Hospital-acquired infections (HAIs) are responsible for significantly higher mortality rates, length of stay and hospital costs, being an increasing cause for concern in healthcare worldwide.1,2
In Brazil, as in most developing countries, there are scarce data on the subject, especially concerning pediatrics. Most cases occur in the pediatric intensive care unit (PICU) and are related to invasive devices, such as central venous catheters, mechanical ventilation and urinary catheters; however, HAIs also happen in intermediate care units and general wards.3
The main objective of this study was to describe the epidemiologic, clinical and microbiologic characteristics of HAIs in the PICU, intermediate care unit and general pediatric wards of a tertiary teaching hospital in São Paulo, Brazil.
MATERIALS AND METHODS
Initial data were obtained retrospectively from the hospital’s Infection Control team and included all HAIs in the Department of Pediatrics from January 2012 to December 2016. Additional data were gathered from patients’ medical charts.
The Infection Control team identified each HAI at the time and place of its occurrence, in accordance with the Centers for Disease Control and Prevention criteria.4 Incidence density and device utilization rates were only calculated for the PICU.
Santa Casa de São Paulo is a tertiary teaching hospital in São Paulo, the largest city in Brazil. Its Department of Pediatrics offers healthcare services to children from birth through 16 years of age and comprises a PICU, an intermediate care unit and 2 general pediatric wards.
We excluded all chronic patients who were eligible for home care at the beginning of the studied period but, for social reasons, needed to remain hospitalized.
Antimicrobial resistance profile is routinely performed by the hospital’s own laboratory, in accordance with Clinical and Laboratory Standards Institute guidelines.5 Data from the year 2012 were missing and were not considered in the antimicrobial resistance analysis.
The study was approved by the hospital’s Ethics Committee.
Results are expressed as median and interquartile range or as proportion of the total number of patients or HAI cases. Statistical analysis included Kruskal–Wallis test for the comparison of non-normal distribution variables and χ2 test for comparison of the categorical ones. Alpha was set at 0.05. All statistical analyses were done using Prism 7 software.
There was a total of 718 HAIs in 448 patients. Once exclusion criteria were applied, we were left with 614 HAIs in 369 patients. Forty-three percent occurred in the PICU, 35% in the intermediate care unit and 22% in the general wards.
Incidence density of HAI in the PICU was 14.32 cases per 1000 patients-day.
Table 1 shows data on comparison of gender, age, duration of hospitalization before HAI, underlying conditions, invasive device association with HAI, microorganism identification and lethality.
Of the underlying conditions, chronic neuropathy (18.1%) was the most common, followed by congenital heart disease (14.0%), neoplasia (10.3%) and chronic nephropathy (7.8%).
The most common sites of HAIs were primary bloodstream infection (BSI), pneumonia and urinary tract infection (UTI), responsible for 31%, 20% and 20% of all cases, respectively. When considered separately in each unit or ward, BSI was the most common infection site in the PICU and the general wards, whereas pneumonia was the most common in the intermediate care unit. In the PICU, pneumonia was the third most common infection site. Of all other HAI sites, the most frequent were surgical site (predominantly neurosurgery, 66% of cases), skin and vascular infections, as well as peritonitis and diarrhea.
Both PICU and intermediate care unit had significantly (P < 0.0001) higher rates (92.7% and 85.7%, respectively) of central line-associated BSI compared with that in the general wards (61.4%).
Gram-negative bacteria were predominant in every setting, representing 63%, followed by Gram-positive bacteria, 19%, and fungi, 13%. The remaining 2%–3% were polymicrobial and gastrointestinal rotavirus infections, respectively. Figure 1 shows the percentage of type of microorganism, according to the site of infection and unit/ward.
The PICU UTIs had a significantly higher percentage of fungal etiology than both intermediate care unit and general wards (P< 0.0001). There were no other statistically significant differences among each setting regarding infection etiology.
We found 35% of carbapenem resistance among Gram-negative bacteria. The intermediate care unit had statistically significant lower proportions of carbapenem resistance (18%, P = 0.012) than the PICU (44%) and the general wards (35%).
As for Gram-positive bacteria, 33% of Enterococcus isolates were vancomycin resistant (40% of PICU, 0% of intermediate care unit and 33% of general wards isolates). We also found 50%, 40% and 67% of methicillin-resistant Staphylococus aureus in the PICU, intermediate care unit and general wards, respectively (P = 0.40).
This is one of the largest studies on pediatric HAIs conducted in our country. Most of our HAI cases were diagnosed during hospitalization in the PICU. Intensive care unit patients have critical conditions leading to immunologic compromise and are more frequently submitted to invasive procedures and empiric broad-spectrum antibiotics, all of which are known to be risk factors for HAIs.6
We found that patients in the general wards were significantly older than patients in the other settings, which is probably due to the fact that the younger children tend to develop more severe symptoms, needing intensive or intermediate care.
Invasive device association and HAI lethality were significantly higher in the PICU than in the other settings, which is in accordance with the severity of the PICU cases when compared with other settings.
BSI was the most common HAI in the PICU and general wards, and the third most common in our intermediate care unit. Most BSI cases in all settings were device associated, which points to the need of reviewing our central venous catheter policies in nonintensive care units.
Nosocomial pneumonia was the third most common HAI in our PICU, which was unexpected, given that the use of mechanical ventilation (the most important risk factor for nosocomial pneumonia) is much higher in this setting than in any of the others.7 We also found a low incidence density of ventilator-associated pneumonia (VAP) in our PICU (2.25 cases/1000 ventilator-days), when compared with other developing countries (7.9–18.7),6,8 which suggests probable underdiagnosis of VAP in our PICU.
Gram-negative bacteria were the most common microorganisms found in all settings, which is in accordance with developing countries’ literature on this subject.2,6,8–10 Fungal etiology in UTIs was higher, in proportion, in the PICU, probably because of higher utilization of urinary catheters.
As for antimicrobial resistance, Gram-negative carbapenem resistance proved to be significant in all settings and vancomycin-resistant Enterococcus (VRE) and methicillin-resistant S. aureus prevalence was also noteworthy, especially in the PICU and general wards.
General wards high antimicrobial resistance was unexpected and suggests a demand for improvement in HAI and multidrug-resistant microorganisms’ dissemination prevention strategies.
This study is important to understand HAIs in our hospital, our country and, hopefully, other developing countries. Only with this kind of knowledge may we design better prevention and treatment protocols for HAIs and its complications.
1. Wisplinghoff H, Seifert H, Tallent SM, et al. Nosocomial bloodstream infections in pediatric patients in United States hospitals: epidemiology, clinical features and susceptibilities. Pediatr Infect Dis J. 2003;22:686–691.
2. Becerra MR, Tantaleán JA, Suárez VJ, et al. Epidemiologic surveillance of nosocomial infections
in a Pediatric Intensive Care Unit of a developing country. BMC Pediatr. 2010;10:66.
3. Rosenthal VD, Al-Abdely HM, El-Kholy AA, et al; Remaining authors. International Nosocomial Infection Control
Consortium report, data summary of 50 countries for 2010-2015: device-associated module. Am J Infect Control. 2016;44:1495–1504.
4. Horan TC, Andrus M, Dudeck MA. CDC/NHSN surveillance definition of health care-associated infection and criteria for specific types of infections in the acute care setting. Am J Infect Control. 2008;36:309–332.
6. Arnoni MV, Berezin EN, Martino MD. Risk factors for nosocomial bloodstream infection caused by multidrug resistant gram-negative bacilli in pediatrics
. Braz J Infect Dis. 2007;11:267–271.
7. Elward AM, Warren DK, Fraser VJ. Ventilator-associated pneumonia in pediatric intensive care unit patients: risk factors and outcomes. Pediatrics
8. Lopes JMM, Tonelli E, Lamounier JA, et al. Prospective surveillance applying the National Nosocomial Infection Surveillance methods in a Brazilian pediatric public hospital. Am J Infect Control. 2002;30:1–7.
9. Abramczyk ML, Carvalho WB, Carvalho ES, et al. Nosocomial infection in a pediatric intensive care unit in a developing country. Braz J Infect Dis. 2003;7:375–380.
10. Porto JP, Mantese OC, Arantes A, et al. Nosocomial infections
in a pediatric intensive care unit of a developing country: NHSN surveillance. Rev Soc Bras Med Trop. 2012;45:475–479.