Coronavirus Pandemic Evaluation of environmental cleaning quality: an observational study at a tertiary hospital in Wuhan, China

Introduction: The COVID-19 pandemic highlights the role of environmental cleaning in controlling infection transmission in hospitals. However, cleaning practice remains inadequate. An important component of effective cleaning is to obtain feedback on actual cleaning practice. This study aimed to evaluate the cleaning process quality from an implementation perspective. Methodology: An observational study was conducted in a tertiary public hospital in Wuhan, China and 92 cleaning processes of units housing patients with multidrug-resistant organism infections were recorded. The bed unit cleaning quality and floor cleaning quality were measured by six and five process indicators respectively. Descriptive statistics were used to describe the cleaning quality. Results: For bed unit cleaning quality, the appropriate rates of cleaning sequence, adherence to cleaning unit principle, use of cloth, use of cloth bucket, separation of clean and contaminated tools, and disinfectant concentration were 35.9%, 71.7%, 89.7%, 11.5%, 65.4%, and 48.7%, respectively. For floor cleaning quality, the appropriate rates of adherence to cleaning unit principle, use of cloth, use of cloth bucket, separation of clean and contaminated tools, and disinfectant concentration were 13.4%, 50.0%, 35.5%, 11.0%, and 36.7%, respectively. Conclusions: The cleaning staff showed poor environmental cleaning quality, especially the floor cleaning quality. The findings can help reveal deficiencies in cleaning practices, raise awareness of these deficiencies, and inform targeted strategies to improve cleaning quality and hospital safety.


Introduction
Healthcare-acquired infection (HAI) is a substantial risk to patient safety, causing morbidity and mortality and increasing hospital stay [1]. The pooled HAI prevalence was 7.6% in high-income countries and 10.1% in the developing world [1]. Along with it, the emergence of antimicrobial-resistant strains gradually becomes a global public health problem. One global report estimated that antimicrobial-resistant strains claimed about 700,000 lives annually. Unless action is taken, the number of deaths could balloon to 10 million lives a year by 2050 [2]. Furthermore, the ongoing COVID-19 pandemic has caused more than 109 million confirmed cases and 2.4 million deaths as of February 19, 2021 [3].
Infection control measures are crucial to controlling the spread of HAI and COVID-19 in hospitals [4,5], among which environmental cleaning is an essential component [5]. Given that some pathogens can survive weeks to months on unclean surfaces, contaminated surfaces can therefore be directly or indirectly involved in the transmission of pathogens [6]. Cleaning can reduce the environmental reservoir of these pathogens and interrupt such transmission [7,8].
However, environmental cleaning remains inadequate, and the suboptimal cleaning quality is prevalent [9][10][11]. Previous studies found that only 49% high-risk surfaces were appropriately cleaned across 23 acute care hospitals [9], and about 35% surfaces in patient bedrooms were cleaned [10]. In addition, 63% of surfaces were culture positive for Clostridium difficile after cleaning; 90% of drawer handles were still contaminated [11]. Furthermore, recent studies found extensive contamination of ward surroundings by SARS-CoV-2 [12][13].
An important component of effective cleaning is to obtain feedback on actual cleaning practice. Cleaning quality is usually measured by methods such as visual inspection, fluorescent marker, adenosine triphosphate (ATP) detection, or microbial culture [14]. Visual inspection was proved to be unreliable and it is limited by the invisible nature of the cleaning outcome [14,15]. Fluorescent marker can only determine whether a surface has been cleaned, but provides no indication on the cleaning efficacy [10]. Microbial culture is timeconsuming and costly and takes a long time to obtain results [14,15]. Fluorescent marker, ATP, and microbial culturecannot fully evaluate cleaning quality because of limited sampling points. Furthermore, these methods cannot assess the potential cross-contamination risks among different patient units caused by cleaning activities. Evaluating the cleaning process quality froman implementation perspectivecan make a comprehensive evaluation of cleaning quality, which may help make up for the above deficiencies. Moreover, process quality is a crucial part of the quality model, comprising three levels of structure, process, and outcome quality [16].
Therefore, this study aimed to evaluate cleaning process quality through observing the whole process of cleaning activities, to help identify and improve deficiencies in cleaning practices and ultimately improve infection prevention and control.

Study design and participants
A cross-sectional observational study was carried out at a tertiary public teaching hospital in Wuhan, Hubei Province, Central China from January to March 2019. The hospital has more than 6,000 inpatient beds and provides over 250,000 inpatient and 300,000 outpatient services every year. The Regulation for Cleaning and Disinfection Management of Environmental Surface in Healthcare [17] and Standard Operating Procedure for Environmental Cleaning (established by this hospital) were implemented in this hospital.
The cleaning staff who worked in inpatient wards were included in this observational study. The cleaning processes of cleaning units housing patients with multidrug-resistant organism infections were observed and recorded. The cleaning unit is the environmental surface adjacent to a patient, including bed, bedside table, and medical instruments. [17]. The observation of each cleaning activity took about 20 min.

Measurement
The cleaning process quality was measured using process indicators: six indicators for bed unit cleaning quality and five indicators for floor cleaning quality (Table 1) [17,18]. About 400-700 mg/L of the chlorinecontaining disinfectant used at the hospital was considered appropriate [17]. Greater than 700 mg/L was

Statistical analysis
Descriptive statistics were used to describe the sample and calculate the frequency and percentage of process indicators for cleaning quality. The analyses were performed by SPSS 19.0 (SPSS Inc., Chicago, IL, USA).

Basic information about the sample
A total of 92 cleaning processes of 25 cleaning staff were recorded in 14 wards of 11 departments. The 11 departments were orthopedics, thoracic surgery, hepatobiliary surgery, intensive care unit, neurosurgery, respiratory medicine, infectious disease, rehabilitation medicine, cardiothoracic surgery, neurology and traditional Chinese medicine. The frequency of observations for each cleaning staff ranged from 1 to 6, with an average of 3.68. Table 2 shows the appropriate rate of process indicators for bed unit and floor cleaning quality. For the disinfectant concentration for bed unit cleaning, 48.7% was appropriate, 15.4% was insufficient (less than 400 mg/L), and 35.9% was excessive (Greater than 700 mg/L). For the disinfectant concentration for floor cleaning, 36.7% was appropriate, 43.3% was insufficient, and 20.0% was excessive.

Discussion
In this study, we evaluated the environmental cleaning quality at a tertiary hospital from an implementation perspective. Overall, the cleaning staff showed poor environmental cleaning quality. For bed unit cleaning quality, the appropriate rates of three indicators (cleaning sequence, use of cloth bucket, and disinfectant concentration) were below 50%. For floor cleaning quality, the appropriate rates of all five indicators were 50% or below. Greater efforts are needed to improve cleaning quality for infection control.
The findings related to poor cleaning quality are consistent with previously reported suboptimal cleaning quality measured by ATP or microbial culture in hospitals [9][10][11]. These findings suggest that hospital environment may frequently become contaminated and couldbe a source of pathogen spread. It is plausible to assume that hospital surface contamination would be severe when COVID-19 broke out. Previous studies found extensive contamination of hospital environment by SARS-CoV-2, and about 30% of surface samples were positive for SARS-CoV-2 in units specialized for confirmed patients in China [12,13]. These potentially highly contaminated surfaces may account for the early cases of healthcare-acquired transmission among healthcare workers and visitors when environmental cleaning protocols were not widely implemented and healthcare workers were not aware of the potential risk of indirect transmission of SARS-CoV-2 [12].
In the present study, the value of evaluating cleaning quality with process indicators was immediately apparent in that it helped identify the deficiencies in cleaning practice. Although regular training was carried out for cleaning staff in the surveyed hospital, they still showed poor cleaning quality. Of note, 39.3% surface samples of COVID-19 patients' surroundings were still contaminated with SARS-CoV-2 in April in China [13], which indicated that training and audit based on the results from ATP and microbial culture may not be adequate to achieve optimal cleaning practice. Good outcome quality depends on the process quality as cleaning protocols and best cleaning practices are useful only if they are actually followed [19]. According to our findings, the suboptimal cleaning quality may be due to inappropriate cleaning sequence, inappropriate use of cleaning tools, inappropriate disinfectant concentration, and mixture of clean and contaminated tools. Therefore, targeted audit, feedback, and training of cleaning staff regarding specific deficiencies in cleaning practice would help improve cleaning quality more effectively and benefit the prevention and control of HAIs and COVID-19 pandemic, ultimately improving hospital safety. This study has some limitations. First, the relatively small sample from only one hospital may limit the generalization of our findings. A multicenter study would be conducted in the future if more data are available. Second, the Hawthorne effect may exist during the observation, which may lead to the overestimation of cleaning quality. Third, the outcome cleaning quality was not measured. In future studies, the evaluation of cleaning quality combined with process indicators and outcome indicators should be taken into account.

Conclusions
The present study explored the level of cleaning process quality and revealed some deficiencies in cleaning practice. Our study supports the need for greater efforts to improve environmental cleaning quality, and more attention should be paid to process indicators of cleaning quality. Data about cleaning process quality can inform evidence-based targeted strategies to improve cleaning practice more effectively. Future intervention research in combination with process and outcome cleaning quality is needed to explore the role of process indicators in improving cleaning quality.