Given that MRSA can frequently spread inside geriatric LTCFs via cross-transmission among elderly residents, those with longer periods of residence are considered to have a higher risk for MRSA colonization, which would result in a higher prevalence of MRSA. However, our findings conversely indicated that the prevalence of MRSA was lower among residents with longer residence periods than those with shorter periods. Furthermore, among residents living in LTCFs ≥ 2 months from their initial admission, only 1 in 60 was identified as having positive conversion to MRSA colonization whereas about half had negative conversion. Findings from previous studies remain controversial as to whether the high prevalence of MRSA among LTCF residents is mainly owing to importation from outside the facility via admission of residents who are colonized with MRSA or whether the high prevalence is owing to cross-transmission inside of LTCFs. [33, 34, 39, 40] Our findings indicated that MRSA was most likely to be carried into the LTCFs via transfer of residents rather than spread via cross-transmission inside the LTCFs.
Previous studies have reported that among LTCF residents, 20–50% are potentially persistent nasal MRSA carriers, but about 50% of these residents exhibit negative conversion with time. [23, 40] Another article also indicated that MRSA within the human nasal cavity can disappear over time,  which is consistent with our results. If residents indeed acquire MRSA via transmission from persistent carriers inside LTCFs, those with longer periods of LTCF residence would be at greater risk for MRSA acquisition, which would prove our hypothesis that LTCF residents with longer periods of residence have a higher prevalence of MRSA. Some previous studies have assumed that geriatric LTCFs are potential reservoirs for MRSA because of the high MRSA prevalence in these facilities. [24, 29–35] However, our study indicated that admission of residents with nasal MRSA colonization might be a primary contributor to the high prevalence of MRSA among LTCF residents.
There are two possible routes via which MRSA may be introduced to LTCFs from outside the facility. The main route would likely be importation from hospitals, which is largely supported by the results of previous studies. [35–38] In addition to these reports, one study previously found that LTCFs with a larger number of hospitals located nearby had a higher prevalence of MRSA than LTCFs with fewer hospitals nearby.  Another possible route is via the general population.  The prevalence of MRSA in the Japanese general population is estimated to be ≤ 5%. [46–48]; however, this prevalence might be higher among elderly adults owing to frailty, in comparison with younger people.  Furthermore, regional MRSA epidemics of community-acquired strains have recently occurred in Japan, which might affect MRSA prevalence. [45, 48, 50, 51] A previous study found regional differences in the MRSA prevalence. [44, 52] Our results showed that the prevalence of MRSA differed among LTCFs, which might reflect differences in the MRSA prevalence among the general population where each LTCF is located.
Care providers in geriatric LTCFs should consider that frail residents who are initially admitted to the facility are most likely to have nasal MRSA colonization, especially those admitted from hospitals. Universal precautions against MRSA transmission should therefore be appropriately applied by all LTCF staff. Nasal application of mupirocin for residents identified as having MRSA at the initial admission might be effective for the prevention of MRSA transmission inside an LTCF; however, this should be carefully assessed owing to the high costs as well as the possible increase in MRSA with resistance to mupirocin.  Among various prevention strategies, thorough hand hygiene as well as appropriate use of gloves is recommended to maximize deficient medical resources in LTCFs and to reduce excess costs [54, 55]
This study includes some limitations. First, we did not obtain complete information on the background of residents, such as sex, age, general condition, and medical history owing to the ethics protocol followed; the ethics review board did not grant approval to obtain this information. Among these factors, general status and medical history might affect the prevalence of MRSA.  Second, we also could not obtain information regarding from where residents had been transferred to the LTCFs. Third, not all residents initially admitted to the LTCFs were included in Cohort 2 because some did not agree to undergo secondary MRSA testing. Furthermore, we performed MRSA testing only twice after admission. Multiple tests with longer follow-up might be required to accurately assess status of MRSA colonization. Fourth, we only performed nasal MRSA testing. Some residents have MRSA on their skin or in wound sites, which might result in a higher prevalence. Finally, we did not exclude residents from the analysis who had previously been discharged from an LTCF but who were subsequently readmitted.