Introduction
Carbapenemase-producing organisms are epidemiologically significant due to easily transferred mobile genetic elements such as plasmids and transposons. Reference Chiotos, Han and Tamma1 Carbapenemases have been responsible for rapid global spread of carbapenem-resistant Enterobacterales. Reference Nordmann, Naas and Poirel2 New Delhi Metallo-Beta-Lactamase (blaNDM) is one such carbapenemase first described in 2009 in India. Reference Yong, Toleman and Giske3 At the time of this outbreak, Enterobacterales carrying blaNDM had been reported in 34 states of the United States with only 7 reported cases of blaNDM-carrying Pseudomonas aeruginosa from four states but none from Texas. 4 We describe a clinical and molecular epidemiologic investigation of the first reported outbreak of blaNDM-carrying P. aeruginosa in the State of Texas and describe the control measures that were effective in quickly containing the spread of blaNDM-carrying organisms at the spinal cord injury center (SCI) and the acute care hospital (ACH) within Veterans Affairs North Texas Health Care System. We also report transmission of blaNDM-1-carrying P. aeruginosa to one roommate and blaNDM-5-carrying Escherichia coli to another roommate of the same index patient.
Methods
Setting
Veterans Affairs North Texas Health Care System (VANTHCS) serves veterans from 38 counties in Texas and 2 counties in Southern Oklahoma. It has 835 operating beds including an academic ACH and a SCI with 30 beds.
Investigation period
This outbreak investigation occurred from July 2018 to October 2018.
Clinical and molecular epidemiologic investigation
A case was defined as any patient admitted to the SCI or the ACH in whom a blaNDM-carrying organism was either cultured clinically or a rectal screen with a real-time polymerase chain reaction (PCR) assay (Xpert® Carba-R) was positive for blaNDM. The PCR assay from rectal screen is only able to identify the presence and type of the carbapenemase gene but unable to specify the genera of bacteria carrying the gene. An indirect contact was defined as a patient who was cared for by hospital staff who also cared for the index patient. A clinical epidemiologic investigation was launched when the index case was identified with a urine culture positive for blaNDM-carrying P. aeruginosa. Universal contact isolation was initiated in SCI and all non-critical shared equipment on SCI and ACH unit underwent supervised disinfection. A systematic approach was undertaken with SCI roommates screened initially with a rectal swab for real-time PCR testing to identify any of the five carbapenemase genes including blaNDM expanding investigation to patients outside the index patient’s room once the screen was positive. We reviewed electronic medical record and staffing schedules to identify patients who shared healthcare staff with the index patient in the ACH and approached those patients for consent to screen. We performed phased point prevalence testing with rectal screening three times, four weeks apart. Detection of carbapenemase production and molecular characterization of blaNDM gene using PCR assay were performed as described earlier. Reference Paul, Dhar and Maurya5 Whole genome sequencing (WGS) was performed by the Centers for Disease Control and Prevention (CDC) with short-read sequencing on all three (Illumina MiSeq) and long-read sequencing on one Pseudomonas isolate (PacBio). The project was deemed as quality improvement by an institutional review process and the need for research approval was waived.
Results
Our investigation revealed that the index patient was transferred to the SCI from Thailand following a three-month hospitalization due to injuries sustained from a motor vehicle accident while vacationing overseas. He was transferred the day after admission to the ACH within the same campus due to hypokalemia where he stayed for 7 days and then transferred to SCI sharing a room with two other patients for 31 days before a urine culture grew blaNDM-carrying P. aeruginosa. Rectal screens performed immediately after this were positive for blaNDM in the index patient and one of the two roommates. Twenty-nine days later, the roommate with positive screen had a urine and a coccyx culture positive for blaNDM-carrying P. aeruginosa, and the second roommate had a urine culture positive for blaNDM-carrying E. coli. Treatment was unnecessary during this outbreak as positive cultures were deemed secondary to colonization. A third patient who was an indirect contact in the same unit as the index patient in ACH tested positive for blaNDM with rectal screening (Figure 1). A total of 54 patients were identified as indirect contacts to the index patient. Of the 54 patients, 28 patients underwent rectal screening, the remaining either refused or were unable to get tested. In addition, point prevalence rectal screening was conducted in three phases in the SCI and included a total of 30 patients. All these tests were negative. Whole genome sequencing revealed that index patient and roommate 1 had blaNDM-1-carrying P. aeruginosa whereas roommate 2 had blaNDM-5-carrying E. coli. Whole genome sequencing did not reveal any plasmids in blaNDM-1-carrying P. aeruginosa. No further spread occurred. Our investigation ended after twelve weeks with all the rectal screens testing negative for blaNDM during the three phases of point prevalence testing. Risk factors and comorbidities, time between exposure to positive test, clinical outcome, and resistance genes for all patients identified to be carrying the blaNDM are outlined in Table 1.
Figure 1. Diagrammatic representation of cases showing blaNDM-carrying organisms in indirect and direct contact with the index patient.
NDM, New Delhi Metallo-Beta-Lactamase.
Table 1. Clinical epidemiologic data and whole genome sequencing in case patients identified to have blaNDM
NA, not applicable; NDM, New Delhi Metallo-Beta-Lactamase.
Index patient with NDM-1 Pseudomonas aeruginosa clinical culture and screen positive. 
NDM-1 Pseudomonas aeruginosa clinical culture and screen positive. 
NDM-5 E. coli clinical culture and screen positive. 
NDM screen positive with real-time PCR; no culture positive.
Discussion
P. aeruginosa carrying blaNDM has been infrequently detected within the US. However rapid spread of carbapenemase-producing organisms has occurred globally likely owing to international travel followed by local spread. For example, the first KPC-producing Klebsiella pneumoniae was discovered in the United States in 1996 from a hospital in North Carolina. Reference Yigit, Queenan and Anderson6 Since then, there has been exponential growth in KPC prevalence in the US (CDC, https://arpsp.cdc.gov/profile/antibiotic-resistance). In addition to sporadic international travel-related cases, occasional domestically acquired NDM P. aeruginosa have been reported. Reference Gray, Beaird, Smith, Schaenman and Yang7 Our efforts to control the blaNDM-carrying organisms have quickly mitigated further spread within our SCI and ACH.
One of the roommates of the index patient had a urine culture positive for E. coli with blaNDM-5. There was no plasmid detected to indicate intergenus transfer of bla NDM. It is very likely that the index patient carried both blaNDM-1-carrying P. aeruginosa and blaNDM-5-carrying E. coli and transmitted one each to each of his two roommates. blaNDM-5 was first recovered in the United Kingdom in a patient with a history of travel to the Indian subcontinent. Reference Hornsey, Phee and Wareham8
Our study has its limitations. First, this is a single-centered, observational study in a specific patient population, which may not be generalizable to other care settings. Nevertheless, we demonstrated that timely interventions helped us limit the outbreak to a small number of patients. Reference Lin, Froilan and Makhija9,10 By promptly identifying and isolating direct and indirect contacts, we were able to limit the transmission to the three contact patients with two of them being index patient’s roommates. Second, we only screened indirect contacts whom we defined as those that had common healthcare staff with the index patient. We did not explore colonization in patients who may have shared non-critical medical equipment and therefore cannot ensure that there was no transmission to the patients who were not screened. However, we have not seen any subsequent infections develop in any of our veterans. In conclusion, timely interventions including prompt isolation of colonized patients were effective in curbing transmission of blaNDM-carrying organisms. It is very likely that our index patient was colonized with both blaNDM-1-carrying P. aeruginosa and blaNDM-5-carrying E. coli and transferred one each to each of his roommates; in addition, transferred blaNDM-carrying organism/s to another patient who is an indirect contact, either through common staff or shared equipment.
Acknowledgments
We thank the VA North Texas healthcare staff and the Texas Department of State Health Services for their help during this outbreak investigation. We thank Dr. Alison Laufer Halpin, PhD from the Centers for Disease Control and Prevention (CDC) and the CDC for whole genome sequencing of the NDM isolates.
Financial support
No financial support was provided relevant to this article.
Competing interests
All authors report no conflicts of interest relevant to this article.
Open access
