MDR organisms represent a worldwide threat in ICU hospitalized children; they affect disease infection control and are accompanied with high mortality rates. Treatment cost is correspondingly increased secondary to the prevalence of resistant pathogens requiring more expensive therapies [2]
The prevalence of infections sustained by MDR bacteria in ICU patients varies in different regions of the world. In North America, a study on critically ill patients with pneumonia (DEFINE study) reported a 14.1% rate of MDR infections, while a large study on nosocomial bloodstream infections conducted in 24 ICUs distributed worldwide (EUROBACT study) showed on average a 47.8% MDR rate, including 20.5% and 0.5% of isolated microorganisms with extensively drug-resistant (XDR) and pan-drug-resistant (PDR) patterns, respectively, with major variations between different countries ranging from 8% (Australia) to more than 75–80% (Turkey, Greece, Croatia, Serbia) [9, 10].
In the present study, the incidence of colonization of MDR bacteria among all admitted patients was 26/282 (9.2%) which was comparable to the incidence reported by a study done in three PICUs of one tertiary children’s hospital in Italy 8.72% (79/906) [11], which is significantly lower than that of older reports with incidence of MDROs that ranged from10 to 25% in PICU [12, 13].
Our study conducted on 282 PICU patients, the culture results revealed 68 isolates only 26 (38.2%) were MDROs and the rest 42(61.8%) were non-MDROs. This rate was lower than that reported even in other regions in Egypt as reported in a study done in 2 pediatric ICUs in Pediatric Hospital-Cairo University [14] 98/106 (92.45%) and in Neonatal and Pediatric Intensive Care Units of Beni-Suef University Hospital145/169 (85.8%) [2]. Moreover, our rate was lower than the reported prevalence in King Chulalongkorn Memorial Hospital-Thailand 30/58 (52%) [15]. This major difference can be explained by variations in the sample size, different demographic regions, and inadequate implementation of infection control measures [2].
Possible explanations for the high percentage of MDROs in some PICUs can be extrapolated by abuse of antibiotics in the outpatient settings, treating viral infections with antibiotics, inappropriate dose, and incomplete course of antibiotics [2]. It has been stated that the university/teaching hospitals that usually operate as referral hospitals generally report higher infection rates [16].
In a review of several studies, Cosgrove showed that there is an association between the development of antimicrobial resistance in some microorganisms and the increase in mortality rates, the duration of hospital stay, and the cost of health care. Inadequate therapy or delay in therapy and the presence of underlying disease were thought to be responsible for the adverse outcomes associated with antimicrobial-resistant infections [17].
Similarly, a statistically significant higher mortality was detected in our patients colonized with MDROs 9/26 (34.6%) versus MDROs non-colonized patients 32/256 (12.5%) [P=0.002]. Moreover, MDROs infection has negative significant risk with discharged patients ([OR] 0.269; [95% CI] (0.111–0.656); p = 0.002).
Additionally, patients infected with MDROs did have significantly greater PICU stay than those non-infected [median (IQR), 16.5 (10.7–22), 5 (4–8), P=0.00] and have longer ventilation [median (IQR), 15.5 (10–18), 3 (2–10), P=0.00].
Our results matching the previous studies, which have reported that the patients with multi-drug resistant infections (MDRI) were hospitalized and treated in the ICU for considerably longer stay and had lower survival rates compared to other patient groups [11].
In the present study, the mortality rate was 41/282(14.5%) which was matching that reported in a study done in Beni-Suef university hospital (Egypt) 10/80 (12.5%) [2]. However, higher rates were detected in other developing countries ranged between 10–53.6%. In contrast, very low mortality rates were reported in the developed countries [2].
In our study the incidence of HAI 68/282 (24.1%) which is comparable to pediatric ICUs in Pediatric Hospital-Cairo University106/378 (28%) [16 ]. In Europe, incidence of HAI ranges from 1% in general pediatric wards to 23.6% in PICUs [18], while a nation point prevalence study of PICU in the USA found the incidence to be 11.9% [19].
A study on the Extended Prevalence of Infection in Intensive Care (EPIC) II revealed that 51% of patients were considered to be infected while in ICU. Infections were of respiratory origin in 64% of cases. The most frequent isolated microorganism was Staphylococcus aureus (20.5%). However, the overall predominance was for Gram-negative organisms as a group: 62.2% (Klebsiella spp., Pseudomonas spp., E. coli, Enterobacter spp., and Acinetobacter spp.) [20].
This is in harmony with this study as we detected only MDR gram-negative bacteria (MDR-GNB), MDR Acinetobacter was isolated in half of the cases colonized with MDR strains followed by MDR Klebsiella 12/28 (42.8%), and MDR Pseudomonas was isolated from only 2/28 (7.1%) patients. Like our study, predominance of MDR-GNB was also reported by some studies such as those done in Italy [21], Tunisia[ 22], and the Philippines [22].
Over the years, and as a result of the increasing lack of effective antimicrobial agents against resistant gram-negative microorganisms, MDR gram-positive microorganisms have been much less than MDR gram-negative strains [23].
Nowadays effective treatment against MDR bacteria is few or missing for specific PDR strains. Beta-lactam antibiotics were considered the first-line treatment against many microorganisms and this was due to their high safety profile and broad efficacy, for many decades. However, bacterial production of β lactamase enzymes progressively increased globally making beta-lactams ineffective as first-line treatments for nosocomial infections in many countries of the world. In the last years, the use of Carbapenems as first-line empiric treatments in many critically ill patients grew rapidly leading to a noticeable increase in the incidence of Carbapenem-resistant bacteria, by various mechanisms of resistance [24].
In our PICU, Piperacillin Tazobactam and Colistimethate sodium combination was effective in 13/26 (50%) while Imipenem and Colistimethate sodium was effective in 9/26 (34.6%).
Acinetobacter baumannii (AB) is a frequent cause of nosocomial acquired infection in critically ill patients [24]. In fact, it is considered to be the third microorganism that is responsible for ventilator-associated pneumonia in the European ICU patients, after S. aureus and P. aeruginosa [25]. Although Carbapenems are usually considered the first-line agents for the treatment of severe infections caused by AB, their usage is becoming limited in many areas because of the increasing resistance [23]
These data matching our study as MDR Acinetobacter is responsible for half 14/28 (50%) of the MDROs isolations mainly causing pulmonary infections in the majority of cases 10/32 (71.4%) and was responding to antibiotic combinations primarily to carbapenems (Imipenem) and Colistimethate sodium in 6/26 (42.8%) of patients and to a lesser extent to Piperacillin Tazobactam and Colistimethate sodium combination in 5 (35.7%) patients.
In many European regions, Klebsiella pneumoniae producing Carbapenemase is considered as one of the most common MDR gram-negative microorganisms in critically ill patients [26, 27]..
In this study, MDR Klebsiella pneumoniae was the second most common organism after MDR Acinetobacter baumannii isolated from 12/28 (42.8%) and foremost sensitive to Piperacillin Tazobactam and Colistimethate sodium combination in 8/12 (66.6%) and causing both pulmonary infection in 5/12 (41.6%) and sepsis 4/12 (33.3%).
Pseudomonas aeruginosa (PA) is one of the commonest causes of health-care-associated infections and is responsible for severe bloodstream, pulmonary, urinary tract, and soft tissue infections in ICU cases [28]. On the contrary, MDR Pseudomonas aeruginosa was the least organism isolated in our study in only 2/28 (7.1%) patients.
Our study revealed that the majority of MDROs were isolated from sputum in more than half of the patients 19/32 (59.3%) followed by whole blood in 10/32 (31.2%) and finally urine in 3/32 (9.4%). Conversely, Wang et al. showed that out of the 79 cases of MDRIs, 43 (54.4%) cases were detected in the whole blood, 23 (29.1%) cases in sputum [11], matching several previous studies which detected most of MDR isolates at blood cultures (69.7%) [2]. Similar findings were obtained in other studies in Egypt [29,30,31] and other different countries (including China, Mexico, South Africa, and Kenya) [32,33,34].
Identifying risk factors for infection development caused by multidrug-resistant bacteria can help health care providers prevent nosocomial infections. This is much more important when we consider the slow development of new effective anti-microbial agents and the increasing prevalence of MDRI, especially in the PICU [11].
Regarding risk factors for acquiring nosocomial infections due to resistant organisms in our PICU, multivariate logistic regression analyses showed significant relationships between MDROs and age under 1 year (odds ratio [OR] 2.4554; 95% confidence interval [95% CI] (1.072–5.625); p = 0.043) and underlying pulmonary diseases (OR 2.417; 95% CI (1.014–5.761); p = 0.592).
These risk factors were also reported by another study, which concluded that acquiring nosocomial infections due to resistant organisms in PICU patients is more likely in patients with transplants and those with underlying lung disease [17].
Similarly, age under 2 years along with the length of hospital stay (more than 3 days) was identified as risk factors for acquisition of MDR during hospitalization in PICU in Tunisia [ 22]. Young age as a risk factor might reflect an inherent risk to acquire MDR by environmental contamination [ 22].
However, other risk factors were detected by several studies such as Atta et al. [35], who found that hematologic diseases and healthcare-associated infection had a significant relationship with colonization of multi-drug-resistant gram-negative bacteria (MDR-GNB). In another study in PICU in Italy which concluded the length of PICU stay, the duration of mechanical ventilation > 5 days, parenteral nutrition, coma, urinary catheter indwelling, snd invasive operation, two or more antibiotics use were associated with MDROs [11].