To the Editor—According to a prospective multicenter cohort study, when the proportion of patients in contact isolation increases, compliance with contact isolation precautions decreases.Reference Dhar, Marchaim and Tansek ¹ European Society of Clinical Microbiology and Infectious Diseases guidelines recommend contact precautions for extended-spectrum β-lactamase (ESBL)–producing Enterobacteriaceae with the exception of ESBL-producing Esherichia coli (ESBL-EC) due to low transmission rates in nonepidemic settings (except intensive care units [ICUs] and hematolocigal units).Reference Tacconelli, Cataldo and Dancer ² ESBL-EC caused a therapeutic challenge, and a high rate of carbapenem usage, resulting in the first clinical strain of carbapenem-resistant E. coli at our hospital.Reference Metan, Zarakolu, Cakir, Hascelik and Uzun ^{3 ,} Reference Gulmez, Woodford and Palepou ⁴ We implemented contact precautions for all patients colonized or infected with ESBL-EC until 2016. However, with this approach, a high number of patients required contact precautions.

Hacettepe Adult Hospital is a tertiary care center with 700-beds. The adult hospital has 3 surgical ICUs with 42 beds and a medical ICU (MICU) with 9 beds. Hacettepe Oncology Hospital has a total of 114 beds, including an 8-bed ICU and a 24-bed bone marrow transplantation (BMT) unit. Transfer of patients with cancer between 2 hospitals is common. All rooms have a single bed in the BMT unit. The other hospital rooms have 1 or 2 beds. There is no standard distance between beds in rooms with 2 beds, but usually this distance is no shorter than 1.5 meters. Alcohol-based hand rub and gloves are available at the bedside for all patients. A patient-based infection control program has been in place for more than 20 years. An infection control nurse visits all patients hospitalized at ICUs daily to detect ICU-acquired infections. Patient-based nosocomial infection surveillance is conducted for certain surgical procedures. Hospital-wide nosocomial bacteremia surveillance is performed by following the microbiology laboratory results daily. Infection control nurses contact the wards when contact precautions are required according to the culture results. Patients who need contact precautions are electronically flagged in the hospital’s electronic web system until the infection control team cancels them. Cautionary cards are placed at the bedside when a patient is isolated. Any clustering of multidrug-resistant bacteria in the same ward is discussed with an infectious disease specialist, and outbreak investigations begin as quickly as possible. Contact precautions are implemented for all patients infected or colonized with ESBL-producing Enterobacteriaceae, carbapenem-resistant (CR) Enterobacteriaceae, multidrug-resistant Acinetobacter baumannii, or multidrug-resistant Pseudomonas aeruginosa. Contact precautions were abandoned for patients infected or colonized with ESBL-EC from January 1, 2016, because the evidence for the usefulness of contact precautions for ESBL-EC is low. For neutropenic cancer patients, we recommend 1-bed rooms. When a 1-bed room is not available, these patients are not admitted to rooms with patients colonized or infected with ESBL-EC. The medical devices used were either for single use or were disinfected before use with neutropenic patients.

During the study period, the identification of the isolates was performed using VITEK MS matrix-assisted laser desorption/ionization–time-of-flight mass spectrometry (MALDI-TOF MS). Antimicrobial susceptibility testing was performed with VITEK 2 (BioMérieux, Marcy-l'Étoile, France), and results were interpreted according to Clinical and Laboratory Standard Institute recommendations. When specific tests for ESBL production were not reported, E. coli and K. pneumoniae resistant to third-generation cephalosporins were considered ESBL producers for infection control purposes.

We compared the rates of ESBL-EC, ESBL-producing K. pneumoniae, CR E. coli, and K. pneumoniae between 2015 and 2016 to determine the results of our intervention. OpenEpi version 3.01 (www.openepi.com) was used for data analysis. Infection density rate (IDR), conditional maximum likelihood estimate (CMLE) of rate ratio (RR), 95% confidence intervals (CI), and P values were calculated. The Fisher exact test was used to compare IDRs among years. P<.05 was considered statistically significant.

The IDR did not increase for ESBL-EC after cessation of contact precautions in our hospital. Also, no change was observed for IDR for ESBL-producing K. pneumoniae or for CR K. pneumoniae between 2015 and 2016. An increase in CR E. coli bacteremia at the Oncology Hospital was observed, but it was not statistically significant (Table 1).

TABLE 1 Infection Density Rates and Rate Ratios of Pathogens

NOTES. CI, confidence interval; CMLE, conditional maximum likelihood estimate; CR, carbapenem (ertapenem or imipenem or meropenem) resistant; ESBL, extended-spectrum β-lactamase; IDR, infection density rate.

^a The Fisher exact test was used.

A recent Swiss study showed the safety of cessation of contact precautions for ESBL-EC in a setting where compliance with standard infection control precautions and hand hygiene is high.Reference Tschudin-Sutter, Frei and Schwahn ⁵ Compliance with infection control precaution is highly variable in our hospital. The rate of compliance with hand hygiene before patient contact is nearly 90% in the oncology ICU and BMT units; however, it was 30%–60% in the surgical ICUs. Nevertheless, we did not observe an increase in the rate of ESBL-EC bacteremia.

This study has some limitations. First, we did not compare the types of ESBL-EC infection other than bacteremia between 2015 and 2016, but no clusters of ESBL-EC infections were detected in any of the wards during surveillance activities. Bacteremia surveillance is the only type of surveillance that is performed hospital-wide, so we decided to compare the bacteremia rates. Also, we did not have access the molecular epidemiology of ESBL-EC because it is very difficult to analyze the genetic relatedness of ESBL-EC in daily practice for infection control purposes.

Despite all limitations, our study showed that, in a middle outcome country where compliance to infection control precaution is highly variable, cessation of contact precautions for ESBL-EC did not result in a negative outcome. However, infection control teams practicing in crowded hospitals under high workload with insufficient staff should be cautious because ESBL-EC outbreaks are common.