Editorial 

April 1, 2024

Controlling Multidrug-Resistant Organisms Across Patient-Sharing Networks

Christopher J. Crnich

JAMA. Published online April 1, 2024. doi:10.1001/jama.2024.0267

Antimicrobial resistance is a major public health problem that threatens to reverse many of the medical therapy advances achieved over the past 4 decades.¹ Infections caused by multidrug-resistant organisms (MDROs) are estimated to cause 35000 deaths in the US every year² and increase medical costs by $4.6 billion.³Efforts to stem the emergence and spread of MDROs have primarily centered on the implementation of preventative interventions in acute-care facilities. While rates of colonization and infections caused by MDROs are high in many hospitals, MDRO colonization rates among residents of long-term care facilities (LTCFs) are often far higher than those observed in tertiary care referral facilities.^4,5 Residents of LTCFs are frequently hospitalized⁶ and LTCFs that provide high volumes of postacute care have repeatedly been implicated in regional outbreaks caused by epidemiologically important MDROs.^7-9 These findings point to a critical need to expand MDRO efforts beyond the hospital setting.

In this issue of JAMA, Gussin et al¹⁰ examine the implementation of a universal MDRO decolonization intervention in 16 hospitals and 19 LTCFs (16 skilled nursing facilities and 3 long-term acute care hospitals) in Orange County, California. The intervention involved twice-daily application of an intranasal iodophor for 5 days every other week combined with routine bathing using a chlorhexidine (CHG)-containing product. All residents of participating LTCFs were exposed to the decolonization intervention while its application in hospitals was limited to patients undergoing contact precautions. After implementation of the intervention, MDRO colonization among residents and patients of participating facilities was 33% lower (odds ratio [OR], 0.67; 95% CI, 0.50-0.89) in long-term acute care hospitals, 23% lower (OR, 0.77; 95% CI, 0.69-0.86) in skilled nursing facilities, and 14% lower (OR, 0.86; 95% CI, 0.75-0.98) in hospitals compared with baseline. Rates of incident MDRO clinical cultures decreased 23% (risk ratio, 0.77; 95% CI, 0.63-0.99) from baseline in long-term acute care hospitals and rates in skilled nursing facilities and hospitals were 30.4% (95% CI, 16.4%-42.1%) and 12.9% (95% CI, 3.3%-21.5%) lower compared with facilities in Orange County that did not implement the MDRO decolonization intervention. Finally, in intervention compared with control skilled nursing facilities, infection-related hospitalizations, costs, and deaths were all lower.

Intervention hospitals and LTCFs in Gussin et al¹⁰ were selected based on their degree of patient sharing.¹¹ Interfacility transfer of residents and patients connects health care facilities within a region and affects MDRO prevalence.¹² Consequently, a hospital engaged in aggressive MDRO containment efforts may continue to experience increasing numbers of MDROs if other facilities within its patient-sharing network do not make the same efforts. The study by Gussin et al¹⁰ highlights the value of coordinating MDRO control efforts across a given patient-sharing network and particularly the benefit of centering these efforts on LTCFs.

Another aspect of the intervention used in Gussin et al¹⁰ that deserves mention is the universal application of the intervention to residents of participating LTCFs. Vertical interventions, based on active surveillance and isolation of colonized individuals, have been applied in LTCFs to successfully control the regional spread of methicillin-resistant Staphylococcus aureus, vancomycin-resistant Enterococcus, and carbapenem-resistant Enterobacterales. However, large-scale screening of resident populations can be costly and the application of contact precautions when colonization is identified may adversely affect resident well-being. Horizontal interventions have broader outcomes that can simultaneously reduce an individual’s risk of becoming colonized and/or infected with different types of MDROs.¹³ Along those lines, the intervention in Gussin et al¹⁰ appears to have had a reasonably balanced beneficial outcome in multiple target MDROs, including methicillin-resistant Staphylococcus aureus, extended spectrum β-lactamase producers, and carbapenem-resistant Enterobacterales. Moreover, universal application of the intervention, rather than targeting specific high-risk individuals, may have helped routinize practice in LTCFs, which may explain the high levels of adherence to CHG bathing (86.3%) and application of intranasal iodophor (69.5%) observed in this study.

The study by Gussin et al¹⁰ is not without its potential limitations. First, costs associated with procuring iodophor and CHG products as well as added staff time to conduct the decolonization procedures are not enumerated, and formal cost-effectiveness studies examining the intervention from a variety of payor perspectives are warranted. Second, while the results of a 2023 cluster-randomized trial conducted in 28 LTCFs¹⁴ may address causal inference questions raised by the quasi-experimental design component in the study by Gussin et al,¹⁰ the mechanisms by which the MDRO decolonization intervention achieves its effect on clinical outcomes remain ambiguous. Specifically, the relative contributions of intranasal iodophor application and bathing using CHG are not clear as their combination was not compared with the use of either agent individually. Additionally, it can be argued that simply routinizing bathing practices in LTCFs can decrease a resident’s risk of MDRO colonization and the lack of detail on bathing practices in control facilities is a missed opportunity. Third, depending on the nature and intensity of support provided by the investigative team to intervention health care facilities, there is potential for off-target outcomes in facility infection prevention behaviors (eg, indwelling device management and wound care practices) that may be associated with some of the observed clinical outcomes. Given the uncertainties around cost benefit and causal mechanisms, it would appear additional studies are needed before widescale adoption of the MDRO decolonization intervention used in Gussin et al,¹⁰ can be recommended.

These concerns aside, the outcomes observed by Gussin et al¹⁰ are clinically meaningful and highlight the value of implementing similar MDRO prevention efforts across hospitals and LTCFs in the same patient-sharing network. Defining the boundaries of these networks, the limited administrative integration among most facilities within these networks, and lack of obvious financing mechanisms are major barriers that need to be overcome before the benefits of network coordinated MDRO prevention intervention can be fully realized. Nevertheless, the power of patient-sharing networks in the amplification and spread of MDROs cannot be denied. We are all in this fight against MDROs together and Gussin et al¹⁰ are to be commended for reporting that going it alone is no longer an option.