Study design and setting

This was a quasi-experimental before-and-after cohort study based on a cohort of all residents admitted to an NH in South Korea. The study site was purposively selected considering the feasibility and absence of any infection control intervention. The facility had a 130-bed capacity and approximately 70–80 workers. The NH did not have infection control guidelines, and no IPC program or professionals were dedicated to IPC activities before intervention. The institutional review board reviewed and approved this study (No. 1711/003–015, 1911/001–008). The intervention described below was implemented for 12 months, from January 1, 2018 to December 31, 2018 while data was collected between December 2017 and February 2020, after written informed consent was obtained from the workers and residents. Due to the health status of LTCF residents, individuals who had difficulty exercising autonomy required consent from proxies such as family members.

Intervention

The multi-component IPC program was based on the IPC manual of the WHO [6] comprising evidence-based guidelines; education and training; surveillance; multi-modal strategies; monitoring and feedback; workload and staffing; and built environment, materials, and equipment (Fig. 1). The specific activities for each component are as follows:

Fig. 1

The timeline of this study. IPC = infection prevention and control

IPC guidelines

Based on the guideline development process of the Scottish Intercollegiate Guidelines Network [14], the guidelines were developed [15]. During the development, the infection control guidelines for LTCF provided by the Society for Healthcare Epidemiology of America (SHEA)/ Association for Professionals in Infection Control and Epidemiology (APIC) guideline for infection control in LTCF [16] were used as a foundation, along with guidelines and literature on infection control [17,18,19]. Printed guidelines were provided to the NH. The IPC practice algorithm [15], structuring the infection control procedures, was posted in units, including nursing stations and corridors, to facilitate application in daily practice.

Education and training

Education specific to the IPC was implemented for all employees. Nursing staff, including nurses and certified caregivers, participated in 6 h of education delivered over 3 weeks, with one 2-hour session each week. The educational content included the characteristics of HAIs in LTCFs, standard precautions such as hand hygiene, transmission-based precautions, IPC guidelines, algorithms, nursing and assessment for infection prevention and control, occupational infection prevention and control (including immunization, medical evaluation, and management of exposure), visitor restrictions, environmental infection control, management of outbreaks, surveillance, and communication regarding infection-related information between healthcare facilities. An additional two-hour training session were provided after six months.

Surveillance

Facility-based surveillance was performed by an external infection control nurse once weekly to detect HAI occurrences and outbreaks. The nurse had approximately 5 years of experience working as an infection control nurse at a tertiary hospital and played a role as part of the intervention in this study, but was not involved in data collection, analysis, or evaluation of the effectiveness of this intervention. The infection control nurse monitored residents for symptoms and signs of HAIs.

Multi-modal strategies

Five elements of the WHO multi-modal strategy were applied, namely, system change, training and education, monitoring and feedback, reminders, and a culture of safety.

System change: Before intervention, alcohol-based hand sanitizers were only placed at central stations of each unit, limiting access to hand sanitizers at the point of care. To address this issue, alcohol-based hand sanitizers, including pocket-sized hand rubs and paper towels, were supplied during the intervention. An educational kit for HH, such as a UV lamp for effective hand-washing demonstrations, was placed in the units.

Training and education: In addition to the aforementioned education sessions for staff, the infection control nurse provided immediate on-site education and training to improve the staff’s infection control practices for 2 h per week.

Monitoring and feedback: The infection control nurse monitored HAI occurrence and infection control practice by conducting direct observations to assess hand hygiene, proper glove use, environmental management, and compliance with IPC guidelines. Timley feedback was provided to staff on-site based on the results of these observations and the results of HH compliance were analyzed monthly and posted in units.

Reminder and communication: Posters and checklists related to IPC activities were posted in each unit; the infection control nurse reminded the staff of proper IPC practices and infection control guidelines. The posters featured WHO’s hand hygiene indications, examples of nursing practices for each indication, proper techniques of hand hygiene, respiratory etiquette, visitor restrictions and the IPC practical algorithm.

Culture of safety: An HH campaign was held for all employees at the NH in February 2018 to increase awareness of the importance of HH. Heads of nurses and certified caregivers were encouraged to participate in the intervention and lead their members actively. Employees with excellent IPC practices were selected monthly as role models, the list was posted in units, and incentives were awarded.

Monitoring and feedback

The IPC practices of the NH were monitored weekly, and the infection control nurse provided timely feedback.

Workload and staffing

The NH did not have dedicated infection control personnel before the intervention. Thus, we assigned the infection control nurse, who was hired by the research project, to the facility as a facilitator for implementing the IPC program. The infection control nurse visited the NH once weekly.

Built environment, materials, and equipment

Products for HH, such as pocket-sized alcohol sanitizers and antimicrobial soaps containing chlorhexidine gluconate, were readily available at the point-of-care. Cleaning checklists were developed and used for environmental infection control on resident rooms, common areas, and bathrooms. These checklists offered guidelines for cleaning practices, detailing which environmental surfaces required cleaning, as well as specifying proper disinfectant use and verification of expiration dates.

Measurements

The effect of the intervention was evaluated using resident- and staff-based outcomes. Resident-based outcomes were evaluated by the incidence of respiratory tract, urinary tract, GI, and SST infections—the most common infections in LTCFs [15, 20, 21]—defined by McGeer’s revised criteria [22], which were used for surveillance purposes. The study periods for the resident outcome evaluation were divided into three phases: ‘pre-implementation phase’ preceding the intervention, ‘implementation phase’ as the first evaluation of the intervention’s effect (with surveillance conducted at 6 and 12 months after intervention began), and ‘sustainability phase’ as the evaluation of long-term effects (with surveillance conducted at 3, 6, and 12 months after intervention was finished). The incidence of HAIs was calculated based on the number of new cases occurring over a one-month at each surveillance point, and the infection risk was analysed accordingly. A case that satisfied the criteria was classified as a definite case, while those that did not meet the criteria but in which the resident presented with symptoms of a possible infection were classified as probable cases. Infections included (1) definite and probable cases, and those (1) with no evidence of infection at the time of admission, (2) attributed to the NH, and (3) in residents on more than two calendar days after admission. The authors evaluated resident-based outcomes (MHL, YMY, and EYN). Initially, two authors independently carried out the surveillance, and then the results were reviewed and discussed with the third author to determine whether each case met the criteria.

Staff-based outcomes included the self-reported level of knowledge and compliance with infection control practices measured before and after the intervention. Knowledge of infection control practices was measured using a tool modified by Baek [23] according to the revised guidelines of the Hospital Infection Control Practices Advisory Committee [24], based on the tool designed by Suh and Oh [25]. This tool consists of a total of 29 items assessing standard precaution, HH, personal protective equipment, respiratory etiquette, placement of patient, environmental management, and sharp injury prevention. Each item was measured on a dichotomous scale: 1 point for correct answer and 0 points for wrong answer. A higher score indicated greater infection control knowledge. The reliability of the tool in this study was assessed based on the Kuder–Richardson 20 (KR-20) = 0.71. The performance of infection control practices was measured using a tool of Park and Lim [26], based on the infection control guidelines for elderly care facilities developed by Kim and Chun [27]. This tool comprised seven categories: HH, personal hygiene, disinfection, medication management, urinary tract infection management, respiratory infection management, and environmental management, and included 35 items rated on a 4-point Likert scale (1-never to 4-always). A higher score indicated a higher level of infection control performance. The reliability of this tool was Cronbach’s alpha = 0.96 in the developer’s study [26] and 0.96. in this study.

Statistical analysis

Considering the characteristics of LTCFs, older adults who resided for ≥ 14 days were included in the final analysis. Data analysis was conducted using R 4.3.1 (R Foundation for Statistical Computing, Vienna, Austria; http://www.r-project.org/). For comparisons between groups (time points) in Tables 1, 3, and 4, one-way analysis of variance (ANOVA), independent sample t-tests, and chi-squared tests were performed. A Poisson regression model was used to predict the risk of infection based on the number of infections across the three phases of the study periods. However, a substantial proportion of the data had zero counts, making applying a standard Poisson model inappropriate. Therefore, the number of infections was modeled using a zero-inflated Poisson model with the ‘pscl’ package for analysis in Table 2. When conducting the analysis using the zero-inflated Poisson model, we adjusted for sex and age as covariates. We treated the number of resident days for each phase of study periods as an offset, representing at-risk days. All tests were conducted with a significance threshold set at p < 0.05.

A power analysis was performed using the G*Power 3.1.9 program [28] based on Poisson regression, with a focus on the odds ratio (OR) of resident outcomes. The analysis indicated that all resident outcomes achieved a power of 0.9 or higher, suggesting an adequate sample size, except for short-term total infection (OR = 1.06).