The Polish Prevalence of Infection in Intensive Care ( PPIC ) : A one-day point prevalence multicenter study

Results. Among 205 ICU patients (193 adults and 12 children), 134 infections were found in 101 patients (99/193 adults (51.30%) and 2/12 children (16.70%)), and bacterial colonization in 19/205 (9.3%) patients. In 66.42% of the cases, more than 1site of infection was diagnosed. On the day of the study, 75.40% of the diagnosed infections had positive microbiological results. The most frequent were respiratory tract infections (53.73%), wound infections (18.65%) and bloodstream infections (14.92%). Most of the infections (64.10%) were caused by Gram-negative bacteria (GN), followed by Gram-positive bacteria (GP; 31.80%) and fungi (4.10%). The most frequently reported GN microorganisms were Enterobacteriaceae (44.7%). Methicillinresistant Staphylococcus aureus (MRSA) infections were found in 8.80% of the patients. Antibiotics were administered to 75.60% of the adult patients, in 69.20% as targeted treatment. Mechanical ventilation, central vein catheterization and urinary bladder catheterization were used in 67.80%, 85.85% and 94.63% of the patients, respectively.


Introduction
Infections in critically ill patients are the main reasons for a lack of therapeutic success and increased mortality in intensive care units (ICUs) all over the world.5][6][7][8] Among all these studies, one-day prevalence studies are favored because they can be carried out quickly and easily in different medical centers.The prototype for this type of analysis was the European Prevalence of Infection in Intensive Care (EPIC) Study. 4here have been no large studies of this type focused on ICU patients either in Poland or in Eastern and Central Europe.The objective of our research was a one-day study of the prevalence of infections in ICUs in Warsaw and the Mazovian region of Poland (about 7 million citizens).

Study population and data collection
Our study was a questionnaire-based survey analysis of the epidemiological status of critically ill patients who were hospitalized in ICUs on Tuesday, June 25, 2014.The study covered 205 patients in 28 ICUs in Poland.The study protocol was approved by the institutional Bioethics Committee and performed in accordance with the Declaration of Helsinki.We asked 15 questions about infections and pathogens "occurring in the ICU", and about all other aspects of therapy important to our analysis.The questionnaire was sent to 33 ICUs in 500-to 1000-bed university and municipal hospitals, as well as to smaller district hospitals (up to 250 beds).Two large pediatric hospitals (PD) were also included in our study.Of the 33 ICUs that received the questionnaire, 85% completed the questionnaires and were included in the study.The patient characteristics are shown in Table 1.
We performed global epidemiological analyses in 28 ICUs (205 patients) as well as detailed analyses in Warszawa (WA -11 hospitals), large Mazovian provincial hospitals (MPH -5 hospitals) and district hospitals (DH -12 hospitals).Additionally, we compared some results of our study to the European Prevalence of Infection in Intensive Care (EPIC) and EPIC II studies. 4,5

Definitions and diagnostic methods
Infections were diagnosed based on the criteria of the Centres for Disease Control and Prevention's National Healthcare Safety Network (CDC/NHSN) and the European Centre for Disease Control (ECDC). 9All materials submitted for microbiological analysis were sampled and assessed qualitatively and quantitatively according to accepted standards.The susceptibility of microorganisms was determined in accordance with the recommendations of the European Committee on Antimicrobial Susceptibility Testing (EUCAST). 10Colonization was defined when microorganisms were found at a normally sterile site on the patient, without clinical or laboratory signs of infection.Patients were considered surgical if emergency surgery was performed immediately before admission or if elective surgery was performed within 1 month before admission.All other patients − e.g. with respiratory, cardiac or renal insufficiency − were considered medical.The incidence of infections was calculated as the number of patients with infections per 100 hospitalized patients.

Statistical analysis
The statistical analyses were performed using STA-TISTICA software v. 10 (StatSoft Inc., Tulsa, USA).Descriptive statistics were computed for all study variables.Discrete variables are expressed as counts (percentages) and mean ± standard deviation (SD).The data was analyzed using the χ 2 test or χ 2 test with Yates's correction, as appropriate.P-value <0.05 was considered statistically significant.
The prevalence of infections and primary sites of infections in the different types of hospital are shown in Table 2.
Mechanical ventilation, central venous catheters and urinary catheters are considered risk factors for infections in critically ill patients.These factors are shown in Table 3.

Discussion
One-day point prevalence studies of infections in ICUs have been performed mainly in the USA, North European countries and Australia. 4,5,113][14][15] A one-day point prevalence study (PPS) and incidence study in Poland evaluated the epidemiology of infections in Polish long-term care facilities. 16he first Polish one-day PPS on infections in ICUs was performed in 59 ICUs in 1994. 17The most important differences in comparison to this study include the predomination of GP bacteria in the earlier study, which were isolated in 61.8% of the patients (of these, 67.0% were MRSA species).In our study, the percentage of GN bacteria was nearly double in comparison with the findings of the earlier Polish survey (35.2%);Candida infections were also twice more prevalent (2.6%). 17ecause Poland lacks a history of one-day infection PPS performed in ICUs, we could compare our data only to the one-year prevalence study of Polish ICUs, which analyzed 1,043 critically ill patients with sepsis. 18In this earlier study, patients with GN bacterial sepsis were less frequent than in our study (48.0%vs 64.1%; P. aeruginosa 14.2% vs 4.1%; A. baumannii 15.3% vs 5.3%).The main site of infection found in the previous study was the abdominal cavity (47.0%), which is a contrast to the present study, where respiratory tract infections were predominant. 184][15] However, in other studies (mainly analyzing HAIs), bloodstream infections were predominant. 12,19e analyzed data from 28 Polish ICUs, whereas the EPIC study analyzed data from 1,417 European ICUs; the EPIC II study analyzed data from 1,265 ICUs in North and South America, Western Europe, Asia, Oceania, Australia, and Africa.4,5 The incidence of infections in our study was comparable both to EPIC (44.80%) and EPIC II (51.40%).The incidence of lung infections in our study was also similar to EPIC II (36.30% vs 32.60%).Nevertheless, in our study, respiratory tract infections constituted 53.03% of all infections, in comparison to 63.50% in EPIC II.The incidence of blood infections in our study was also similar to the EPIC II study (10.40% vs 7.80%), as was the incidence of UTIs (8.80% vs 7.30%).We did not find any difference in the incidences of respiratory tract infections and bloodstream infections when we compared our study to the EPIC study.The incidence of UTIs in our study was higher than in the EPIC study (8.80% vs 3.62%). 4,5A comparative analysis of these 3 PPSs is shown in Table 4. Data from the European Centre for Disease Control and Prevention point prevalence survey showed that the prevalence of HAIs in pediatric ICUs was 15.5%.19 That was supported by our observations; nevertheless, our study showed lung infections as the most common, while in the ECDC point prevalence survey, bloodstream infections were the most common type of infection (45.0%).19 The EPIC and EPIC II studies reported the incidence of infections caused by GN bacteria as 32.0% and 62.0%, respectively.4,5 Only the EPIC II microbiological results were consistent with our findings, in which GN bacterial infections were predominant.The main GN pathogens in our study were members of the Enterobacteriaceae family.This is similar to other studies, such as those from Brazil (33.8%) and Japan (27.6%).13,14 A high number of infections caused by A. baumannii, as well as resistance to many groups of antibiotics among bacteria, was noted in studies by Weiner et al. 20 and Harris et al.21 We did not observe this phenomenon in our study.The relatively small number of A. baumannii infections in our observations may result from sample size; Weiner et al. analyzed information from 4,515 hospitals.Moreover, our data was collected more than 3 years ago, when the number of such infections was lower in comparison to the present day.In addition, the number of infections caused by A. baumannii may be a picture of the epidemiological situation only on the particular day the study was performed.
The frequency of MRSA infections was lower in our study in comparison to the results of one-day multicenter PPS from Turkey (18.20%) and Brazil (16.90%) as well as the EPIC study (20.00%), and was similar to the frequency observed in EPIC II (10.20%). 4,5,13,15We found no infections caused by colistin-resistant A. baumannii, although the literature includes data on A. baumannii resistance to colistin amounting to 2.95%. 22The low percentage of infections caused by GN pathogens that produce ESBL in our study was similar to the observations of Coque et al. 23 The rate of C. difficile infections (only 1 isolated pathogen) was lower in our study than in Bartlett's work. 24This may result from the methodology of our study.We decided to analyze only C. difficile infections confirmed with microbiological tests.Moreover, according to data of Kübler et al., 18 metronidazole is administered to septic patients in Polish ICUs very often because about half of the infections in the critically ill originate in the abdominal cavity.Metronidazole is effective against C. difficile, so it may be the reason such infections were not noted in this oneday study.
In summary, the most effective way to control infection problems in ICUs is to strictly follow antiseptic rules (hand hygiene, the use of alcohol-based hand rub solution and HAI monitoring) and to assess compliance with protocols related to these infections that are being implemented in Polish hospitals. 6,25,26ur study had several limitations.First, the analysis could have been affected by the respondents' level of carefulness in completing the questionnaires.Second, our study included multiple hospitals, but the ICUs were situated within 1 geographic area, and the epidemiology of infections in the rest of Poland may not be the same.Third, it is quite noticeable that the data on pediatric patients is rather small: there were 12 patients, and only 2 of them had infections.This is not a representative sample, and all the comparisons involving this group should be considered very cautiously.Fourth, preventive methods could have influenced the rate and epidemiology of HAIs, and this factor was not analyzed.Fifth, we did not precisely analyze the origin of infections, so some could be community-acquired and some hospital-acquired.Also, the use of medical devices may promote infections in hospitalized patients.However, because of the methodology of the study and the nature of the data received, it was not possible to perform analyses that would take these factors into account.Sixth, we did not analyze survival because we completed our observations within 24 h.Finally, some limitations resulted from the methodology of the study; nevertheless, the small number of published studies with the same methodology indicates that there is limited research in this field and shows that our analysis is very important for this part of Europe.

Table 1 .
Patient characteristics

Table 2 .
Prevalence of infections in different types of hospitals Data is presented as number of patients and percentage value.WH -Warsaw hospitals; MPH -Mazovian provincial hospitals; DH -district hospitals; PD -pediatric departments.

Table 3 .
Device utilization ratios in different types of hospitals

Table 4 .
Comparative analysis of point prevalence infection studies PPIC vs EPIC; p* -PPIC vs EPIC II; PPIC -Polish Prevalence of Infection in Intensive Care; EPIC -European Prevalence of Infection in Intensive Care.