Impact of Implementing a Rapid PCR-based Algorithm for Carbapenemase-producing Enterobacterales (CPE) and Infection Control Bundle in a Tertiary Hospital

Background:

Carbapenemase-producing Enterobacterales (CPE) are critical-priority pathogens associated with increased morbidity, mortality, and healthcare costs. Screening and isolation are recommended infection control measures, yet delays inherent to culture-based methods can hinder timely decision-making and overuse limited isolation rooms.

Objective:

To compare the time to initiation and discontinuation of contact isolation-from swab request to result availability and isolation status update-before and after implementing a rapid PCR-based diagnostic protocol for CPE identification.

Design and Setting:

Retrospective, quasi-experimental before-after study at the Hospital Italiano de Buenos Aires, Argentina. The pre-intervention period covers October 1, 2023-April 30, 2024; the post-intervention period covers October 1, 2024-April 30, 2025.

Intervention:

Incorporation of real-time PCR testing (BD MAX™ System) for CPE genes (bla_KPC, bla_NDM, bla_VIM/IMP, bla_OXA-48-like) into the existing CPE surveillance and isolation reevaluation workflow. The infection control team coordinates sample requests and response actions.

Primary Outcomes:

Time difference (in hours) from surveillance swab request to isolation implementation.

Time difference (in hours) from surveillance swab request to isolation discontinuation.

Secondary Outcomes:

Time differences stratified by weekday versus weekend.

Time differences according to sampling time (08:00-16:00 vs 16:00-08:00).

Time differences according to immunosuppression status.

Time differences in ICU versus general ward settings.

Data Collection:

Four timestamps will be extracted from the electronic health record (EHR): swab request, laboratory check-in, final laboratory result, and change in isolation logo. These will allow computation of operational intervals (request → action), collection delay, processing delay, and action delay.

Statistical Analysis:

Continuous variables will be summarized as medians and interquartile ranges. Median time differences between pre- and post-intervention periods will be compared using mixed-effects linear regression adjusted for immunosuppression, ICU admission, day of the week, and public holidays. Analyses will be conducted using Stata v16.

Ethical Considerations:

The study is retrospective and minimal-risk, involving only secondary use of clinical data. It has been submitted to the CEPI (Comité de Ética de Protocolos de Investigación), Hospital Italiano de Buenos Aires (PRIISA 15728), with waiver of informed consent under CIOMS 2019 Guideline 10.

Expected Impact:

By quantifying real-time process improvements after PCR implementation, this study will provide evidence on diagnostic turnaround times and operational efficiency in infection control practices.

Inclusion criteria

• Adult patients aged ≥18 years
• Patients screened for carbapenemase-producing Enterobacterales (CPE) carriage by perianal swab within the first 5 days of hospital admission
• Patients newly identified as CPE-colonized, leading to initiation of contact isolation
• Patients found to be decolonized, leading to discontinuation of contact isolation
• Patients with indication for active surveillance at hospital admission:

• Transfer from another healthcare facility
• Hospitalization in another healthcare center within the previous month
• Patients undergoing active surveillance during hospitalization:

• First surveillance swab in high-risk neutropenic patients (HAR flag)
• First surveillance swab in immunosuppressed units, such as hematopoietic stem-cell transplant wards
• Patients evaluated for discontinuation of contact precautions who meet all of the following:

• Prior CPE-positive surveillance sample
• At least 3 months since the last positive result and last hospitalization
• No systemic antibiotic exposure during that period

Exclusion criteria

• Lost samples
• Insufficient samples
• Invalid laboratory test results requiring repeat sampling