Impact of the COVID-19 pandemic on the incidence of multidrug-resistant bacterial infections in an acute care hospital in Brazil

Dear editor,

We appreciate the contribution of Lima et al to the discussion of this important issue. While their results appear different, it is important to appreciate the numerous differences in our methodologies when making that comparison. (1) While we both analyzed the incidence of MDR infections, Lima et al looked only at ICU populations, while we looked more broadly at all inpatient units (ICU and non-ICU). (2) Lima et al considered CRAB, while we looked at 5 pathogens, noting differences in CRAB and MRSA. (3) The median of Acinetobacter baumannii MDR infections calculated by Lima et al is per patient-day/100 beds, while our median was calculated using the number of MDR infection cases in the period as numerator and the number of patient-days in the same period multiplied by 1,000. (4) While we both used the same statistical test to compare medians (Mann-Whitney), we used different denominators.¹ (5) Our definitions of MDR are also different. While Lima et al considered MDR to be a pathogen resistant to at least 1 agent in 3 or more categories of antimicrobials,² we used the definition for A. baumannii of resistance to the carbapenem class, which is one of the main drugs for the treatment of serious infections caused by them.

Our approach to analysis also differed due to different objectives. We agree with Lima et al that time series analysis is the preferred statistical method for before and after studies, but the purpose of this study was to verify whether there was a statistical difference in the incidence of resistant bacteria between prepandemic years and during the early months of the pandemic. While time series events on preliminary or indicator data can be useful³—as Lima et al demonstrate, we are deferring a time series analysis of our data until after the pandemic is over.

We were able to identify the outbreaks and both are well described in the article, but the increased use of the number of invasive devices is only a hypothesis, as neither study compared their usage rates in the pre and pandemic periods, although we know that COVID patients are more severe and require a greater number of invasions.

We recognize the limitations of our study, as it is a single center, but with exclusive dedication to the care of COVID for 5 months. We believe that in this period exclusively dedicated and with 200 beds of ICU the impact of COVID-19 can really be measured, despite the biases described.

In summary, the results found by Lima et al may show a transient impact on the incidence of A. baumannii MDR by their definition, but they are different from our results. We agree that the impact of COVID-19 on hospital-acquired MDR infections may be heterogeneous across healthcare settings.