Distribution and antibiotic susceptibility profiles of A. baumannii strains isolated from hospital inanimate surfaces and hands
Among the 34 strains, 28 were from the environment, and 6 were from ICU workers’ hands. As shown in Table 2, the environmental isolates were most frequently isolated from mice and keyboards, ventilators, IV pumps, patient files, and bedrails, with detection rates of 17.86% (5/28), 17.86% (5/28), 14.29% (4/28), 10.71% (3/28), 10.71% (3/28), respectively. Approximately 67.65% (23/34) of the strains were extensively-drug-resistant (XDR), and 11.76% (4/34) were multidrug-resistant (MDR). In particular, carbapenem resistance was detected in 75% (21/28) of the environmental isolates and 83.33% (5/6) of the finger isolates.
MLST analysis with the Pasteur and Oxford scheme
Using the Pasteur MLST scheme, 33 (97.06%) A. baumannii isolates were identified as having eight sequence types (STs), except for one isolate that could not be assigned to a distinct ST and was defined as STnew (Table 4). The most frequent ST was ST2, which was found in 22 isolates (22/34, 64.71%), followed by ST242 (4 isolates, 11.76%) and ST93 (2 isolates, 5.88%), and one each was assigned to ST/195, 882, 1641 and ST25. MLST analysis also revealed that 84.62% of CRAB (22/26) belonged to ST2. Twenty-two strains of ST2,17 were from the high-touch surfaces of the patient’s surroundings, 4 were from health care workers’ hands, and 1 was from the telephone of the nurse station. Among the 4 strains of ST242, 2 were from the doctor’s keyboard, 1 was from the printer of the nurse station, and 1 was from the sleeves of the nurse (Table 3).
MLST analyses based on the Oxford scheme identified 7 STs for 28/34 strains (82.35%). Four strains could not be typed since the gdhB gene could not be assigned to the number of alleles, while two strains (5.88%) could not be assigned to a distinct ST and were defined as STnew1 or STnew2. ST195 was the most prevalent sequence (52.94%, 18/34), followed by ST208 (11.76%, 4/34) and ST1986 (5.88%, 2/34). The abundances of ST368, 1897, 712 and ST229 were determined once. Among the 18 strains of ST195, 14 were from the high-touch surfaces of the patient’s surroundings, and 4 were from health care workers’ hands (Table 3).
Compared to the Oxford scheme, ST2 (Pasteur) corresponds to three STs (Oxford), namely, ST195, ST368 and ST208 . In addition, one strain of ST33 and four strains of ST242 in the Pasteur scheme were not be defined as distinct STs in the Oxford scheme (Table 3).
According to the results of the goeBURST algorithm, CC195 (Oxford) was the largest CC, including 23 strains (2-15, 17, 24-30, 32) of 3 STs (ST195, ST368, and ST208). In addition, ST195, ST368 and ST208 were closely related genetically because they shared six of seven housekeeping genes tested, differing only in the the gpi gene (Figure 1B, Table 3). Conversely, ST195 (Oxford) had 2 different STs (ST195, ST2) in the Pasteur scheme. The goeBURST algorithm used in the present study showed that CC2 (Pasteur) was the largest CC (Figure 1A), including 23 strains (2-15, 17, 24-30, 32) in two STs (ST2 and ST195). Although there was no complete correlation between STs from both schemes, the CCs of the Oxford scheme and the Pasteur scheme corresponded to each other.
ERIC and REP-PCR analyses
REP-PCR analysis revealed the presence of 4 clusters and 11 single types (Figure 2). The predominant type was A, which contained 13 isolates (38.24%). Among them, 12 strains (11 environmental strains and one hand strains)were obtained from ICU1, and 1 environmental strain was obtained from ICU2. Moreover, 5 isolates were clustered into Group B (4 environmental strains and one hand strain from the NICU), followed by Group C (3 isolates from NICU) and Group D (2 isolates from the NSICU). Genotyping of A. baumannii isolates revealed 4 different ERIC-PCR groups and 10 single types. The majority of isolates belonged to either type A (12/34, 35.29%; 10 environmental isolates from ICU1, 1 environmental strain and 1 hand strain from ICU2) or type B (8/34, 23.53%; 3 environmental strains and 2 hand isolates from ICU1; 2 environmental strains and 1 hand strain from ICU2). Both types C and D contained 2 isolates each. Moreover, compared to the STs of the Pasteur scheme, two profiles of REP-PCR (A, C) and four profiles of ERIC-PCR (A, B, D, E) belonged to ST2.
Detection of drug-resistance genes
PCR analysis demonstrated that the OXA-51 gene was the most prevalent subgroup. The gene was detected in 32 of 34 (94.12%) A.baumannii isolates. The frequencies of the acquired OXA carbapenemase genes OXA-23, OXA-24 and OXA-58 were detected in 24 (70.59%), 4 (11.76%) and 3 (8.82%) isolates, respectively. In addition, the OXA-143 gene was negative in all the isolated strains. Among the metallo-β-lactamase genes, the most widespread was VIM, which was found in 79.41% (27/34) of the isolates, followed by SIM in 47.06% (16/34), NDM in 41.18% (14/34) and IMP in 20.59% (7/34). OXA-23 and VIM were the most common carbapenemase genes among the CRAB isolates (88.46%, 23/26). OXA-51, OXA-23 and VIM coexisted in 67.65% (23/34) of the isolates.
Distribution of qac genes
All A. baumannii isolates were assayed for the presence of qac genes. The distribution of the qac genes is shown in Table 4. Six isolates (17.65%) of A. baumannii carried no qac genes. The most prevalent gene was qacΔE1 (79.41%, 27/34), which could be detected in CRAB (88.46%, 23/26) and non-carbapenem-resistant A. baumannii (40%, 4/8). The coexistence of qacΔE1, OXA-51, OXA-23 and VIM was only detected in 76.92% (20/26) of the CRAB isolates. The qacH gene was also detected in CRAB (30.77%, 8/26). In addition, the qacA/B gene was detected in 11.76% (4/34) of the isolates, qacC/D in 11.76%, qacE in 5.88% (2/34), qacF in 5.88% and qacG in 2.94%. However, the qacJ gene was not detected in any of the isolates.