ARTÍCULO ORIGINAL / ARTIGO ORIGINAL
Nosocomial infections in two hospitals
in Uberlandia, Brazil

Infecções hospitalares em dois hospitais em Uberlândia, Brasil
Ânderson Silveira Duque1 - Ana Flávia Ferreira2 - Renata Cristina Cezário3 - Paulo Pinto Gontijo Filho4

1Medicine Student, Universidade Federal de Uberlândia/UFU (Uberlândia Federal University), MG, Brazil.
2Graduate in Biological Sciences, Universidade Federal de Uberlândia, MG, Brazil.
3Ph.D. Researcher in Immunology and Parasitology, Universidade Federal de Uberlândia.
4Ph.D, Head Professor, Clinics Hospital, Microbiology Laboratory, Universidade Federal de Uberlândia, Uberlândia, MG, Brazil.

Research developed in the Microbiology Laboratory - Immunology, Microbiology and Parasitology Area -
Universidade Federal de Uberlândia, MG, Brazil.

Rev Panam Infectol 2007;9(4):14-18

Conflicto de intereses: ninguno

Recibido en 30/1/2007.
Aceptado para publicación en 10/9/2007
Resumo
Introdução: Infecções hospitalares (IHs) estão associadas com um aumento na morbidade e mortalidade, hospitalização prolongada e custos. Objetivo: Analisar em dois hospitais (universitário e privado) as diferenças entre infecções hospitalares e fatores de risco em subgrupos (clínicos, cirúrgicos e críticos) de pacientes e a colonização por Staphylococcus spp. e bacilos Gram-negativos multirresistentes aos antibióticos. Metodologia: Foi realizado um estudo prospectivo incluindo pacientes de diferentes unidades em dois hospitais de Uberlândia, um público, universitário e outro privado, no período de maio/2003 a maio/2004. Os espécimes clínicos provenientes de pacientes foram coletados com swabs de narina, boca e cavidade anal e subcultivados em ágar manitol salgado com 6μg/mL de oxacilina e ágar MacConkey com 2μg/mL de ceftazidima. Resultados: As taxas de prevalência de IH foram altas (20,87%), com predomínio em pacientes críticos (~40%), porém sem diferença significativa entre os dois hospitais. As infecções mais freqüentes foram: infecções urinária (36,84%) e de sitio cirúrgico (31,58%). Os fatores de risco significativos foram: hospitalização por mais de sete dias, presença de dois ou mais procedimentos invasivos e uso de antimicrobianos. A principal diferença entre os dois hospitais foram as taxas de cirurgias eletivas e não limpas, e da prescrição de antibióticos. A taxa de colonização por Staphylococcus spp. (12%) e bacilos Gram-negativos multirresistentes (19%) foram semelhantes em ambas as instituições; a distribuição dos fenótipos de resistência foi mais uniforme nos subgrupos clínicos, cirúrgico, críticos do hospital universitário do que no hospital particular, onde foram mais altas na UTI. Conclusão: Estudos de prevalência podem favorecer informações quanto à real situação das infecções hospitalares, conduzindo à implementação de medidas de prevenção e controle de IHs.

Palavras-chave: Infecção hospitalar, fatores de risco, monitoramento, epidemiologia.

Abstract
Introduction: Nosocomial infections (Nis) are associated with an increase in morbidity and mortality rates, extended hospitalization and costs. Objective: Analysis of the differences in two hospitals between nosocomial infections and risk factors in subgroups (clinical, surgical and critical subgroups) of patients and colonization by Staphylococcus spp., and Gram-negative bacilli, multiresistant to antibiotics. Methodology: A prospective study was carried out, comprising patients of different wards in two hospitals, a public one and a private one, in Uberlândia city, Minas Gerais State, from May/2003 through May/2004. Clinical specimens from patients were collected with nostrils, mouth and rectal cavity swabs, and subcultivated in agar Manitol Salgado com 6 µg/Ml of oxacylin and MacKonckey agar with 2 µg/mL of ceftzidime. Results: The NI prevalence rates were high (20.87%) with predominance in critical patients (~40%); however, no significant difference between the two hospitals was found. The most frequent types of infections were urinary infection (36.84%) and surgical site infections (31.58%). Significant risk factors were hospitalization time longer than seven days, the presence of one or more invasive procedures, and the use of antibiotics. The major difference between the two hospitals was detected in the rates of elective and nonclean surgeries and prescription of antibiotics. The rate of colonization by Staphylococcus spp. (12%) and Gram-negative bacilli (19%) was similar in both health institutions; the distribution of resistance phenotypes was more uniform in the clinical, surgical and critical subgroups of the university hospital than in the private hospital, where their higher incidence occurred in the ICU. Conclusion: Studies of prevalence may favor information as to the real status of nosocomial infections, leading to the implementation of NI prevention and control measures.
Key words: NIs, risk factors, monitoring, epidemiology.

Introduction
Nosocomial Infections (NIs) are associated with an increase in morbidity, mortality, more extended hospitalization and significant economic costs. Surveillance of NIs has been widely accepted throughout the world as a primary step toward prevention of NIs.(1) In developed countries NIs represent from 5% to 10% of total hospitalizations.(2) In the US, estimates range from 25,000 to 100,000 deaths per year as a result of NIs, which represent a yearly burden of over 7.5 billion dollars.(3,4) In Brazil, this scenario is even worse because of the lower availability of both human and financial sources.(5,6)

The Brazilian hospital network is extremely heterogeneous as far as number of beds, services available, and patients are concerned. We may highlight differences in terms of macro-regions, large and small sized hospitals, and the nature of hospitals. All these differences are eventually reflected at rates and characteristics of NIs.(7) Larger hospitals tend to present higher rates of NIs.(8,9)

This study aims at analysing differences between NIs and risk factors for these types of infections, by considering subgroups of clinical, surgical, and critical patients in two hospitals: a public university hospital, and a private hospital, both in Uberlândia, Minas Gerais State, Brazil.

Methods
Patients and Hospital: Colonization/infection by multidrug-resistant bacteria in patients hospitalized in surgical unit, clinical-medical unit and adults intensive care unit critical (ICU), was studied. The patients were admitted both in Hospital Santa Genoveva, a private (300-bed) institution, the University Hospital, of the Faculty of Medicine, Uberlândia, Minas Gerais State, Brazil, a teaching hospital with 500-bed capacity also offering tertiary care. Study design: Throughout the study three “cross-sectional” repeated prevalence investigations were carried out within four-month intervals, from May/2003 to May/2004. Clinical and demographic information about each patient, including intrinsic (age and length of hospitalization) and extrinsic risk factors (surgery, invasive devices and use of antimicrobial agents) was obtained from each patient’s record. It was used case-control study where the case represented patients with clinic and microbiologic diagnostic of nosocomial infection and the control patients hospitalized in the same period without diagnostic of infection. Patients admitted in the day the inquiry was carried out were excluded. Nosocomial infection is defined as an infection that is neither present nor incubated when the patient is admitted to a hospital or any other health care facility; 48-72 hours after admission is generally deemed indicative of nosocomial infection.(10) Research on patients colonized/infected by multidrug resistance bacteria: Oropharyngeal, nasal and rectal swab samples were transported within tubes containing 1 mL of Trypticase Soy Broth (TSB) and about 2 mL of urine were collected by aspiration with a syringe of patients with vesical catheter. This procedure involved all patients hospitalized at the moment the inquiries were carried out. Microbiological techniques: Colonization or infection was researched by specimens of nostrils, mouth and rectal. Mannitol salt agar containing 6 µg/mL for oxacylin Staphylococcus aureus and coagulase-negative Staphylococcus, and MacConkey agar, with 2 µg/mL of ceftazidime, for Gram-negative bacilli isolations. Quantitatively analyses were performed with Trypticase Soy agar (TSA), CLED and Sabouraud agar for urine and cultures with growth equal or superior to 104 UFC/mL were considered positive.(11) Statistical analysis: The univariate statistical analyses were done with Epi-Info software, version 6.04, and values of p<0.05 were considered to indicate statistical significance. Categorical variables were assessed with x2 and Fisher’s Exact test, and Student’s t test for continuous comparative data. Ethical approval: Ethical approval to conduct the study was obtained from the Institutional Ethics Committees of the participating hospitals.

Results
Overall, 233 patients from the public institution and 112 in the private institution, distributed by medical clinics (127 and 36), surgery (71 and 52), and ICU (35 and 24) were included in the research. The prevalence rate of NI was high in both hospitals (20.90%), (public and private) corresponding to 21.89% and 18.75%, respectively. ICUs of both hospitals were the places with the highest prevalence rates of NIs in comparison with all the sectors analyzed (table 1).



Predominant infectious syndromes in both hospitals were: urinary tract infections, surgical site infection, pneumonia, and sepsis (table 2).



The patient’s characteristics and variables that were examined as possible risk factors with corresponding matched univariate analysis are displayed in table 3. The presence of 3 or more invasive procedures present in the private hospital (OR, 8.70; CI95, 1.86-42.86) and in the university hospital (OR, 6.26; CI95, 2.77-14.23), showed the highest risk of NI (HCU: OR 6.26; CI95 2.77-14.23, p<0.05; HSG: OR 8.70; CI95 1.86-42.86, p<0.05).The most patients with Vesical Probe, Central Vascular Catheter and/or Ventilatory Prosthesis presented significance for the development of NI (p<0.05). In the university hospital, about half (49.02%) of the patients with NI had VP.



Surgeries were classified as clean or not-clean, and as elective or non-elective, with a higher number of not-clean and non-elective in the public hospital, which corresponded, respectively, to 57.14% and 51.19% (p<0.05), while in HSG most of the surgeries were clean and elective, reaching, respectively, 72.88% and 84.74% (table 4).



In HSG (Hospital Santa Genoveva, Uberlândia), the use of antimicrobial drugs was more frequent (61.60%) than in HCU (Hospital das Clínicas de Uberlândia) (40.77%), highlighting the prophylactic indication (53.62%) in the first one (table 4). The frequency of patients colonized by Gram-positive cocci (methicillin resistant Staphylococcus aureus and coagulase-negative Staphylococcus) in both hospitals was similar. However, in HCU the amount of colonized patients was similarly distributed over the three components. In terms of multi-resistant Gram-negative bacilli, while without significancy, colonized patients were usually seen in the ICU, mainly in the private hospital (table 4).

Discussion
In Brazil, information about NIs is scarce.(12) For this reason, studies like this one are very important, aiming at evaluating differences between two general hospitals, a private one and a public one – a teaching hospital - representative of the Brazilian hospital reality. In the only inquiry carried out in the country, involving general hospitals in several macro-regions,(7) a private hospital was included, in spite of this kind of institution being responsible for about 70% of the available hospital beds.(13) In the present study, the total rate of NI found was high, approximately 20%, when compared to countries such as the US, with different frequencies in clinical (11.25% in both hospitals), surgical (17.55%) and critical (47.97%) patients, but without significant difference between institutions (p>0.05).

Among the most frequent NIs, in order of frequency, it is possible to highlight the urinary tract infections, surgical site infections, pneumonias, and blood stream infections. In Brazil, on the other hand, pneumonias are usually referred to as the most frequent event and they are responsible for about 30% of occurrences of NI, followed by surgical site infections with 15.6% of NIs in general hospitals.(7) One of the reasons for this discrepancy is the low use of microbiological criteria to define NIs, which are mainly defined by clinical and radiological data, in cases of pneumonia, in part due to the lack of laboratories in the majority of hospitals, in Brazil.(14) NIs, especially pneumonias found in critical patients, are frequently difficult to be diagnosed when the professional bases his/her diagnostic on clinical criteria only(15) and their use has poor specificity, in the face of the several undistinguishable pulmonary pathologies.(16) Additionally, uroculture is needed to confirm bacteriuria since most of them are asymptomatic.(17,18) Moreover, because they usually occur in critical and/or unconscious patients using vesical probe, the diagnostic is not perceived, even when they are symptomatic, if microbiological diagnostic sources are not used.(19)

Factors predisposing to nosocomial infections present variables related to the patients’ susceptibility and to extrinsic variables, from which we may underline invasive procedures as the use of ventilatory prothesis, central vascular catheter, and vesical probe.(9) In developed countries, especially in the USA, the hospitalized population is different from the Brazilian one, with its major portion being represented by elderly people, and with a higher proportion of beds for critical patients.(20) In Brazil, diseases affecting people searching seeking admission, especially into public hospitals, are a consequence of the poor social status of the general population, and in general patients are younger, victims of trauma and/or community infectious diseases. Another point to be considered is the great diversity of hospitals, which differ greatly in terms of their nature, whether public or private, in terms of patients, and services available.(7) This study confirmed these differences in the two major hospitals in a medium sized town with about half a million inhabitants, showing in the private hospital patients admitted for lower grade morbidities, which is reflected by a shorter hospitalization period, under 3 days (65%), while in the public institution, the proportion of this kind of hospitalization period was around 20%. On the other hand, the analysis of invasive procedures as risk factors, in both hospitals, evidenced its importance, especially as to the use of the vesical probe and bacteriuria. Another key aspect observed in this investigation was the difference among surgical procedures performed in both hospitals. While in HSG, clean and elective surgeries prevailed (>70%), in HCU nonclean and nonelective surgeries were more frequent (>50%), and these are more associated with NI.(21)

However, surgery site infection rates in clean surgeries were very high and similar (~29%) in both hospitals.
Nowadays, one of the major medical-hospital issues is the increasingly expressive frequency of microorganisms resistant and multidrug-resistant to antibiotics. This is a global reality representing a challenge, since infections by these organisms are more severe and have higher

costs.(22) The use of antimicrobial drugs in Brazilian hospitals is usually intensive (about 48% in this investigation) and poorly judicious,(11) besides being predominantly empiric. Moreover, the use of antibiotic-prophylaxis is excessively used including clean surgeries,(23) as this investigation has showed, in the private hospital (>50% in HSG). The abusive use, be it therapeutic or prophylactic, be it associated with a longer hospitalization period, and with invasive procedures are factors of risk for infections by resistant bacteria.(22,24,25,26)

In the impossibility of obtaining clinical isolates, mouth, nostrils, and intestine microbiota were analyzed, since they are major sites of hospital pathogens. This research, considering both hospitals, showed a proportion of about 12% patients colonized by Staphylococcus sp. and 15% by multidrug-resistant aerobic Gram-negative bacilli. The greatest proportion of resistant Staphylococcus sp. was isolated from critical patients (22%), but in the university hospital, this microorganism was more disseminated, since it was detected in similar frequency from the three components.

Summing up, in spite of divergences between patients with and without NI, the differences between both hospitals were not statistically significant. The result of concomitantly performing active surveillance and urine culture from patients with vesical probe was high frequencies of NI in clinical, surgical, and critical patients, in both hospitals.

References
1. Jarvis WR. Epidemiology of nosocomial infections in pediatric patients. Pediat Infect Dis J. 1998;6:344-351.
1. 2- Haley RW. The scientific basis for using surveillance and risk factor data to reduce nosocomial infection rates. J Hosp. Infect 1995;30:3-14.
2. Bergogne-Bérézin L. Les infeccions nosocomiales: nouveaux agents, incidence, prevéntion. Presse Med 1995;24:89-97.
3. Smith SD, Doebbeling BM. Costs of nosocomial infections. Curr Opin Infec Dis.1996;9:286-290.
4. Wenzel RP. Epidemiology of hospital-acquired infection. In: Ballows A (ed): Manual of Clinical Microbiology. Washington DC: American Society for Microbiology, Washington 1994. p. 147-150.
5. Wey S. Comentarios generales sobre el problema de las infecciones hospitalarias en el hospital Sao Paulo, Brasil. In: Wey S (ed). Desarrollo y fortalecimiento de los sistemas locales de salud. Pan American Health Organization,
Washington 1991, p. 74-76.
6. Nettleman MD. Global aspects of infection control. Infect Control Hosp Epidemiol. 1993;14:646-648.
7. Prade SS. Estudo brasileiro da magnitude das infecções hospitalares em hospitais terciários. Revista de Controle de Infecção Hospitalar 1995;2:11-23.
8. Lizioli A, Privitera G, Alliata E, Banfi EMA, Boselli L, Panceri ML et al. Prevalence of nosocomial infections in Italy: result from the Lombardy survey in 2000. J Hosp Infect. 2003;54(2):141-148.
9. Klavs I, Bufon T, Skerl M, Grgic-Vited M, Zupanc TL, Dolinsek M et al. Prevalence of and risk factors for hospital-acquired infections in Slovenia – results of the first national survey, 2001. J Hosp Infect. 2003;54(2):149-157.
10. Garner JS, Jarvis WR, Emori TG, Horan TC, Hughes JM. CDC definitions for nosocomial infections. Am J Infect Control 1988;16:128-149.
11. Isenberg HD. Aerobic Bacteriology. In: Isenberg HD (ed). Clinical Microbiology Procedures Handbook. Washington, DC: American Society For Microbiology, Washington 1992, p 101-202.
12. Iney SB. Infection control in a country with annual inflation of 3.6%. Infect Control Hosp Epidemiol. 1995;16(3):175-178.
13. IBGE, 2002. Estatísticas da Saúde: Assistência Médico-Sanitária 2002. Available:<http://www.ibge.gov.br/home/estatística/população/condicaodevida/mas/defaulttab.shtm>: Accessed in: October 24, 2005. Weinstein RA. Nosocomial infection update. Emerg Inf Dis. 1998;4:416-240.
14. Gontijo Filho PP. Definições de infecções hospitalares sem a utilização de critérios microbiológicos e sua conse­qüência na vigilância epidemiológica no Brasil. News Lab. 2002;53:120-124.
15. Mayhall CG. Nosocomial urinary tract infection. In: Mayhall CG (4 ed): Hospital Epidemiology and Infection Control. Lippincott Williams and Wilkins, Philadelphia 2004, p.139-153.
16. Alcón A, Fabregas N, Torres A. Hospital acquired pneumonia: etiologic considerations. Infect Dis Clin N Am. 2003;17:679-695.
17. Johnson JR. Microbial virulence determinants and the pathogenesis of urinary tract infection. Infect Dis Clin N Am. 2003;17:261-278.
18. Mayhall CG. Ventilator-associated pneumonia or not? Contemporary diagnosis. Emerg Infec Dis. 2001;7(2):2000-2004.
19. Wagenlehner FMG, Naber KG. Hospital acquired urinary tract infection. J Hosp Infect. 2000;46:165-170.
20. Weinstein RA. Nosocomial infection update. Emerg Inf Dis 1998;4:416-420.
21. Cruse PJE. Surgical wound infection. In: Gorbach SL, Bartlett JG, Blacklow NB (ed). Infectious Diseases. WB Sounders Company, Philadelphia 1992, p. 758-764.
22. Kollef MH. Hospital-acquired pneumonia and de-escalation of antimicrobial treatment. Crit Care Med. 2001;29(7):1473-1475.
23. Couto HG, Lima HV, Leles CCV, Gontijo Filho PP. Critérios microbiológicos no diagnóstico de infecções hospitalares e de prescrição de antimicrobianos no Hospital de Clínicas da UFU. XX Congresso Brasileiro de Microbiologia, 1999, Salvador. Bahia, p.60.
24. Richards MJ, Edwards JR, Culver DH, Gaynes RP. Nosocomial infections in medical intensive care units in the United States. National Nosocomial Infections Surveillance System. Crit Care Med. 1999;27:887-892.
25. 25.Trovillet JL, Chastre J, Vuagnat A, Joly-Guillou ML, Combaux D, Dombret MC, Gibert C. Ventilator-associated pneumonia caused by potentially drug-resistant bacteria. Am J Respir Crit Care Med. 1998;157:531-539.
26. Kollef MH. The importance of antimicrobial resistance in hospital-acquired and ventilator–associated pneumonia. Curr Anas Crit Care 2005;16:209-219.

Correspondence:
Dra. Renata Cristina Cezário
Laboratório de Microbiologia - ARIM - Av. Pará, 1.720 - Umuarama, Uberlândia, MG, Brasil.
e-mail: cezariorenata@yahoo.com