This was a hospital based, prospective study conducted during the period of September 2012 to June 2014 at the Department of Microbiology of a tertiary care centre in Puducherry, India. All the clinical specimen received from paediatric outpatients and inpatients were included in the study and data were collected from those yielding growth on culture. Repeat isolates from the same patients and from the same site/specimen were excluded. Ethical clearance was obtained from the Institute Ethics Committee for the study. Informed consent was obtained from parents. Standard methods were followed for sample processing, culture, and antibiotic susceptibility testing [9].

Blood cultures were done by inoculating 2-5 mL of blood, collected aseptically from median cubital vein into 20 mL of Brain Heart Infusion (BHI) broth. In case of neonates 1 mL of blood was inoculated into 5 mL of broth. It was incubated at 37°C for 18 hours and then subcultured onto Blood Agar (BA), Chocolate Agar (CA) and MacConkey (MAC) agar. Urine culture was done by standard loop technique, a semiquantitative method on Cystine Lactose Electrolyte Deficient (CLED) agar. Body fluids were centrifuged at 3000 rpm for 20-30 minutes and the sediment was inoculated onto BA, CA and MAC agar. Exudate samples were inoculated on BA and MAC. All plates were incubated at 37°C for 18 hours in 5-10% CO₂ jar. Those specimens which yielded growth was included in the analysis and the details were entered in Microsoft (MS) excel sheet. The antibiotic susceptibility pattern was analysed and expressed in percentages. All the clinically significant isolates were tested for susceptibility for various antibiotics (Hi-Media Laboratories Ltd, Mumbai, India) by Kirby-Bauer disc diffusion test as per CLSI 2012 guidelines [10]. Antibiotic panels tested for different organisms were as per CLSI guidelines. An organism was considered as ESBL producer when the zone size for ceftriaxone, cefotaxime and ceftazidime was ≤19 mm, 22 mm and 17 mm respectively [10].

For Quality Control of antibiotic susceptibility test, Staphylococcus aureus ATCC^®25923^™, Escherichia coli ATCC^®25922^™ and Pseudomonas aeruginosa ATCC^®27853^™ were used [10].

An organism was considered MDROs when it was resistant to more than or equal to three (≥3) drugs of different classes [11,12]. The common pathogens from various clinical samples and their susceptibility pattern were analysed. The relative occurrence of MDRO from various clinical specimens was also studied.

The study was an observational study and the resistance patterns of organisms were calculated as percentages and proportions using MS Excel 2010 version.

A total of 5443 samples were received, from 3881 patients of paediatric age group, including urine (n=2826), blood (n=2208), exudates (n=52), respiratory specimens (n=141), stool (n=167) and body fluids (n=49) and 561 bacteria were isolated from these samples with the majority being from urine 419 (74.7%) and blood 86 (15.3%). The remaining bacteria were isolated from other specimen as shown in [Table/Fig-1]. Out of 561 isolates, 278 (49.5%) were MDRO. Out of the 419 bacteria isolated from urine, 238 (56.8%) were MDRO. Of the 86 isolates from blood, 24 (27.9%) were MDR. The details are as shown in [Table/Fig-1]. E. coli and Klebsiella spp. were the common pathogens isolated as shown in [Table/Fig-2]. Nearly 86% of the MDRO were isolated from urine culture as shown in [Table/Fig-3] and E. coli was the most common MDRO. The sensitivity pattern of these MDRO is as shown in [Table/Fig-4] and [Table/Fig-5]. Out of 561 culture positives, 39 were from neonates, 105 from one month to one year age group (infants) and 417 from 1-12 years age group.

Infections due to antibiotic resistant bacteria are of grave concern. Drug resistant bacteria are not only seen in hospital setting but also in the community. The increasing prevalence of drug resistant pathogens has become a global challenge, particularly among developing and under developed nations. Added to this are the poor infection control practices and abuse of antibiotics favouring the resistant pathogens in causing infections.

This study was undertaken to know the prevalence of MDROs in paediatric infections. Nearly 68% of the MDROs are from inpatients and 32% from outpatients suggesting the prevalence of resistant pathogens in community as well. In present study, urine and blood were the most commonly received clinical specimen and MDROs were also most commonly isolated from these specimens. About 561 isolates were obtained from blood, urine, exudate, respiratory, stool and body fluids, out of which 278 (49.5%) were MDROs. Nearly 57% of the isolates from urine and 28% of the isolates from blood were MDROs. Of the 278, 238 MDROs (86%) were isolated from urine. E. coli, Klebsiella spp. and Enterococcus spp. were the common MDROs in our study, where majority of these were isolated from urine samples.

Studies from various parts of India had similar findings with E. coli, Klebsiella spp. and Enterococcus spp. being the most common uropathogens in children [3,12,13]. The susceptibility of enterobacteriaceae in a study was 96%, 51% and 58% towards imipenem, piperacillin and cefoperazone-sulbactam. Enterococcus spp. showed a resistance of 88%, 72% and 86% towards ciprofloxacin, amoxicillin and High Level Aminoglycoside Resistance (HLAR-G) respectively [3]. A study from Northern India showed that 42% of the Gram negative isolates were ESBL producers [12]. Similarly in the present study, 46% of the Gram negative organisms were ESBL producers where nearly 75% of them were E. coli isolated from urine.

Coagulase Negative Staphylococci (CONS), E. coli and Klebsiella spp. were the common MDR pathogens from blood culture. Salmonella typhi and S. paratyphi A were other significant isolates from blood culture. However, they were sensitive to the antibiotics tested. Enterobacter spp. and Klebsiella pneumoniae were common Gram negative pathogens and CONS and S. aureus were common Gram positive pathogens causing neonatal septicemia in various studies across India [14-17]. These studies showed that Gram negative bacilli were sensitive to ciprofloxacin and amikacin and were highly resistant to third generation cephalosporins. Studies have shown that septicaemia due to drug resistant Gram negative bacilli are important cause for neonatal mortality [18,19].

A study conducted on septicaemic neonates in Bijapur, Karnataka showed that K. pneumoniae was highly resistant to ciprofloxacin (71%), amikacin (67%) and aztreonam (92%) [1].

A study in Taiwan showed that nearly 20% of the bacteremia among neonates in Neonatal ICU (NICU) was caused by MDR Gram negative bacilli. K. pneumoniae was the most common causative agent followed by E. coli. Prior broad spectrum antibiotic therapy and underlying renal disease were some of the risk factors associated [20]. Acinetobacter spp. and K. pneumoniae were the most common MDR pathogens in PICU in a study in Pakistan. These organisms showed high resistance to ceftriaxone, amikacin and piperacillin-tazobactam [5].

A study from Karnataka states that MDR Acinetobacter spp. were also potential pathogens causing septicaemia in neonates [21]. However, in our study we did not encounter MDR Acinetobacter spp. from blood though few were isolated from urine.

A study on neonatal septicaemia in a tertiary care centre in Maharashtra showed that K. pneumoniae and S. aureus were the common pathogens and that 28% of K. pneumoniae and E .coli were ESBL producers. Multidrug resistant pattern was observed in their study and ciprofloxacin and aminoglycosides were found to be useful drugs [4].

The MDR Gram negative bacilli in our study showed good sensitivity to amikacin (73-88%), pipercillin-tazobactam (82-100%) and imipenem (82-100%). Sensitivity to other antimicrobials was poor (<30%). Among the MDR Gram positive cocci, Enterococci were the significant one which showed high sensitivity to ampicillin (73%), chloramphenicol (90%), linezolid (100%), teicoplanin (100%) though 40% of the Enterococci were Vancomycin Resistant Enterococcus (VRE).

Studies have shown that in infections due to MDR bacteria, an appropriate antibiotic combination therapy needs to be prescribed based on the local antibiogram, both for empirical as well as therapeutic purpose [22].

Though the study was done prospectively, clinical correlation and follow up of patients were not feasible. Hence, only laboratory data were analysed. Molecular analysis for resistant isolates has been sent to AIIMS, New Delhi, the results of which are awaited.

This study reveals significant prevalence of MDR pathogens among paediatric population. E. coli, Klebsiella spp. and Enterococcus spp. were the common MDROs encountered in the study, mostly isolated from urine culture. MDROs are present among outpatients as well, warranting judicious use of antibiotics and adequate infection control measures to prevent spread of these potential pathogens.

Financial support: The authors are thankful to Indian Council of Medical Research (ICMR), New Delhi for funding this extramural project.