This was a retrospective study conducted at Institute of Nephro-Urology, a tertiary care referral hospital inside Victoria Hospital Campus, Bangalore. The data from July 2018 to July 2020 was retrieved from patients’ medical records. The process of data retrieving and analysis was performed from January 2020 to July 2020.

Inclusion criteria: All CKD patients admitted in Renal Intensive care unit

Exclusion criteria:

A total of 69 samples of ETA were received in Microbiology laboratory and processed immediately using standard microbiological procedures ike microscopy, culture and antimicrobial susceptibility testing.

The samples were first subjected to Gram’s staining and observed for the presence of number of pus cells and epithelial cells for assessing the clinical significance of sample as per Bartlett criteria, based on the relative number of Squamous Epithelial Cells (SEC) and inflammatory cells, seen microscopically per Low-power field (LPF). As per the criteria, a Q score which is a sum of +/- assigned values was calculated. A score of +2 is assigned, if White Blood Cells (WBCs) were more than 25/LPF and a score of +1, if 10-25 per LPF.

A score of -2 was assigned if the SEC were more than 25 per LPF and a score of -1 if 10 to 25 per LPF. A “+” Q score indicates that the quality of the specimen is acceptable, whereas a “-” or 0 Q score indicates excessive oral contamination and an unacceptable sample [20,21]. Irrespective of the Q score, the samples were processed using standard semi quantitative culture methods [22].

ETA samples were mechanically liquefied and homogenised by vortexing for one minute with 1 mL of 0.9% saline solution. The samples were then inoculated using semiquantitative calibrated loop method on 5% defibrinated sheep BA, Chocolate Agar (CA), MacConkey agar and SDA for quantifying the colony forming units per milliliter (cfu/mL) [22].

BA and chocolate agar plates were incubated in candle jar and the other plates in standard bacteriological incubator at 37°C. SDA plates were inoculated in duplicate and incubated at both, 37°C and at room temperature for up to one week as per standard guidelines [21,22]. Plates were evaluated for growth at 24 and 48 hours [22].

Following incubation, the colonies were quantified and identified using standard biochemical tests. Colony forming units of ≥10⁵ cfu/mL were considered significant for tracheal aspirates and ≥10⁴ cfu/mL for bronchoalveolar lavage [23].

All microorganisms were identified using standard laboratory methods [21]. Antimicrobial susceptibility testing was performed and interpreted by Kirby Bauer disc diffusion method as per current CLSI guidelines with appropriate quality control protocol [24].

SPSS version 23.0 (IBM, Armonk, US) software was used for data entry and analysis. Frequencies and percentages were calculated using descriptive statistics. Tables and charts were constructed using Microsoft Excel and Word. Chi square test was used to calculate p-value for age and gender statistics. p-value less than 0.05 was considered as statistically significant.

A total of 69 patients with underlying CKD were included during the study period. Out of which, 46 (66.66%) were males and 23 (33.34%) were females [Table/Fig-1]. There was no significant difference between males and females among patients showing blood culture growths (p-value=0.793). Most of the patients were in the age group between 18 to 35 years (39.13%) [Table/Fig-2]. There was no significant difference based on age group (p-value=0.123).

Predisposing risk factors such as Diabetes mellitus (64%) was seen in the majority of patients in present study, followed by alcohol abuse (45%) and hypertension (42%) [Table/Fig-3].

Out of the total 69 samples, 69.56% (48) were culture positive with significant bacterial growth (colony count >10⁵/mL). Out of the positive cultures, 5 (10.41%) were polymicrobial and 43 (89.58%) were monomicrobial growths [Table/Fig-4].

The total isolates were 53. Predominant isolates were gram negative bacteria 49 (92.4%) followed by fungal species (6%) and gram-positive cocci (2%) [Table/Fig-5]. Among the gram negative bacteria, most common was Acinetobacter baumannii (45%), followed by Klebsiella pneumoniae (23%) and Escherichia coli (11%). Other less commonly found organisms were gram negative non fermenting bacilli (7%) and Enterobacter species (6%) [Table/Fig-5].

Most of the gram-negative isolates were multidrug resistant. A total of 77.55% of the isolates were resistant to amikacin, 91.84% were resistant to cefotaxime and ceftazidime (3^rd generation cephalosporins), 91.84% were resistant to amoxicillin-clavulinic acid (betalactam + betalactamase inhibitor combination) and 89.8% were resistant to ciprofloxacin (fluoroquinolone). Meropenem (carbapenem) resistance accounted to 73.5% [Table/Fig-6].

Among the 24 isolates of Acinetobacter baumanii, 95.8% was resistant to amikacin and gentamycin, cefotaxime, ceftazidime, amoxicillin-clavulinic acid and ciprofloxacin. Resistance to meropenem and imipenem (carbapenems) was seen in 83.68% of isolates [Table/Fig-7]. Among the 12 isolates of Klebsiella pneumoniae, all were resistant to amikacin, gentamycin (aminoglycosides), cefotaxime, ceftazidime, amoxicillin-clavulanic acid and ciprofloxacin. Resistance to meropenem, imipenem, ertapenem was 100% [Table/Fig-7]. E.coli showed variable sensitivity and resistance pattern. 16.6% resistant to amikacin and meropenem, 100% resistant to cefotaxime, ceftazidime and ciprofloxacin. [Table/Fig-7]. All isolates were sensitive to polymyxin B and colistin [Table/Fig-7].

All isolates of E.coli, Klebsiellapneumonia and Enterobacter species were sensitive to tigecycline. However, the study found one isolate of Acinetobacter baumannii which was resistant to tigecycline (4.17%) [Table/Fig-7]. Among the gram positive bacteria, only one strain of methicillin sensitive Staphylococcus aureus was isolated.

Among the fungal isolates, one isolate each of Candida albicans, Candida non-albicans and Trichosporon species were isolated. A 33.33% of Candida species showed resistance to fluconazole. All were sensitive to voriconazole, amphotericin B, caspofungin and micafungin [Table/Fig-8].

Blood culture of ETA samples which did not isolate any organism was performed. No organisms were isolated. Five bacteria were isolated from blood samples of patients whose ETA samples isolated predominantly single bacteria [Table/Fig-9]. Only two of these blood cultures grew the same bacteria as the ETA culture (Acinetobacter baumanii and E.coli) [Table/Fig-10].

Offlate, resistant strains are emerging in the clinical setting due to inadvertent use of antibiotics [25]. These resistant strains are a predominant cause of nosocomial infection of patients admitted in intensive care unit. Patients in ICU are subjected to invasive procedures such as mechanical ventilation which increases the risk of nosocomial infections [26]. If not treated early, VAP is associated with severe morbidity and mortality. For early and appropriate treatment, it is important that we have a good empirical antibiotic policy. Aetiological agents vary widely based on the study population. Predisposing risk factors such as Diabetes Mellitus (DM) (64%) [Table/Fig-3] was seen in the majority of patients in present study. In a study, by Dey A and Bairy I, [17], a total of 28 patients had co-morbidities out of which 15 (60%) had VAP. In another study by Rajasekhar T et al., prolonged hospital stay (73%) was the most common risk factor [27]. DM is the most common cause of CKD and is a proinflammatory state. Hence, it is the most common risk factor in this study. This study had a positive culture rate of 69.56% [Table/Fig-4]. A study by Simoni P et al., had a positive culture rate of 100% [28]. Cardenosa Cendrero JA et al., had a positive culture rate of 89% [29]. This difference in culture positivity rate can be due to the different study population or colonisation of the ET tube [30].

This study showed predominant growth of gram negative bacilli (92%) compared to gram positive cocci (2%) [Table/Fig-5]. This was in concordance with studies by Gupta P et al., and Chandra D et al., where in 86% and 85% of isolates were gram negative, respectively [31,32]. Also, a study by Ebenezer R et al., in burns patients demonstrated higher gram negative tracheal isolates compared to gram positive isolates [33]. There is increased incidence of gram negative species in nosocomial infections. They are difficult to treat and hence cause prolonged ICU stay. Acinetobacter baumannii was the predominant isolate in this study. In other studies by Caskurlu H et al., and Swati A et al., Acinetobacter species was the most common isolate [34,35]. This is due to its ability to spread easily by aerosols and high degree of resistance to the surrounding environment and antimicrobials [36].

The second most common isolate was Klebsiella pneumoniae followed by E.coli in this study. In a study, by Chandra D et al., Klebsiella (32.35%) was the second most common bacteria isolated [32]. In this study, all isolates of Acinetobacter baumannii were sensitive to Colistin and Polymyxin B (100%) [Table/Fig-7]. It showed 95.83% sensitivity to tigecycline and 37.5% sensitivity to cefoperazone-sulbactam. A study by Malik M et al., demonstrated 69% sensitivity to cefoperazone-sulbactam [37]. Anusha N et al., showed Acinetobacter was highly sensitive to ciprofloxacin and imipenem [38]. The emergence of multidrug resistant strains may be due to inadvertent use of antibiotics. This study showed Klebsiella pneumoniae to be 100% sensitive to colistin, polymyxin B and tigecycline [Table/Fig-7]. Ahmad H et al., showed Klebsiella pneumoniae to be sensitive to cefoperazone- sulbactam (81.8%) and piperacillin-tazobactam (71.4%) in their study [39]. Decrease in sensitivity of Klebsiella may be attributed to cross infections and unjust use of antibiotics [40]. In this study, other gram-negative bacilli such as E.coli and Enterobacter showed variable sensitivity profile. E.coli showed 83.33% sensitivity to aminoglycosides, cefaperazone-sulbactam and carbapenems. A study by Anusha N et al., showed similar results. E.coli was sensitive to amikacin and meropenem in their study [38].

In the total duration of two years, the study found only one isolate of gram-positive cocci, which was methicillin sensitive staphylococcus aureus. This is consistent with the study results of Gupta P et al., [31]. Analysis of the resistance profile in this study, showed Acinetobacter to be 95.8% resistant to aminoglycosides, third generation cephalosporins and fluoroquinolones. A 83.68% were resistant to carbapenems. These results were consistent with the study results of Gupta P et al., and Ahmad H et al., [31,39]. One isolate of Acinetobacter was resistant to tigecycline in present study. This denotes an emergence of an alarming situation, wherein there is increased incidence of multidrug resistant and extremely drug resistant microorganisms in the tracheal aspirates which in turn increases their morbidity and mortality [17]. Other gram-negative isolates such as Klebsiella pneumoniae showed 100% resistance to aminoglycosides, third generation cephalosporins, fluoroquinolones and carbapenems. E.coli and Enterobacter showed variable resistance patterns. They were sensitive to broad spectrum antibiotics. This is in concordance with the studies by Gupta P et al., and Malik M et al., [31,37]. This points to the fact that there is an increase in the rates of carbapenem resistance.

The number of islolates from blood cultures were different when compared to ETA samples which predominantly showed single isolate. However, same organism as that of the ETA cultures was isolated from only two blood culture [Table/Fig-9,10]. Other studies have demonstrated a correlation between 0% to 80% [41,42]. ICU infections are most commonly caused by multidrug resistant organisms. CKD population is already immunocompromised and increasingly susceptible for infections as they have associated comorbidities such as diabetes and are at increased risk of exposure to pathogens due to frequent hospital visits. The antibiogram of these isolates should be periodically monitored and the empirical treatment protocols updated regularly. Every center should study the antimicrobial resistance patterns prevalent in their ICU and enforce infection control practices.

This study was a single centre experience. Extrapolation of the results to the entire population may be questionable.

It is clear from the previous and present studies that the resistance pattern of organisms causing VAP has been changing along with time. It varies between geographical areas and also between the study population. CKD population is at increased risk of infection. Off late, there has been an emergence of multidrug resistant strains due to which the treatment of VAP in CKD has become a major challenge. There is an increase in Extended spectrum beta-lactamase (ESBL), Metallo-beta-lactamase (MBL) and AmpC production by Gram-negative Bacteria (GNB) in CKD population. There has been a trend towards an increase in VAP due to preexisting comorbidities in CKD patients and prolonged intubation. Hence, the antibiotic policy needs regular updating for CKD patients. The need of the hour is to prevent VAP in CKD population caused by multidrug resistant strains by formulating appropriate empirical antibiotic policy and also hospital infection control practices.