Comparing Visual Inspection, Performance Observation, and Aerobic Colony Counts for Evaluating Hospital Cleanliness

Design and Setting We perform a prospective study which is conducted in a 17-bed adult medical intensive care unit with approximately 352 annually admissions at a tertiary hospital in Taiwan from July 2021 to December 2022. Cleaning performance observation will be performed according to the environmental checklist for monitoring terminal cleaning released by Centers for Disease Control and Prevention in 2014, to select sampling areas of high-touch surfaces in patient rooms, including bed controls, bed rails, cart pulls, light switches, machine controls, tray tables and call buttons.

Data collection The housekeepers start to do terminal cleaning after patients are transferred or discharged, the researcher also starts to observe directly the practices of cleaning at the same time and recorded "compliant" or "not compliant" whenever surfaces are cleaned and disinfected properly or on the contrary. The high-touch surfaces are disinfected with solution which contained Sodium dichloroisocyanurate (NaDCC). Right after the surfaces are cleaned, the researcher inspect the surfaces and record "clean" or "not clean" whenever surfaces are free of dust, debris, and stains immediately after terminal cleaning or on the contrary.

Ten minutes after surfaces are cleaned and disinfected, Hygiena Ultrasnap Surface ATP test will be applied to swab a standard 10*10 cm area on selected surfaces by samping templet. After completing sampling, swabs are replaced back in swab tube and all liquid in bulbs was expelled to activate the reaction. Once activated, samples were read right after shaking swab tube for 5-10 seconds. The ATP readings less than 10 RLU (relative light unit, RLU) are interpreted as "pass".

The designated cleaner and the researcher will be assigned to clean high-touch room surfaces and to be in charge of evaluating and sampling, respectively.

Statistical analysis Descriptive statistics are calculated including numbers, percentage, mean and standard deviation. Inferential statistics are analyzed with SPSS (version 22) to determine correlation between results of performance observation, Visual inspection and ATP interpretation by using Chi-square test or Fisher's exact test. The phi coefficient are calculated to measure strength between two sets of data. The phi coefficient refers to the meaning of Pearson's Correlation coefficients, values 0.01-0.19 as no or negligible, 0.2-0.39 as weak positive, 0.4-0.69 as medium positive, 0.7-1.0 as strong positive.12 The ATP readings were compared between "not compliant" and "compliant", and between "not clean" and "clean" respectively by using Mann-Whitney U test.