Universal decolonization for methicillin-resistant Staphylococcus aureus and carbapenem-resistant Acinetobacter in elderly homes: A large cohort of over 16,000 residents in Hong Kong

Abstract

Background

We conducted a decolonization program for methicillin-resistant Staphylococcus aureus (MRSA) and carbapenem-resistant Acinetobacter (CRA) among residential care homes for the elderly (RCHE) residents by providing universal decolonization using nasal povidone-iodine and chlorhexidine baths at the RCHEs and during hospitalization.

Objectives

1. To investigate the effectiveness of decolonization of MRSA and CRA in elderly homes. 2. To assess the incidence of MRSA and CRA infections bacteremia, associated morbidity and mortality in elderly homes. 3. To investigate MRSA infection reduction and risk ratio amongst the elderly residents.

Study design

This quality improvement program adopted a quasi-experimental design.

Methods

A total of 20,741 RCHE residents were instructed to apply 10 % povidone-iodine nasal ointment once daily, Monday to Friday, every alternate week, while 2 % chlorhexidine gluconate solution was used for bathing on alternate days. Group 1 included RCHE residents who underwent decolonization at the RCHEs and during hospitalization, whereas Group 2 received decolonization at the RCHEs only. The residents who stayed in the RCHEs other than the 330 invited residents RCHEs were considered the control group when they admitted to their network hospitals. We compared the incidence of any MRSA and CRA infections, bacteremia, and associated death in 2023 with the rates during the baseline period from 2017 to 2019, using Poisson exact test.

Results

A total of 257/330 (77.9 %) RCHEs continued the program, involving 16,190 residents. The mean utilization of povidone-iodine ranged from 79.4 % to 96.2 %, whereas chlorhexidine utilization ranged from 79.9 % to 97.2 %. MRSA infection was reduced in Group 1 with a risk ratio of 0.878 (95 %CI: 0.776–0.992, p = 0.035) compared with that at baseline. For bacteremia, larger reductions were noted in Group 1 and Group 2, with risk ratios of 0.719 (95 %CI: 0.448–1.115, p = 0.158) and 0.721 (95 %CI: 0.418–1.192, p = 0.207), respectively, compared to the control 0.785 (95 %CI: 0.621–0.984, p = 0.036). Regarding mortality associated with MRSA, a differential reduction of 51.1 %, 18.7 % and 22.2 % were observed for Group 1, Group 2 and the control respectively. For CRA, no statistically significant reduction was detected in the intervention groups for infection, bacteremia or death.

Conclusions

We demonstrated a reduction in hospital admissions due to MRSA infections following the decolonization program for residents in RCHEs and during hospitalization.

1. What this study adds

• ●This is the largest study involving over 16,000 residents, using a quasi-experimental design, to investigate the effectiveness of decolonization of MRSA and CRA in elderly homes, showing fairly good compliance.
• ●Decolonization by provision of 10 % povidone-iodine nasal ointment and 2 % chlorhexidine gluconate solution was effective in reducing hospitalization due to community-onset MRSA infection.
• ●A reduction in MSRA bacteremia and associated mortality was detected, yet the risk ratios were not statistically significant, probably because of the small number of patients with these adverse outcomes.

2. Implications for policy and practice

• ●Decolonization could potentially help tackle Staphylococcus aureus, the top pathogen causing the largest number of attributable deaths due to antimicrobial resistance among adults.
• ●Further consideration of feasibility, staff compliance, residents' acceptance and resource implications is required before full implementation in real-world settings.

3. Introduction

In Hong Kong, the prevalence of methicillin-resistant Staphylococcus aureus (MRSA) colonization in residential care homes for the elderly (RCHE) was only 2.8 % in 2005 [1], which sharply increased to 21.6 % in 2011 [2]. The prevalence of MRSA colonization continued to rise to 30.1 % in 2015 and then to 37.9 % in 2017, as shown by different local studies in RCHEs [3,4]. In 2015, the prevalence of carbapenem-resistant Acinetobacter (CRA) was estimated to be 6.5 % with a high background rate of MRSA (32.2 %) [5]. Infection with MRSA and other multidrug-resistant organisms (MDROs) can lead to prolonged illness and high mortality. For example, a review of 1133 MRSA bloodstream infections in 26 public hospitals in Hong Kong between 2010 and 2012 reported 367 deaths, resulting in a 30-day mortality rate of 32.4 % [6]. It is common for elderly residents in RCHEs to be admitted to public hospitals due to pre-existing chronic illness. Patients with multiple risk factors for MDROs are more prone to acquire MDROs during their hospital stays. When these residents return to RCHEs, they may further spread MDROs to others in the same institution due to crowded environments, poor personal hygiene, or suboptimal infection control practices by the staff. This could create a vicious cycle of MDRO transmission between public hospitals and RCHEs.

Control of MDRO in hospitals and RCHEs mainly rest on the transmission-based precaution including proper hand hygiene, wearing gown and glove when conducting personal care to patients, use of dedicated medical equipment, assigning dedicated toilets or bathing facilities, cleansing and disinfection of the environment. Isolation and cohorting of MDRO carriers may be difficult given the high prevalence of the MDRO in these settings with high occupancies. Antimicrobial stewardship programme has been in place in public hospitals to minimize occurrence of resistant bacteria. Research has shown that topical antiseptic agents, such as mupirocin nasal ointment and chlorhexidine gluconate (CHG) bathing, are effective in decreasing MRSA colonization or clinical infection in hospital settings [[7], [8], [9], [10]]. A systematic review indicated that CHG bathing reduces the risk of MRSA colonization and bacteremia in the intensive care unit (ICU) by 41 % and 36 %, respectively [10]. However, there is limited evidence regarding the decolonization of MDROs in RCHEs or long-term care facilities (LTCF). In a randomized controlled trial involving 2021 participants in the community (including both LTCF residents and non-LTCF residents), MRSA decolonization with nasal mupirocin and CHG for 6 months led to a 30 % lower risk of MRSA infection compared to education alone [7]. Another trial found that universal decolonization for MRSA was more effective than targeted decolonization in the intensive care unit (ICU) setting [11]. Moreover, CHG bathing has also been demonstrated to reduce colonization of multidrug-resistant Acinetobacter in the ICU [12]. It is worth exploring whether these measures are effective in other settings, particularly those institutions with a high prevalence of MDROs. Therefore, we conducted a decolonization program among RCHE residents in Hong Kong with the aim of reducing MRSA and CRA acquisition and carriage by providing universal decolonization using nasal application of povidone-iodine and CHG baths in a real-world setting. The objectives of the present study include 1) to investigate the effectiveness of decolonization of MRSA and CRA in elderly homes; 2) to assess the incidence of MRSA and CRA infections bacteremia, associated morbidity and mortality in elderly homes; and 3) to investigate MRSA infection reduction and risk ratio amongst the elderly residents.

4. Methods

4.1. Recruitment of RCHEs and provision of decolonization

This quality improvement program adopted a quasi-experimental design for the decolonization of MRSA and CRA, targeting residents in RCHEs and hospitalized patients. The program was conducted under the supervision of the Expert Committee on Antimicrobial Resistance, which comprises a panel of clinical microbiologists, hospital infection control experts, and public health physicians, in collaboration with the Centre for Health Protection, Department of Health of the Government of Hong Kong Special Administrative Region (HKSAR). Invitations were sent to 330 RCHEs in the catchment areas of three major hospital networks in Hong Kong (Networks I, II and III). The invitation letter outlined the situation regarding MDROs, the associated disease burden, and the potential benefits of decolonization. Information sheets were also prepared for RCHE residents and their relatives to enhance their understanding of the program and promote compliance. Universal decolonization was offered to all RCHE residents on a voluntary basis. Verbal consent of the residents, family members or relatives were obtained before starting any decolonization. As a public health and quality improvement initiative, this decolonization program was exempt from human participant research oversight by the Expert Committee on antimicrobial resistance and was continuously monitored by the Centre for Health Protection, Department of Health of the HKSAR.

The decolonization intervention included the application of 10 % povidone-iodine nasal ointment and the use of 2 % CHG bath solution, based on reference from the previous study [11]. Residents were instructed to apply the povidone-iodine ointment to their nostrils twice daily once a week. Additionally, they were advised to use the CHG solution for bathing, including showers and hair washing, on alternate days or according to the RCHE's prevailing schedule. When these residents from Group 1 were admitted to Hospitals A to D (Supplementary Table 1), hospital staff also administered decolonization during hospitalization. In contrast, residents in Hospitals D and E (Group 2) did not receive decolonization during their hospital stay. This design allowed for a comparison of the effectiveness of decolonization between residents receiving decolonization in both RCHEs and hospitals (Group 1), with those receiving decolonization only in RCHEs (Group 2). The residents who stayed in the RCHEs other than the 330 invited RCHEs were considered the control group when they admitted to their network hospitals.

4.2. Training and compliance monitoring

To ensure proper implementation of the decolonization program, training materials were developed to educate RCHE staff on the appropriate techniques for applying povidone-iodine nasal ointment and using CHG solution. Special attention was given to the correct amount of CHG applied and thorough cleaning of hidden areas, such as the axilla, groin, and areas near any tubes in-situ. Online briefing sessions were conducted to explain the importance and details of the program. On-site visits were made to all participating RCHEs to enforce application techniques for decolonization products and infection control measures, including hand hygiene and environmental disinfection. Regular audit visits, conducted approximately every three months, monitored compliance by reviewing dispensing records for povidone-iodine, checking the stock of CHG bath solutions, and using test strips to confirm the application of povidone-iodine to residents’ nostrils. These visits also focused on reinforcing infection control measures, identifying any practical difficulties and resolving them. During audits, 5 % of residents from each RCHE, or a maximum of 10 residents, were selected for the iodine strip test. Reasons for refusal to participate in the program as recorded by the RCHE staff were also reviewed.

A pilot involving 23 RCHEs in Healthcare Networks I and II commenced in September 2021. The regimen was adjusted from 4 % to 2 % CHG due to dry skin experienced by participating RCHE residents. The program gradually rolled out to the remaining 127 RCHEs in Networks I and II from January 2022, and to 180 RCHEs in Network III in June 2022. It is important to note that the program faced significant disruptions during the COVID-19 pandemic in February and March 2022, especially during the Omicron variant surge, which led to isolation and quarantine measures for both RCHE residents and staff. To accommodate staff logistics and capacity, the application of povidone-iodine was switched to once daily, Monday to Friday, every alternate week since March 2023. Following a gradual return to normal operations, the program was fully implemented in 2023.

4.3. Microbiological tests in hospitals and evaluation of the program

The program was evaluated by measuring the positive rates of the clinical samples among participating residents admitted to hospitals (Hospitals A to D) for MRSA and CRA cultures were performed as previously described [13]. The bacterial species were confirmed by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS; Bruker Daltonics, Bremen, Germany). The susceptibility of the isolates was determined using the Kirby–Bauer disc diffusion method or E test, and the results were interpreted according to the guidelines of the Clinical and Laboratory Standards Institute. Information on infection, bacteremia and death associated with MRSA and CRA was retrieved from the Clinical Data Analysis and Reporting System of the Hospital Authority. RCHE residents were considered to have community-onset MRSA and CRA infections if clinical specimens taken within 48 h of admission were positive. Bacteremia due to MRSA or CRA was defined as positive isolation of MRSA or CRA from blood samples during the hospitalization episode. Mortality associated with MRSA or CRA was defined as death within the same hospitalization period where MRSA or CRA were detected in blood samples, indicating the presence of MRSA or CRA bacteremia.

4.4. Statistical analysis

The characteristics of the RCHEs under different healthcare networks, including the type and size of the institution, were compared with those not participating in the program, using chi-square tests or analysis of variance. We calculated means and 95 % confidence intervals for the utilization of povidone-iodine nasal ointment and CHG among RCHEs in each cluster during the participation period. To reflect compliance with the usage of povidone-iodine, we reported the number and percentage of residents who passed the iodine test under each healthcare network. We determined the incidence of MRSA and CRA infections, bacteremia and associated death among residents admitted to hospitals using resident-days as the denominator. Resident-days are estimated based on resident count data from the Social Welfare Department and their follow-up periods. For infection, bacteremia and associated death due to MRSA and CRA, the incidence in 2023 was compared with that in the baseline period from 2017 to 2019. The incidences from 2020 to 2022 were not used for analysis due to the disruption of COVID-19. The incidences in the two periods were compared using the Poisson exact test, and statistical significance was considered if the two-sided p-value was <0.05, using SPSS version 27, Chicago.

5. Results

5.1. RCHE participation in the program

Supplementary Table 2 shows the participation of the RCHEs. Among the 330 RCHEs, 9 (2.7 %) refused to join. Ten (3.0 %) RCHEs closed during the implementation period and 54 (16.4 %) RCHEs withdrew mostly because of a lack of manpower to support the program. At the end of 2023, 257/330 (77.9 %) RCHEs continued the program. In the 257 RCHEs, 16,190 residents (78.1 %) out of the 20,741 residents participated the program. Table 1 illustrates the characteristics of the RCHEs across the three participating healthcare networks and the non-participated healthcare networks. For RCHEs across the three participating healthcare networks, approximately 75.0 %–80.0 % of the RCHEs were private homes, with a mean capacity of 95.1–120.8 residents while occupancy ranged from 70.3 to 83.2 residents. There was no statistical difference in the proportion of RCHEs classified as private homes between the three participating healthcare networks and the non-participating healthcare networks. The RCHEs under the three participating healthcare networks were slightly larger in size than those under the non-participating healthcare networks (mean capacity: Network I: 111.2, Network II: 120.8, Network III: 95.1 vs non-participating healthcare networks: 90.2). The RCHEs across Network I and II had a lower occupancy (Network I: 70.3, 95 % C.I.: 66.8–73.9; Network II: 74.0, 95 % C.I.: 70.5–77.5) than the homes across non-participated healthcare networks (80.0, 95 % C.I.: 78.2–81.8), with a p-value of less than 0.05.

Table 1.

Characteristics of Residential Care Homes for the Elderly (RCHE) participated in the decolonization program.

	RCHEs under hospital cluster				p-value
	Network I	Network II	Network III	Non-participating RCHEs
No. of RCHE	60	90	180	468	–
Proportion of RCHE classified as private homesa	78.3 %	80.0 %	75.0 %	74.8 %	0.71b
RCHE capacity (Mean (95 % CI))	111.2 (95.2–127.2)	120.8 (99.2–142.3)	95.1 (86.8–103.3)	90.2 (85.0–95.4)	<0.005c
RCHE occupancy (Mean (95 % CI))	70.3 (66.8–73.9)	74.0 (70.5–77.5)	83.2 (80.3–86.1)	80.0 (78.2–81.8)	<0.005c

^aRCHE in Hong Kong are categorised based on the mode of subsidy: subvented homes, contract homes, non-profit self-financing homes, and private homes. Private homes are residential care homes operated by private organizations.

^bChi-square test.

^cANOVA.

5.2. Compliance monitoring

The compliance regarding the use of povidone-iodine and CHG in RCHEs under the three healthcare networks shown in Fig. 1a, Fig. 1b, Fig. 1ca–c. During the participating period, the mean utilization of povidone-iodine ranged from 81.9 % to 95.8 % for the RCHEs under Network I, from 79.4 % to 96.2 % for the RCHEs under Network II, and from 82.9 % to 86.8 % for the RCHEs under Network III. Similarly, the mean utilization of CHG ranged from 82.2 % to 97.2 % for RCHEs under Network I, 87.6 %–95.1 % for RCHEs under Network II, and 79.9 %–87.9 % for RCHEs under Network III. The results of the iodine test for cross-checking compliance with the use of povidone-iodine are shown in Table 2. The positive rates for iodine testing were very similar among the three healthcare networks, ranging from 74.0 % to 75.5 %. Refusal reasons for joining the program mainly included the unpleasant smell of povidone-iodine and skin reactions to CHG, such as dry skin or rash after its application (Table 3).

Fig. 1a.

Utilization rate of povidone-iodine and chlorhexidine among Residential Care Homes for the Elderly residents (Network I hospital).

Fig. 1b.

Utilization rate of povidone-iodine and chlorhexidine among Residential Care Homes for the Elderly residents (Network II hospital).

Fig. 1c.

Utilization rate of povidone-iodine and chlorhexidine among Residential Care Homes for the Elderly residents (Network III hospitals).

Table 2.

Iodine test for compliance with povidone-iodine application.

Residential Care Homes for the Elderly (RCHE) under healthcare network	No. of RCHE (No. of residents on povidone-iodine)	No. of residents passed iodine test (%)a
Network I	49 (2681)	216/286 (75.5 %)
Network II	64 (4168)	287/388 (74.0 %)
Network III	143 (9000)	675/907 (74.4 %)

^aA total of 5 % of residents from each RCHE (maximum 10 residents) were randomly selected for iodine test.

Table 3.

Reasons for refusal of povidone-iodine and chlorhexidine application.

	Reason for refusal	No. of residents refused	%
Povidone-iodine (n = 3497)	Dislike the povidone-iodine smell	1297	37.1 %
	Dementia or psychiatric issues	381	10.9 %
	Refused by self/family members	263	7.5 %
	Doubt the effectiveness	221	6.3 %
	Concerned about the side effect	137	3.9 %
	Other reasons	1198	34.3 %
Chlorhexidine (n = 4551)	Skin problems	2390	52.5 %
	Insisted to use own soap	986	21.7 %
	Doubt the effectiveness	514	11.3 %
	Concerned about the side effect	301	6.6 %
	Refused by self/family members	120	2.6 %
	Dementia or psychiatric issues	42	0.9 %
	Other reasons	198	4.4 %

Reasons reported by Residential Care Homes for the Elderly residents for refusing decolonization during the last audit visit were included in the analysis.

5.3. Outcome assessment

Table 4 presents the results of the outcome analysis on infection, bacteremia and associated deaths due to MRSA and CRA. For MRSA, when compared with the baseline period (2017–2019), the intervention period (2023) showed a significant reduction in infections for Group 1 hospitals with a risk ratio of 0.878 (95 % CI: 0.776–0.992, p = 0.035) or a reduction by 12.2 %. A reduction in the MRSA infection rate was also observed for both Group 2 and the control group, but the difference was not statistically significant. For bacteremia, we detected a greater reduction for RCHEs in both Group 1 and Group 2 than in the control, with risk ratios of 0.719 (95 % CI: 0.448–1.115, p = 0.158) and 0.721 (95 % CI: 0.418–1.192, p = 0.207) respectively, although neither were statistically significant. For mortality associated with MRSA, non-significant reductions of 51.1 %, 18.7 % and 22.2 % were observed in Group 1, Group 2 and the control, respectively. For CRA, there was a significant increase in CRA infection in Group 1 and a decrease in CRA infection in the control group, but no statistically significant difference was detected for bacteremia or death (Table 4).

Table 4.

Additional Analysis of Methicillin-resistant Staphylococcus aureus (MRSA) and carbapenem-resistant Acinetobacter (CRA) infection, bacteremia and mortality.

Pathogen	Outcome	Intervention Group/Control	Baseline period (2017–2019)			Intervention period (2023)			Rate ratio		p-value	Percentage change in incidence
			Case count	Resident-days	Rate	Case count	Resident-days	Rate	Estimate	95 % CI
MRSA	MRSA Infection	(Group 1)	1106	17,011,920	0.065	348	6,094,040	0.057	0.878	(0.776–0.992)	0.035	−12.2 %∗
		(Group 2)	708	16,525,375	0.043	221	5,950,595	0.037	0.867	(0.742–1.010)	0.068	−13.3 %
		Control	3472	38,430,485	0.090	1205	14,172,220	0.085	0.941	(0.881–1.005)	0.070	−5.9 %
	MRSA Bacteremia	(Group 1)	101	17,011,920	0.006	26	6,094,040	0.004	0.719	(0.448–1.115)	0.158	−28.1 %
		(Group 2)	77	16,525,375	0.005	20	5,950,595	0.003	0.721	(0.418–1.192)	0.207	−27.9 %
		Control	342	38,430,485	0.009	99	14,172,220	0.007	0.785	(0.621–0.984)	0.036	−21.5 %∗
	Death with MRSA Bacteremia	(Group 1)	40	17,011,920	0.002	7	6,094,040	0.001	0.489	(0.185–1.102)	0.096	−51.1 %
		(Group 2)	41	16,525,375	0.002	12	5,950,595	0.002	0.813	(0.389–1.576)	0.641	−18.7 %
		Control	122	38,430,485	0.003	35	14,172,220	0.002	0.778	(0.518–1.141)	0.208	−22.2 %
CRA	CRA Infection	(Group 1)	374	17,011,920	0.022	353	6,094,040	0.058	2.635	(2.272–3.056)	<0.005	163.5 %∗
		(Group 2)	148	16,525,375	0.009	39	5,950,595	0.007	0.732	(0.501–1.048)	0.082	−26.8 %
		Control	691	38,430,485	0.018	490	14,172,220	0.035	1.923	(1.709–2.162)	<0.005	92.3 %∗
	CRA Bacteremia	(Group 1)	5	17,011,920	0.000	0	6,094,040	0.000	0.000	(0.000–3.046)	0.335	−100.0 %
		(Group 2)	3	16,525,375	0.000	1	5,950,595	0.000	0.926	(0.018–11.529)	1.000	−7.4 %
		Control	12	38,430,485	0.000	6	14,172,220	0.000	1.356	(0.418–3.903)	0.596	35.6 %
	Death with CRA Bacteremia	(Group 1)	0	17,011,920	0.000	0	6,094,040	0.000	N/A	(0.000 - Inf)	1.000	N/A
		(Group 2)	0	16,525,375	0.000	0	5,950,595	0.000	N/A	(0.000 - Inf)	1.000	N/A
		Control	1	38,430,485	0.000	1	14,172,220	0.000	2.712	(0.035–212.858)	0.466	171.2 %

N/A: not available.

∗p < 0.05.

6. Discussion

To the best of our knowledge, this is the largest study involving over 16,000 residents, using a quasi-experimental design, to investigate the effectiveness of decolonization of MRSA and CRA in elderly homes. We have shown that decolonization by provision of povidone-iodine and CHG was effective in reducing hospitalization due to community-onset MRSA infection. We also detected a reduction in MSRA bacteremia and associated mortality, yet the risk ratios were not statistically significant, probably because of the small number of patients with these adverse outcomes. The intervention strategy could potentially help tackle Staphylococcus aureus, the top pathogen causing the largest number of attributable deaths among adults according to a recent global burden study on AMR [14].

Most of the studies examining the effectiveness of decolonization for MRSA were conducted in hospital settings particularly in the ICU [[15], [16], [17], [18]], and there is limited literature reviewing the effectiveness in elderly home settings [7,19,20]. In our program, we demonstrated a 12.2 % reduction in infection due to MRSA in intervention Group 1, where residents received decolonization in both RCHEs and during hospitalization. In a similar quality improvement program involving 35 health care facilities in California, CHG bathing and nasal decolonization using povidone-iodine were associated with significantly lower MDRO prevalence, incident clinical cultures, infection-related hospitalizations and deaths among nursing home residents [20]. In another study by Miller involving 28 nursing homes where residents were assigned to the intervention group, universal decolonization with nasal povidone-iodine and CHG bathing led to a significantly lower risk of transfer to a hospital due to infection than routine care, with a difference in the risk ratio of 16.6 % (95 % CI, 11.0 to 21.8), similar to our findings of 12.2 % [19]. The less frequent use of povidone-iodine (once daily for 5 days every other week) in the current study compared with twice daily for five days every other week in Miller's study may explain the lower percentage reduction of MRSA.

On the other hand, a reduction in infection, bacteremia and associated death was also observed in the control groups in 2023 compared with the baseline period. This may be due to tighter infection control measures after the resumption of normalcy following COVID-19. A local study revealed that, compared with pre-COVID-19 years, there was a significant reduction in the incidence of various infectious diseases in the community, including scarlet fever, pneumococcal infections, chickenpox, and pertussis, during the pandemic period, which had a sustained effect even after the resumption of normalcy in 2023 [21]. However, the magnitude of reduction was consistently greater in Group 1 than in the control group for all three outcomes (Table 4), supporting the genuine effect of decolonization for MRSA rather than any confounding effect observed in the control group. The observed effect was unlikely due to variation of infection control practices which should be the standard practice across all public hospitals in Hong Kong.

Povidone-iodine is considered an alternative decolonization agent because of its broad antimicrobial spectrum with activity against Gram-positive and Gram-negative bacteria, including antibiotic-resistant and antiseptic-resistant strains [22,23]. Clinical trials have shown evidence for the prevention of surgical site infection during preoperative decolonization with intranasal use [[24], [25], [26], [27], [28]]. Compared with CHG, povidone-iodine has a broader spectrum of antimicrobial activity, with rapid and potent activity against MRSA found in chronic wounds, including biofilms [29]. In addition, unlike CHG and mupirocin, available reports have not observed a link between povidone-iodine and the induction of bacterial resistance, suggesting potential benefits in long-term use [22,30]. In our program, we did not show decolonization using povidone-iodine and CHG was effective against CRA infection. This was probably because the contaminated environment plays a more significant role in the disease transmission of Acinetobacter baumannii than in that of MRSA. In a hospital setting, researchers have reported that both the bed linen contamination rate and the acquisition rate of multidrug-resistant Acinetobacter baumannii are higher than those of MRSA [31].

There are several limitations to the current program. First, the study adopted a quasi-experimental design, not a randomized controlled trial. The characteristics of RCHE residents might be different in Group 1, Group 2 and the control, as reflected by the different infection rates at baseline. However, a historical reference has been used for each group to compare the infection, bacteremia and mortality rates, which should partially address any confounding effects, such as better infection control during the COVID-19 pandemic, as explained earlier. Second, compliance with the intervention was crucial to ensure effective decolonization. Non-compliance with the current program might still exist in a real-world setting, as reflected by the results of the iodine tests. Regular audit visits were conducted to ensure adherence to the program, with checking on dispensing records of povidone-iodine and stock of CHG bath solutions.

We have shown a reduction in hospital admissions due to MRSA after the provision of decolonization with povidone-iodine nasal ointment and CHG bath for residents in RCHEs. These findings suggest that a decolonization program using a regional approach for providing antiseptics in both RCHEs and during hospitalization might help alleviate the hospital burden of MRSA. Further consideration of feasibility, staff compliance, residents’ acceptance and resource implications is required before full implementation in real-world settings.

6.1. Ethics statement

The study was performed in accordance with the ethical standards of the in accordance with the ethical standards of the Helsinki Declaration, which was accepted by the World Health Community in 1975 (revised in 2013). Standard protocol of the programme was drafted and written information about the decolonization was provided to the residents, family members/relatives and staff of RCHE. Verbal consent of the residents, family members/relatives were obtained before starting any decolonization. The progress and outcomes of the programme monitored regularly by the Expert Committee on antimicrobial resistance and the Centre for Health Protection, Department of Health, HKSAR, and the program was exempt from human participant research oversight by the Expert Committee on antimicrobial resistance.

Availability of data and materials

The datasets used and analysed during the current study are available from the corresponding author on reasonable request.

Authors' contributions

ESKM: Conceptualization, Data curation, Formal analysis, Project administration, Writing - original draft, Writing - review and editing. WSC: Conceptualization, Project administration, Writing - review and editing. VCCC: Conceptualization, Project administration, Writing - review and editing. DCL: Project administration. KL: Project administration. RWML: Project administration. VWMC: Project administration. EH: Formal analysis. VC: Formal analysis. AL: Project administration. CH: Project administration. JYCL: Project administration. ELKT: Project administration. KKY: Project administration, Supervision, Writing - review and editing. All authors have read and agreed to the published version of the manuscript.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following is the Supplementary data to this article: