Evaluating compliance with infection prevention protocols in red-box rooms via remote video auditing

• Published in: Infection control and hospital epidemiology Vol. 42; no. 10; pp. 1 - 1284
• Main Authors: Singh, Vansha, Malhotra, Prashant, Khatri, Akshay, Khameraj, Aradhana, Rasul, Rehana, Schwartz, Rebecca M, Farber, Bruce
• Format: Journal Article
• Language: English
• Published: United States Cambridge University Press 01-10-2021
• Subjects: Antibiotics; Disease control; Health care; Infections; Medical instruments; Nursing; Patient safety; Prevention; Protective equipment
• ISSN: 0899-823X; 1559-6834
• DOI: 10.1017/ice.2021.180

Inpatients on isolation precautions have less frequent healthcare personnel (HCP) visits, presumably due to the time and effort required to don and doff personnel protective equipment (PPE) to enter the room.1 The “red box” is a 0.28 m2 (3 feet2) space, at the entryway of a patient’s room and facilitates full view and clear communication with the patient.1 Within the red box, HCP remain at a distance of ∼2 m (6 feet) or more from the patient, eliminating the need to use PPE or wash hands while interacting with patients who are on contact or droplet precautions.1 Use of the red-box strategy has been shown to improve isolation practices and to be cost-effective through preservation of PPE.1-5 Prior studies have used direct observers to document compliance with PPE protocols2-5; however, the method has multiple limitations.6,7 We used remote video auditing to compare the rates of HCP compliance with infection prevention protocols in isolation rooms with and without the red box. The following healthcare-associated infection (HAI) rates were monitored as a part of standard infection prevention strategies using National Healthcare Safety Network (NHSN) criteria: MRSA bacteremia, central-line–associated bloodstream infections, catheter-associated urinary tract infections, carbapenem-resistant organisms, and Clostridium difficile. The model did not converge with adjustment. b 95% CI = 95% confidence interval calculated using the Wilson method. c P value based on χ2 test or the Fisher exact test for unadjusted comparisons. d Rate ratio (RR) from binomial regression models, evaluating the effect of the red box vs control on noncompliance (event no.), adjusted for provider type and isolation room type. e Overall glove use gown use and HH total may not add up as certain observations were unviewable. f For doffing and gown use, only unadjusted effects are presented. g The model for doffing only included exit observations. h The model for mask only included observations with droplet or droplet/contact isolation type. [...]this is likely balanced by similar events in control rooms. Because we collected only anonymized data (no patient or HCP identifiers), we could not were unable to adjust for correlation with either patient or HCP.