures) and multidrug resistance were also collected.

We defined HAIs as infections acquired at least 48 h after ICU admission. Bloodstream, respiratory tract and urinary tract microbial identification, antimicrobial susceptibility, multidrug resistance and MIC interpretation were defined as previously described by Cultrera et al. [10]. The isolations referring to blood, respiratory and urine cultures were requested by the attending physicians for patients with suspected secondary infections because of clinical and/or respiratory deterioration associated with suggestive laboratory or radiological findings.

#### *2.4. Statistical Analyses*

Categorical variables are reported as frequency, while continuous variables as mean ± standard deviation or median [interquartile range], according to data distribution (normal/not normal). Considering the absence of evidence regarding HAIs in non-COVID-19 patients during the first wave of the COVID-19 pandemic, we could not calculate a priori the sample size and therefore aimed to enroll the higher number of patients admitted to the ICU during the study period. Patients were assigned to one of the groups (PP and IP) based on the year of ICU admission.

Bivariate comparisons regarding nominal data were conducted using Pearson's chisquare test. The Shapiro–Wilk test was used to verify continuous variables distribution. Student's *t*-test and Mann–Whitney U test were used to compare the two samples (depending on normality distribution). Logistic regression technique was performed to evaluate risk factors associated with HAIs, and the outcome was defined as presence/absence of HAIs during ICU stay. The predictors inserted in the regression model were: the year of admission, positive history of Diabetes Mellitus, Chronic Kidney Disease (CKD), smoking, obesity (defined as BMI > 30), ICU length of stay, duration of invasive mechanical ventilation and duration of steroid therapy.

A linear multiple regression was used to test the association of SAPS II, presence of heart diseases, lung diseases, DM, CKD, year of ICU admission, duration of ICU stay, duration IMV and admission to the ICU after surgery with total antimicrobial use. Significance was set at *p* < 0.05. Statistical Analysis was performed using SPSS 24 (IBM Corp. Released 2016. IBM SPSS Statistics for Windows, Version 24.0. IBM Corp, Armonk, NY, USA) and GraphPad Prism version 8.0.0 for Windows (GraphPad Software, San Diego, CA, USA, www.graphpad.com, accessed on 1 June 2022).

#### **3. Results**

#### *3.1. Baseline Population Characteristics*

Figure 1 describes the patient selection process. Two-hundred and eleven patients were admitted to the ICU during the study period and screened for inclusion. After evaluating for inclusion and exclusion criteria, a total of 83 patients were included in the study. Their main clinical characteristics are described in Table 1.

**Table 1.** Baseline characteristics at ICU admission, comorbidities, and entrance diagnosis.


Data are expressed as Mean ± SMean SD or Number (%), according to the data. PP, pre-pandemic; IP, intrapandemic; BMI, Body mass index; CKD, chronic kidney disease; DM, diabetes mellitus; SAPS II, Simplified Acute Physiology Score II. Italic for categories.

When comparing the PP and the IP groups, the mean age of the two groups was comparable (Table 1). No significant differences were observed either in anthropometric parameters or comorbidities, with the sole exception of chronic kidney disease (*p* = 0.016). No significant differences were seen in the ICU entrance diagnosis (*p* = 0.35) and in the duration of hospital stay before ICU (*p* = 0.52, Table 1).

#### *3.2. Clinical Features*

The clinical characteristics during the ICU stay are resumed in Table 2. ICU length of stay (*p* = 0.92), ICU mortality (*p* = 0.68), the duration of invasive ventilation (*p* = 0.41), VFDs (*p* = 0.12) and the number of patients undergoing non-invasive ventilation, oxygen therapy, tracheostomy and pronation were not different between the two groups.

**Figure 1.** Flow diagram of patient selection process.


Data are expressed as Mean ± SD, Median [IQR] or Number (%), according to the data. PP, pre-pandemic; IP, intra-pandemic; CVC, central venous catheter. Italic for categories.

A similar number of patients between the two groups had a central venous line, a midline and/or an arterial line. In the IP group, no patient had a first central venous line inserted at the femoral site, consequently resulting in an increased number of jugular and subclavian insertion sites, although this did not reach statistical significance. The duration of steroid therapy was significantly shorter in the IP group (7.5 ± 12.3 days (PP) and 3.1 ± 5.8 (IP) (*p* = 0.038)).

#### *3.3. Antimicrobial Therapy*

In the IP group, the total antimicrobial use was significantly shorter (11.6 ± 9.4 days) than in the PP group (21.4 ± 18.7 days, *p* = 0.003), while the days on antimicrobial therapy were similar between the groups (6.6 ± 5.2 vs. 6.6 ± 4.2, *p* = 0.97, Table 3). The year of ICU stay was also independently associated with total antimicrobial use when adjusting for possible confounders in the regression model (Std. Beta 0.280, *p* = 0.003, Table S1). The duration of 2nd and 3rd antimicrobials were significantly shorter in terms of days in the IP group (*p* = 0.037 and *p* = 0.019, respectively).

**Table 3.** Antimicrobial therapy, cultural samples and infections in the study population, divided for year of admission.


Data are expressed as Mean ± SD, Median [IQR] or Number (%), according to the data. PP, pre-pandemic; IP, intra-pandemic; MDR, multidrug-resistant.

A higher number of patients underwent prophylactic therapy (*p* = 0.03) and a lower number of patients underwent empiric therapy (*p* = 0.05) in the PP group (Figure 2). Despite this, the number of days for each therapy was not significantly different, except for target therapy, which decreased from PP to IP (*p* = 0.03). No significative differences could be found between the groups in the antimicrobial class prescription. Nevertheless, there was a tendency towards an increased prescription of Penicillin and Carbapenems and a decreased prescription of antifungals (Figure 3).

**Figure 2.** Comparison of different antimicrobial approaches before and during the pandemic. (**a**) Percentage of patients who underwent prophylactic, empiric and target therapy in the two study populations; (**b**) Number of days undergoing prophylactic, empiric and target therapy in the two years of analysis, i.e., pre-pandemic (2019) and intra-pandemic (2020). PP, pre-pandemic; IP, intra-pandemic.

**Figure 3.** Antimicrobial classes and relative percentages regarding every administered drug in the two years of analysis (pre-Pandemic and intra-pandemic).

#### *3.4. Cultural Isolations*

There were no significant differences between the two groups in the number of cultural samples/patient, positive cultural samples/patient and number of patients developing HAIs and MDR infections, as shown in Table 3. Microbial isolations in blood, respiratory tract and urinary tract were divided in families and differences are summarized in Table S2 and Figure S1. The only significant difference between PP and IP was the increased number of *Candida* spp. isolations (*p* < 0.001) in the IP group, mostly isolated from the respiratory and urinary tract.

## *3.5. Multivariate Analysis*

The multivariate logistic regression analysis on risk factors associated with HAIs is shown in Table 4. In the analysis, only CKD was significantly associated with HAIs (*p* = 0.024), while being admitted to the ICU during the pandemic was not.

**Table 4.** Multivariate logistic regression analysis on risk factors associated with healthcare-associated infections, with outcome defined as presence/absence of HAIs during ICU stay.


Reference in parenthesis; IP, Intra-Pandemic (2020); CKD, chronic kidney disease; DM, diabetes mellitus; ICU, intensive care unit; IMV, invasive mechanical ventilation.
