The Impact of a Hand Hygiene Education Program in Long-Term Care Facilities in Korea

Abstract

Background

Korea’s rapid population aging has increased demand for long-term care facilities (LTCFs). Because LTCFs are vulnerable to infectious disease outbreaks, consistent hand hygiene (HH) practices among caregivers are essential. We evaluated an LTCF-tailored HH education program with measurement of caregiver compliance.

Materials and Methods

HH compliance rate (HHCR) was defined as the proportion of performed HH actions among the observed WHO five moments. In 2025, 17 LTCFs participated; 13 received HH education. Nine LTCFs also participated in 2024 (three received education in 2024, all nine in 2025). Facility-level effects of 2025 education were assessed using generalized linear mixed-effects models, and caregiver-level HH monitoring was analysed separately.

Results

Education was associated with a greater improvement in HHCR from the first to second half monitoring periods in 2025 (interaction adjusted odds ratio, 8.83; 95% confidence interval [CI], 2.25-34.61; P=0.004). In the two-year enrolled facilities (n=9), 2025 HHCR differed by 2024 education status and monitoring periods (three-way interaction, P=0.008). Within 2024, facilities receiving education in 2024 showed greater HHCR improvement (odds ratio, 58.72; 95% CI, 5.42-635.87; P=0.002). In caregiver-level analyses for second half of 2025, HHCR was higher among educated caregivers than among non-educated (median, 80.0% vs. 12.0%; P=0.002). The weakest HH moment was before touching; education was associated with higher compliance before touching, before a clean procedure, and after body fluid exposure risk.

Conclusion

In the multi-district study of 17 LTCFs, an in-person HH education program was associated with improved caregiver compliance. These findings support expanding and sustaining HH education across Korean LTCFs.

Graphical Abstract

Introduction

Korea is undergoing rapid population aging: as of 2024, about 19.2% of citizens are aged 65 or older, expected to rise above 30% by 2036 and 40% by 2051, making it the fastest-aging country in the world [1,2]. This rapid aging has sharply increased the demand for long-term care services. In 2022, 10.9% of older adults received long-term care services, reflecting growing dependence on institutional support. In Seoul, a capital of total 9 million people (about 18.2% of Korea), residents aged 65 and older are expected to reach 1.9 million by 2025, yet there are only about 1.26 long-term care facilities (LTCFs) per 10,000 older adults [3]. Seoul has 237 LTCFs; on average, each employ about 28 paid care workers and 3 nurses, with a capacity of 62 residents. Approximately one caregiver is responsible for the daily living needs of two residents. This limited workforce creates an imbalance in which few caregivers must care for many old residents.

Moreover, approximately 76.0% of employees in Korean LTCFs are paid care workers, and an increasing proportion are over 60 years old [4]. This aging workforce may have gaps in infection prevention awareness and skills [5]. Unlike hospitals, which undergo regular infection control assessments, LTCFs often lack standardized oversight. Because many residents live in close proximity, LTCFs are especially vulnerable to outbreaks of respiratory infections, foodborne illness, and multidrug-resistant organisms, underscoring the need for consistent hygiene practices in everyday care. Indeed, during the COVID-19 pandemic, more than 95.1% of deaths in Korea occurred among those aged 60 or older, and residents of LTCFs accounted for over 35.1% of these deaths [6]. In countries with stronger healthcare support systems, multifaceted staff hand hygiene (HH) interventions (in-person and video education, posters, ready access to sanitizer) reduced lower respiratory infections from 0.97 to 0.53 per 1,000 resident-days over 22 months [7]. These findings suggest that structured HH education can significantly reduce infection risk in LTCFs.

LTCFs remain structurally vulnerable to infection prevention because institutional systems to continuously monitor paid care workers’ practices or to incentivize sustained improvement are insufficient, as noted in a previous study [8]. Wide variation in facility size and staffing makes it difficult to implement uniform infection control systems. These gaps highlight the urgent need for targeted, sustainable infection prevention education for caregivers. Accordingly, the study evaluates the impact of a HH education program designed specifically to Korean LTCFs, combining in-person training with monitoring of caregivers’ HH compliance.

Materials and methods

1. Study design

This study employed a quasi-experimental controlled before-after design and was conducted among LTCFs that agreed to participate in a HH education and monitoring program. The participating facilities were in three districts of Seoul: Songpa-gu, Gwanak-gu, and Dongjak-gu. In Songpa-gu, all 9 LTCFs in the district participated, with monitoring conducted over a two-year period (2024-2025). In Gwanak-gu, 4 of 9 facilities (44.4%) participated; in Dongjak-gu, 4 of 6 (66.7%) participated; the two districts implemented the program for one year (2025). Participating facilities were initially recruited through the Public Health Centers, and each LTCF voluntarily decided whether to participate. Facility locations within Seoul are shown in Supplementary Figure 1. HH compliance was monitored twice per year-once in the first half and once in the second half of the year. Between these monitoring periods, participating LTCFs that opted into the education received a single in-person training session lasting at least 30 minutes for all on-duty staff; detailed monitoring and education procedures are provided in Supplementary Methods. First-half monitoring was conducted on February 11-25, 2025, and second-half monitoring on September 23-October 31, 2025; among educated facilities, the median interval from education to second monitoring was 110 days. Among the 9 Songpa-gu LTCFs that participated the program for two years, 3 received one additional HH education in 2024. Because not all facilities received the HH education and the same caregivers were not tracked pre- and post-education, HH compliance rates (HHCR) were analysed at the facility level, rather than by individual caregiver.

2. Ethics statement

The study protocol was reviewed and approved by the Institutional Review Board of Seoul Metropolitan Government-Seoul National University Boramae Medical Center (approval no. 07-2025-31). Written informed consent was waived because this was a retrospective design that collected no personally identifiable information from caregivers.

3. Definitions

The primary outcome was the HHCR defined as the proportion of observed opportunities for HH in which the action was actually performed, as described in a previous study [7].

$$\text{HHCR (\%)} = \frac{\text{Number of performed HH actions}}{\text{Total number of HH opportunities}} \times 100$$

The five HH opportunities were defined according to the World Health Organization observation methods, with modifications to fit the context of non-medical LTCFs, as follows: (1) before and (2) after touching residents, (3) after contact with the resident’s surroundings, (4) after exposure risk to body fluids, and (5) before a clean procedure (e.g., wound care such as pressure ulcers, insulin administration, tube-feeding management such as nasogastric flushing, suctioning, and instillation of eye drops) [9]. During each HH opportunity, observers recorded glove use and the HH modality (alcohol-based hand rub or soap-and-water handwashing). HHCR was primarily assessed as whether HH was performed at indicated moments. The duration and technical quality of HH were not systematically measured. However, appropriateness of the method was assessed in selected situations; for example, after body fluid exposure risk, use of soap and water was checked, and glove use without accompanying HH was classified as non-compliance. Because HH compliance in LTCFs was expected to be lower than in hospital settings, an additional indicator was evaluated the “at-least-once compliance rate”, defined as the proportion of caregivers who performed HH at least once during observation among those who had one or more opportunities.

Definitions vary by country, but in Korea, institutional long-term care comprises long-term care hospitals (medical) and LTCFs (non-medical), which include larger nursing homes and smaller group homes [10,11]. In the study, care workers were defined as personnel employed in LTCFs who provide direct caregiving services to residents, including nurses, nurse aides, certified paid care workers, and social workers. Consistent with occupation classifications used in Korean LTCFs, staff whose primary duties involve indirect care or only occasional assistance (e.g., physical or occupational therapists) were excluded [12]. Facility-level variables that could potentially influence outcomes were also collected, including total resident count, total staff count (to indicate facility size), occupation type of caregivers, and the number of caregivers who changed employment between observation periods.

4. Statistical analysis

The main outcome was the change in HHCR associated with HH education. Analyses were conducted at the facility level. For the 2025 quasi-experimental cohort (17 LTCFs), changes in HHCR between the first and second half monitoring were compared by education status using a generalized linear mixed-effects model (GLMM) with a random intercept for facility and fixed effects for period (first vs. second), education (yes vs. no), and their interaction. The interaction term was included in the model to evaluate whether the change in HHCR over the monitoring periods differed according to the education status [13]. For multivariable analysis, significant variables in univariable analyses were initially considered. The final model was selected based on the Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC) to ensure optimal model parsimony [14,15]. Among 9 Songpa-gu LTCFs observed in both 2024 and 2025, 3 received education in 2024 and all did in 2025; HHCR were compared longitudinally by 2024 education status across the two years.

In caregiver-level analyses, categorical variables were analyzed using Fisher’s exact test, and continuous variables were compared using the Mann-Whitney U test. All statistical analyses were performed with a significance level of P<0.050. The analyses were performed, and graphs were created using IBM SPSS Statistics (version 26.0, IBM Corp., Armonk, NY, USA) and a plot of the data was constructed using the R software (version 4.5.2, R Foundation for Statistical Computing, Vienna, Austria, http://www.R-project.org).

Results

1. Characteristics of LTCFs and caregivers

Korean LTCFs operate without on-site physicians, and medical procedures, diagnostic tests, or prescriptions cannot be performed in facilities. The composition of caregivers by facility is shown in Supplementary Table 1. Among the 17 LTCFs, 13 (76.5%) received HH education in 2025: all 9 facilities in Songpa-gu, 2 in Dongjak-gu, and 2 in Gwanak-gu. Facilities had a median of 49 residents and 36 caregivers, with a median paid care workers proportion of 66.7% among caregivers. Among facilities that received HH education, the median caregiver participation rate was 45.0%. Compared with facilities that received 2025-education, facilities without education had fewer residents and caregivers. In LTCFs that received education, a median 15.2% of caregivers changed employment between the first and second half observations.

2. Change in HHCR by LTCF education status

Facility-level effects by education status are shown in Figure 1A. In GLMM, univariable analyses indicated that HHCR tended to be higher in facilities that received 2025-HH education and in those with a larger total staff count. In the final multivariable model selected based on AIC and BIC, HH education was significantly associated with a greater improvement in HHCR between the first and second half monitoring periods (interaction adjusted odds ratio [aOR]=8.83; 95% confidence interval [CI] 2.25-34.61; P=0.004; Table 1). In addition, facilities with a larger total staff count had higher HHCR (aOR=1.01 per staff, 95% CI, 1.00-1.03; P=0.035).

Figure 1. Mean change in facility-level (n=17) hand hygiene (HH) compliance rate. (A) Comparison by 2025-education status (received vs. not received). (B) Songpa-gu facilities (n=9): comparison by 2024-education status; all Songpa-gu facilities received HH education in 2025.

The HH compliance rate was defined as the proportion of observed opportunities in which HH was performed. Asterisks indicate statistically significant differences (P<0.050).

Table 1. Change in hand hygiene (HH) compliance rate by long-term care facility: results from a generalized linear mixed-effects model.

Variables	Univariable		Multivariable
	OR (95% CI)	P-value	aOR (95% CI)	P-value
Monitoring period, second vs. first half of 2025	2.17 (1.30–3.64)	0.006	0.36 (0.10–1.24)	0.098
HH education, yes vs. no	2.79 (1.16–6.76)	0.025	0.62 (0.21–1.82)	0.361
Total staff count, per one-person increase	1.02 (1.01–1.03)	0.005	1.01 (1.00–1.03)	0.035
Monitoring period × HH education	-	-	8.83 (2.25–34.61)	0.004

The outcome is the HH compliance rate; the model includes the listed variables.

OR, odds ratio; CI, confidence interval; aOR, adjusted odds ratio.

Among the 9 Songpa-gu LTCFs monitored over two years, 3 received HH education in 2024, whereas all 9 received education in 2025 (Figure 1B). In the multivariable model, the change in HHCR across monitoring periods differed by 2024 education status (three-way interaction P=0.008; Supplementary Table 2). Post-hoc analysis showed facilities receiving 2024-education had greater HHCR improvement within 2024 (OR=58.72; 95% CI, 5.42-635.87; P=0.002). Additionally, among facilities educated in 2024, HHCR declined in the first half of 2025, suggesting decay of training effects over time.

3. Caregiver-level HHCR

In caregiver-level analyses, there was no significant difference between the two groups in the first half; however, in the second half, HHCR was significantly higher among caregivers who received 2025-education compared with those who did not (median HHCR, 80.0% vs. 12.0%, P=0.002; Table 2). By HH moment, the education was associated with significantly higher compliance in before touching, before clean procedure, and after body fluid exposure risk.

Table 2. Comparison of monitored caregiver hand hygiene (HH) practices by education status.

Variables		First half of 2025				Second half of 2025
		Education received (n=57)	Education not received (n=12)	Total (n=69)	P-valuea	Education received (n=50)	Education not received (n=12)	Total (n=62)	P-valuea
Occupation, n (%)					0.508				0.733
	Nurse	0	0	0		2 (4.0)	1 (8.3)	3 (4.8)
	Nurse aide	2 (3.5)	0	2 (2.9)		1 (2.0)	0	1 (1.6)
	Paid care worker	52 (91.2)	11 (91.7)	63 (91.3)		47 (94.0)	11 (91.7)	58 (93.6)
	Social worker	3 (5.3)	1 (8.3)	4 (5.8)		0	0	0
Glove use, n (%)		19 (33.3)	4 (33.3)	23 (33.3)	>0.999	15 (30.0)	6 (50.0)	21 (33.9)	0.308
HH method, n (%)b					0.044				0.026
	Alcohol-based hand rub	17 (29.8)	1 (8.3)	18 (26.1)		22 (44.0)	1 (8.3)	23 (37.1)
	Soap-and-water handwashing	18 (31.6)	4 (33.3)	22 (31.9)		9 (18.0)	4 (33.3)	13 (21.0)
	Both methods	2 (3.5)	3 (25.0)	5 (7.2)		10 (20.0)	1 (8.3)	11 (17.7)
	No HH performed	20 (35.1)	4 (33.3)	24 (34.8)		9 (18.0)	6 (50.0)	15 (24.2)
Compliance according to HH moment, n/N (%)c
	Before touching a resident	14/42 (33.3)	2/10 (20.0)	16/52 (30.8)	0.705	16/22 (72.7)	0/7 (0.0)	16/29 (55.2)	0.001
	Before a clean procedure	3/5 (60.0)	1/2 (50.0)	4/7 (57.1)	>0.999	18/21 (85.7)	0/2 (0.0)	18/23 (78.3)	0.040
	After body fluid exposure risk	18/30 (60.0)	3/4 (75.0)	21/34 (61.8)	>0.999	19/25 (76.0)	0/3 (0.0)	19/28 (67.9)	0.026
	After touching a resident	18/34 (52.9)	5/11 (45.5)	23/45 (51.1)	0.738	23/35 (65.7)	3/9 (33.3)	26/44 (59.1)	0.128
	After touching resident surroundings	3/15 (20.0)	4/5 (80.0)	7/20 (35.0)	0.031	17/28 (60.7)	4/9 (44.4)	21/37 (56.8)	0.458
HH compliance rate, median (IQR), %d		50.0 (0.0–100.0)	50.0 (15.0–85.0)	50.0 (0.0–100.0)	0.987	80.0 (35.0–100.0)	12.0 (0.0–50.0)	75.0 (25.0–100.0)	0.002
At-least-once compliance, n (%)d		36 (63.2)	9 (75.0)	45 (65.2)	0.521	41 (82.0)	6 (50.0)	47 (75.8)	0.054

Values are presented as number (%), number/total number of observed opportunities (%), or median (interquartile range), as appropriate.

^aCategorical variables were compared using Fisher’s exact test, and continuous variables were compared using the Mann–Whitney U test.

^bHH method was classified according to the method used at least once during observation.

^cValues represent the proportion of caregivers with at least one HH action among those with observed opportunities for each HH moment.

^dHH compliance rate was defined as the proportion of observed HH opportunities in which HH was performed. At-least-once compliance rate was defined as the proportion of caregivers who performed HH at least once among those with one or more observed HH opportunities.

IQR, interquartile range.

Among all 131 monitored caregivers observed in the first and second halves of 2025, paid care workers accounted for 92.4% of observed HH opportunities, and gloves were used in 33.6% of observations. The median HHCR was 50.0%, and the at-least-once compliance rate was 70.2%. Across the five moments, compliance ranked (lowest to highest): before touching a resident, after contact with the surroundings, after touching a resident, after body fluid exposure risk, and before a clean procedure. There were no significant differences between the two monitoring periods for HHCR (P=0.242) or at-least-once compliance (P=0.251).

Discussion

The study provides one of the first multi-district, facility-level evaluations of a structured HH education and monitoring program for caregivers in Korean LTCFs, addressing a critical gap in infection prevention evidence outside hospital settings. In 2025, 17 LTCFs were conducted two rounds of HH monitoring, 131 caregivers were assessed. At baseline, the median HHCR was 50.0%, and the at-least-once compliance was 65.2%, indicating low overall adherence, even though the study did not primarily focus on the technical quality of HH. After the 2025-education, HHCR increased significantly; a similar improvement was observed in the Songpa-gu LTCFs following 2024 HH education. In contrast, repeated monitoring without HH education did not produce a significant HHCR increase, and facilities educated in 2024 showed a decline by the first half of 2025, underscoring the need for continuous, rather than one-off education.

HH is a fundamental, cost-effective strategy to reduce transmission of infectious diseases. Despite limited data from Korean LTCFs, international studies show improved outcomes following staff education. In the randomized controlled trial of 66 Dutch nursing-home units, units receiving an infection control training program had significantly lower incidences of gastroenteritis and influenza-like illness than units without the program [16]. Although resources are limited in Korean LTCFs, where medical procedures are largely unavailable and only basic tasks (e.g., simple wound dressing, eye-drop instillation, hygiene care) permitted, the results of the study indicate that even a single education session can improve HHCR.

In the first half monitoring, baseline HHCR was a median 50.0%. The weakest moment was before touching a resident, a pattern also reported in intensive care units and tertiary care hospitals [17,18]. However, the low absolute rates in this study likely reflect limited awareness of the moment. In facilities that received education, compliance improved notably: before touching from 33.3% to 72.7%, and before clean procedure from 60.0% to 85.7%, and after body fluid exposure risk from 60.0% to 76.0% (Table 2). Prior studies have shown that bundle interventions, including reliable sanitizer supply and visual reminders (poster), can further improve compliance [19]. Taken together, targeted education on the weakest HH moments, coupled with bundle interventions, may further increase HHCR in Korean LTCFs.

Prior studies indicate that infection control practices in Korean LTCFs are largely left to individual caregivers’ discretion, reflecting gaps in mandatory education, systematic monitoring, and weak institutional incentives [8,20]. The current HH education program operates with limited resources: after the program started, only two nurses delivered all in-person trainings and monitoring, which limited both the ability to increase education frequency and to include more LTCFs from additional districts. In the Songpa-gu LTCFs observed over two years, HHCR declined over time after 2024-education, and caregiver turnover reached a median 7.8% within a year, underscoring the need for sustained, repeat training. Given Korea’s rapidly aging population and the growing demand for LTCFs, a coordinated support system to provide professional infection control education and support may be warranted. Future work should identify the effect of increasing education frequency and assess whether improvements in HHCR translate into better resident outcomes, such as lower rates of common infections (e.g., respiratory and skin and soft tissue infections).

This study has a few limitations. First, the program aimed to deliver HH education to as many LTCFs as feasible rather than to conduct a prospective controlled trial; as a result, facilities that received education may differ systematically from those without education in factors such as facility size, staffing, and interest in infection control. In addition, individual-level caregiver characteristics (e.g., age, occupation, prior training) could not be adjusted for in the facility-level analysis. Educated facilities also had larger resident and caregiver counts. Notably, caregiver turnover was higher in educated LTCFs (median 15.2%), a factor that would be expected to lower HHCR; yet HH compliance still improved after education. Although the decrease and subsequent increase in HHCR in the Songpa-gu facilities suggest the influence of time-varying contextual factors not fully captured by the facility-level analysis, caregiver-level findings showed a consistent post-education improvement (Table 2). Because most observed opportunities involved paid care workers, heterogeneity related to occupational composition was unlikely to substantially affect the findings. Second, the relatively low participation rate in the HH education sessions warrants cautious interpretation, as the rotating work schedules of paid care workers may limit the coverage achievable with a single-day session, further supporting the need for repeated education. Third, because participation in education was voluntary, facilities that agreed to take part may have been more interested in infection control, introducing potential selection bias. Finally, because LTCFs are non-medical institutions and resident outcomes such as hospitalizations or infection occurrences are sensitive data, it was unable to assess direct resident-level outcomes.

In conclusion, in this multi-district study of 17 LTCFs, an in-person HH education program was associated with improved caregiver compliance. The study addresses a critical literature gap by providing longitudinal data and showing that training effects weaken when paid caregiver turnover is high, underscoring the need for ongoing, standardized training. Considering the practical challenges of repeatedly delivering education, monitoring, and feedback across multiple LTCFs, a coordinated support system may help sustain HH improvement and support bundle-based infection prevention strategies.

SUPPLEMENTARY MATERIALS

Supplementary Methods

Summary of hand hygiene education program and monitoring procedures

Supplementary Table 1

Characteristics of long-term care facilities participating in the 2025 education

Supplementary Table 2

Change in hand hygiene compliance rate (HHCR) among Songpa-gu long-term care facilities (n=9): generalized linear mixed-effects model according to 2024 hand hygiene (HH) education status

Supplementary Figure 1

Long-term care facilities in Seoul participating in the study.