Discussion

The number of reported CRO cases in Alameda County increased from July 2019 to July 2021 (Figure 1), similar to increases observed elsewhere in recent years. ^{Reference Tadese, Darkoh, DeSantis, Mgbere and Fujimoto4,Reference Villegas, Hitchingham, Wilson, Gambrell, Mounsey and Muleta14} Twenty-four percent of the CRO cases over the 2-year period were carbapenemase-producing, and 28% of reportable CRE cases were carbapenemase-producing. This is higher than the percentage found during the same time frame in Tennessee (14%–26%) ^{Reference Villegas, Hitchingham, Wilson, Gambrell, Mounsey and Muleta14} and within the range found by the Multi-site Gram-negative Surveillance Initiative from 2011 to 2015 (16%–91%). ^{Reference Karlsson, Lutgring and Ansari15} In line with US trends for CP-CRE, ^{Reference Karlsson, Lutgring and Ansari15} KPC was the most common carbapenemase found followed by NDM (Table 3). The percentage of CP-CRE in a population varies greatly by geography, ^{Reference Chen, Jean and Lee16} highlighting the need for understanding and sharing local epidemiology.

Most cases were identified from respiratory or urine samples and less than 15% from rectal swabs, an indirect indication that a high proportion of cases identified were clinical samples rather than surveillance samples. CRE isolates were most likely to be isolated from urine, while CRPA and CRAB isolates were most likely to be isolated from respiratory samples. This may have implications for screening patients when there is an outbreak of a specific organism. Although rectal swabs are often used for screening, other sites might be considered for better detection, depending on the organism being screened. Although studies have found the groin and rectal/perirectal regions to be most sensitive for CRE and MDR Enterobacterales, ^{Reference Thurlow, Prabaker and Lin17,Reference Weintrob, Roediger and Barber18} sensitivity of different screening sites is higher for other CROs, such as the skin or buccal mucosa and the skin for CRAB, ^{Reference Nutman, Temkin, Lellouche, David, Schwartz and Carmeli19} while for others, like CRPA, screening site specificity is not well documented. Further studies would be warranted to determine if urine and rectal culture results tend to be concordant to suggest urine as a screening site for CRE, as well as respiratory and rectal cultures as screening sites for CRPA and CRAB.

There are currently many more studies assessing the epidemiology of CRE than CRAB in the United States, and the emergence of CRAB and 1.83-fold increase in our county without a reporting mandate highlights the importance of conducting surveillance for this organism. Although the CDC shows a decrease in hospitalized carbapenem-resistant Acinetobacter cases from 2012 to 2017, ¹ there is still a concern around CRAB due to the high percent producing carbapenemases. ^{Reference McKay, Vlachos and Daniels20} Our analysis found that 59% of CRAB cases were carbapenemase-producing (Table 2) and that the most common carbapenemase genes conferring resistance among CRAB cases were NDM (6/46, 13%), OXA-23-like (7/46, 15%), and OXA-235-like (9/46, 20%) (Table 3). This differs from the gene breakdown of CRAB isolates submitted to the CDC from 2013 to 2017: 49% of which had OXA-23-like, 5% having OXA-235-like, and none having NDM; ^{Reference McKay, Vlachos and Daniels20} likely due to three LTCFs experiencing outbreaks during the study period. Other CROs were also epidemiologically linked to cases in other facilities, demonstrating likely transmission between facilities.

With regard to attribution, half of the patients were at an STACH, while 42% were at an LTCF when their first CRO organism was identified. Although other studies of CRE in the United States have found a majority of cases identified from STACHs, ^{Reference Tadese, Darkoh, DeSantis, Mgbere and Fujimoto4,Reference Marquez, Terashita, Dassey and Mascola21} incidence rates have been found to be higher among LTACHs than STACHs ^{Reference Livorsi, Chorazy and Schweizer22} and higher prevalence rates among LTCFs compared with STACHs. ^{Reference Chen, Jean and Lee16} Transmission of CRE in LTCFs has been well documented, ^{Reference Chitnis, Caruthers and Rao23,24} and prior stays in high-acuity LTCFs have been identified as risk factors for CRE carriage at hospital admissions, ^{Reference Prabaker, Lin and McNally25} indicating LTCFs as a reservoir for CRO colonization. ^{Reference Chen, Jean and Lee16}

Of particular concern are outbreaks and transmission of CP-CROs at LTCFs. The spread of carbapenemase genes via plasmids has been implicated in healthcare facility outbreaks. ^{Reference Kanamori, Parobek and Juliano26} Our analysis found that cases identified at LTCFs were more likely to be CP-CRO cases than those at acute care facilities, although the increased odds were not significant. The vSNFs were found to have significantly higher odds of CP-CRO cases than other facility types. This is similar to other analyses that have identified vSNFs and time on ventilators as risk factors for CP-CROs during outbreaks. ^{Reference Chitnis, Caruthers and Rao23,Reference Prabaker, Lin and McNally25} Due to the heightened potential for transmission and effect on infection control, it is important to test for carbapenemase genes in the vSNF setting. Focusing limited public health resources in high-risk settings such as vSNFs can be extremely impactful in limiting the spread of antibiotic resistance.

Due to the scale of the three vSNF and LTACH outbreaks during the study period, Alameda County worked with these facilities to implement a multi-interventional approach similar to other successful outbreak investigations. ^{Reference French, Coope and Conway27} Facilities closed to new admissions during the outbreak, implemented contact precautions for CRO patients, and ACPHD recommended strategic cohort isolation of patients and staff, provided education to staff, conducted onsite assessment and adherence monitoring to infection control and prevention practices, and conducted point prevalence surveys every two weeks until the outbreaks resolved. Local public health must coordinate regional surveillance and response to prevent and control vSNF/LTACH outbreaks.

This analysis is subject to several limitations. First, most cases occurred during the COVID-19 pandemic during which there was limited staff capacity at the local public health department and at healthcare facilities that may have affected the completeness of reporting. In addition, this analysis includes data for both mandatory and voluntarily reportable diseases. The sensitivity analysis demonstrated relatively high completeness for CRE, but isolate submission for CRE cases was not as complete. Voluntary non-CRE CROs had lower completeness results. Only 60% of non-CRE CROs on record at a local SNF were reported to the health department and included in this analysis, and only one-quarter of those cases had isolates submitted, which may have biased the distribution of CROs detected. CROs from facilities experiencing outbreaks may be overrepresented. These data demonstrate that CROs are an increasing issue and worthy of surveillance and should be considered for mandated reporting beyond CRE. Furthermore, these results data support how mandated reporting allowed for more complete case finding than voluntary reporting.

In addition, because the data analyzed are surveillance data, we are limited to what is entered into CalREDIE to determine the facility the patient was at when they tested positive for a CRO. This may lead to some misclassification of facility type associated with a case, as a case may be tested and reported by an STACH for a patient residing in an LTCF, without the STACH indicating that the patient was transferred from an LTCF. To reduce this misclassification, we utilized patient addresses in the surveillance system to determine if they resided at an LTCF.

In conclusion, the number of reported CROs increased significantly from July 2019 to June 2021 in Alameda County. CRAB cases saw a large increase, driven by regional outbreaks. Like other US jurisdictions, most CRO cases were identified in STACHs, in part due to compliance with local reporting requirements and prior stays at LTCFs that may not have been easily ascertained. LTCFs such as vSNFs are facilities at high risk for outbreaks that can amplify the spread of emerging genes, and public health resources should be focused on reducing transmission of CROs in this setting. These data could also help target where screening for specific CROs occurs (respiratory vs urine samples) to improve effectiveness with limited resources.

Finally, Alameda County’s local health officer order allowed the local public health department to act on multiple CRO outbreaks and partner with facilities to implement a multi-interventional approach to limit the transmission of CROs. It is important to focus on these highly resistant organisms because many of the patients with these CROs had poor clinical outcomes, including death, highlighting the significant health impact of these conditions.