Table 1

Evidence table for the included original research studies

Author, year Study design	Setting Sample size	Intervention	Outcome
Amirov et al, 2017 Controlled before-and-after18	Tertiary care hospital geriatric complex continuing care unit n=122	Chlorhexidine bathing 6 days/week	Over 12 months, there was one new case of MRSA acquisition in the intervention group and seven new cases in the control group, this difference was not statistically significant.
Bessesen et al, 2013 Non-randomised head-to-head19	Two acute care hospitals (one per arm) n=193 300 patient-days across both sites	Contact isolation (gloves+gown for all room entry) vs upgraded standard precautions for patients infected/colonised with MRSA	No difference in incidence density of MRSA acquisition (1.58 vs 1.56 per 1000 patient-days, p=0.98) or MRSA hospital-acquired infection (0.19 vs 0.16 per 1000 patient-days, p=0.78). Annual gown costs higher with contact isolation strategy ($183 609 vs $25 812).
Biehl et al, 2019 Controlled before-and-after20	Oncology wards at 4 German hospitals n=2968 patients	Single room contact precautions	No statistically significant effect on acquisition of multidrug-resistant E. coli
Biehl et al, 2022 Controlled before-and-after21	Oncology wards at 4 German hospitals n=3079 patients	Single room contact precautions	VRE acquisition was 4.8% lower in single room contact precaution patients but this was less than the a priori 10% non-inferiority
Camus et al, 2011 Randomised trial22	2 ICUs in France n=500 patients	Addition of contact precautions (consistent gowns and gloves when entering the room, face masks for close contact) and decontamination to standard precautions	No difference in MRSA acquisition between groups (5.3% vs 6.5%, p=0.58)
Evans et al, 2023 Prospective cohort analysis of non-randomised discontinuation of study practices23	123 acute care hospitals (all Veterans Affairs hospitals) n=5 225 174 patient-days	Optional discontinuation of any combination of MRSA active surveillance testing (AS), contact precautions for patients colonised with MSRA (CPC) and/or contact precautions for patients infected with MRSA (CPI)	Higher hospital-wide MRSA HAI rate when all three practices were discontinued (no AS or CPC or CPI) compared with continuing any combination of these practices (0.22 vs 0.09-0.12 MRSA HAI per 1000 patient-days, p<0.05). Discontinuing all three practices (no AS or CPC or CPI) showed higher rates of MRSA HAI compared with continuing all three practices (AS+CPC+CPI) both in ICU patients (0.65 vs 0.20 MRSA HAI per 1000 patient-days, p<0.001) and non-ICU patients (0.12 vs 0.07 MRSA HAI per 1000 patient-days, p=0.01).
Huang et al, 2019 RCT24	24 centres (17 acute care hospitals, 7 nursing homes) n=2121 patients	Post-discharge hygiene education alone vs patient education plus decolonisation protocol (chlorhexidine mouthwash and bathing; nasal mupirocin) repeated in 5 day courses twice per month for 6 months	Over 1-year follow-up, decolonisation arm had 30% lower risk of MRSA infection (HR 0.70; 95% CI 0.52 to 0.96); 29% lower risk of hospitalisation for MRSA infection (HR 0.71; 95% CI 0.51 to 0.99); 17% lower risk of any clinically judged infection (HR 0.83; 95% CI 0.70 to 0.99); 24% lower risk of hospitalisation for any infection (HR 0.76; 95% CI 0.62 to 0.93)
Kluytmans et al, 2019 Cluster-randomised crossover trial25	16 Dutch hospitals, medical and surgical wards n=10 220	Contact precautions in a single room vs a multiple-bed room	No significant difference in transmission of ESBL producing Enterobacterales to at least one wardmate (3.4%, 90% CI −0.3 to 7.1)
Maechler et al, 2020 Cluster-randomised crossover trial26	4 European university hospitals n=16 091 patients in contact isolation period vs 16 163 patients in standard precaution	Contact isolation targeting ESBL-E infection or colonisation, vs universal standard precautions	Incidence density of ward-acquired ESBL-E was 6.0 events per 1000 patient-days at risk during periods of targeted contact isolation, vs 6.1 per 1000 patient-days at risk during periods of universal standard precautions (p=0.9710) corresponding to incidence rate ratio of colonisation or infection of 0.99 (95% CI 0.80 to 1.22)
Martin et al, 2018 Discontinuation study (before/after)27	Single acute care hospital n=50 268 patient-days	De-implementation of routine use of contact isolation precautions for patients infected or colonised with MRSA/VRE	Non-infectious adverse events (postoperative respiratory failure, haemorrhage/haematoma, thrombosis, wound dehiscence, pressure ulcers, falls/trauma) decreased by 19% (12.3 to 10.0 per 1000 admissions, p=0.022) (infectious outcomes were included in a relevant review)
McConeghy et al, 2017 RCT28	10 nursing homes (5 per arm, pair-matched) n=861 patients at baseline	Multicomponent infection prevention/control bundle with staff education, sanitation supplies, and auditing/feedback dashboard for infection rates and high-touch surface cultures	Total infections 2.9 vs 4.1 per 1000 patient-days (p=0.03), lower respiratory infections 0.8 vs 1.5 per 1000 patient-days (p=0.01); neither reached significance in difference-in-difference analysis. No difference in antibiotic starts or hospitalisation.
Mehta et al, 2013 Controlled before-after study29	Single orthopaedic acute care hospital; control affiliated university hospital n=128 187 patient-days	Preoperative decolonisation protocol (nasal mupirocin and chlorhexidine) plus screening MRSA nares cultures to determine perioperative antibiotic choice	Clinical MRSA culture prevalence density reduced from 1.23 to 0.83 per 1000 patient-days (p=0.026), while control hospital saw no difference over timeframe (1.27 vs 1.24 per 1000 patient-days, p=0.787)
Miller et al, 2023 Cluster RCT30	28 nursing homes (14 in each arm) n=3 109 607 patient-days	Routine bathing vs use of chlorhexidine for all bathing/showering plus nasal povidone-iodine twice daily for 5 day periods (at admission then every other week)	Comparing intervention to baseline period, risk ratio for transfer to hospital due to infection was 1.00 in routine care arm vs 0.83 in decolonisation arm (difference in risk ratio 16.6%, 95% CI 11.0 to 21.8), and risk ratio for transfer to hospital for any reason was 1.08 in routine care arm vs 0.92 in decolonisation arm (difference in risk ratio 14.6%, 95% CI 9.7 to 19.2).
Mitchell et al, 2019 Stepped wedge randomised trial31	11 acute care hospitals in Australia n=4.8 million bed days	A bundle of environmental cleaning strategies	VRE infections were reduced with the intervention (RR 0.63, 95% CI 0.41 to 0.97), there was no statistically significant change in MRSA or C. difficile infection
Popiel and Miller, 2014 Time series analysis of discontinuation32	Single acute care urban tertiary teaching hospital n=23 000 admissions per year	Change from all admissions screened for VRE to only admissions from endemic hospitals or admitted to high-risk wards; reduced contact tracing, discontinuation of cohorting, no VRE surveillance	Coincident with discontinuation of practices there was an increase in the number of new VRE-colonised patients per quarter from <40 to >100 (statistical testing not performed); definite clinical VRE infections rose from 0 to 5 cases per quarter to 10 cases per quarter
Ray et al, 2017 RCT33	15 acute care hospitals (one additional hospital dropped out after randomisation) Sample size not reported	Fluorescent marker room cleaning monitoring and feedback for environmental services staff, vs usual care	No difference in hospital-acquired C. difficile infection at intervention hospitals before vs after protocol implementation
Salgado et al., 2013 RCT34	3 intensive care units n=614 patients	Copper vs standard materials for high-touch surfaces in ICU rooms	Hospital acquired infection and/or acquisition of MRSA or VRE colonisation 7.1% vs 12.3% (p=0.02); hospital-acquired infection only 3.4% vs 8.1% (p=0.013)

CDC, Centers for Disease Control and Prevention; ICU, intensive care unit; MRSA, methicillin-resistant Staphylococcus aureus; RCT, randomised controlled trial; VRE, vancomycin-resistant Enterococcus.