Perceptions and attitudes of the professional staff concerning infection surveillance and control programs in Brazilian hospitals Carla Morales Guerra, MD, MSc,a Monica Parente Ramos, MSc, PhD,a Virginia Zagallo Penna, NR,a Janaina Midori Goto, MD,a Leandro Queiroz Santi, MD,b Valeska de Andrade Stempliuk, MSc, PhD,c Janaina Sallas, MSc,d and Eduardo A. S. Medeiros, MD, PhDa Background: Several countries have conducted studies to assess the status of their infection control programs (ICP) with the objective of improving quality of infection control practices. Methods: To assess the perceptions and attitudes of the health care workers (HCW) concerning ICP in Brazilian hospitals, we conducted a cross-sectional survey using a self-administered online questionnaire during a Web-based course (WBC) on infection control (IC) and antimicrobial resistance (AR). Results: Of 6256 Brazilian HCW registered for the WBC, 1998 were members of infection control committees (ICC) and answered the survey. Eight hundred six (40.4%) respondents said that an ICP was established for more than 10 years in their institutions. Most professionals reported that their hospitals perform microbiologic surveillance targeted at epidemiologically important mul- tidrug-resistant organisms, but the majority underestimated the prevalence of AR. Conclusion: Our survey highlights important information about the perceptions and attitudes of ICC members that may be used to tailor key interventions for implementing effective ICP. It suggests, additionally, that, to achieve countrywide standardized IC mech- anisms in a developing country, authorities should consider the social, cultural, and economical disparities between regions and identify speciﬁc regional needs to make available the resources required to minimize such disparities. Key Words: Health care workers; infection control programs; perceptions; attitudes. Copyright ª 2010 by the Association for Professionals in Infection Control and Epidemiology, Inc. Published by Elsevier Inc. All rights reserved. (Am J Infect Control 2010;38:59-62.) Guidelines issued by international health institutions the objectives of improving quality of care and reducingor professional societies indicate that the basic require- transmission of health care-associated infectionsments for an effective infection control program (ICP) in- (HAI).3-6clude the existence of an infection control committee In Brazil, since the 1980s, the Ministry of Health rec-(ICC); the availability of professionals trained in infection ommends that all hospitals in the country should havecontrol (IC); and the implementation of written proto- an ICP. However, until 1998, some reports observedcols, ongoing surveillance, and educational programs.1 that most Brazilian hospitals did not practice appropri- Since the landmark Study on the Efﬁcacy of Nosoco- ate surveillance.7,8mial Infection Control (SENIC),2 several countries have In the past 5 years, Brazilian authorities have sup-conducted studies to assess the status of their ICP with ported several strategies in an attempt to reach an ef- ﬁcient implementation of the program throughout the whole country’s large territory. Among these mea- sures, the infection control staff of the Federal Univer- From the Federal University of Sao Paulo, Sao Paulo, SP, Brazila; National ˜ ˜ Health Surveillance Agency (ANVISA) Brasilia, DF, Brazilb; Pan American ˜ sity of Sao Paulo, already experienced in e-learning Health Organization (PAHO) Brasilia, DF, Brazilc; and Ministry of Health programs,9 was asked to design a Web-based course in Brazil (MS) Brasilia, DF, Brazil.d (WBC) on infection control (IC) and antimicrobial re- Address correspondence to Carla Morales Guerra, MD, MSc, Rua sistance (AR) targeted at 1000 health care workers ˜ Napoleao de Barros, 690, 2andar, Sao Paulo, SP, 04024-002 - Brazil. (HCW) involved with HAI and AR from different areas E-mail: firstname.lastname@example.org. of Brazil. Conﬂicts of interest: None to report. Taking advantage of this nationwide program, we 0196-6553/$36.00 conducted a survey to assess the perceptions and atti- Copyright ª 2010 by the Association for Professionals in Infection tudes of these HCW concerning infection surveillance Control and Epidemiology, Inc. Published by Elsevier Inc. All rights and control programs in their institutions, with the ob- reserved. jective of describing the characteristics of ICP in doi:10.1016/j.ajic.2009.05.011 Brazil. 59
60 Guerra et al. American Journal of Infection Control February 2010METHODS ICC infrastructure and human resources We conducted a cross-sectional survey with Brazil- Eight hundred six (40.4%) respondents said theirian HCW participating in a WBC on IC and AR orga- ICC had been active for more than 10 years, whereasnized by the Infectious Diseases Division, Federal 34.9% reported that the ICC had been in place for ˜University of Sao Paulo, and held from October 15 to just 5 years or less. This was more often the case inDecember 14, 2007. Members of infection control com- the answers of professionals from smaller hospitalsmittees (ICC) registered for the WBC were eligible to (P , .001).take part in the survey. Overall, 52.9% of the respondents said their hospitals This educational online program was designed by the had only 1 full-time employed (40 hours per week) infec-Division of Infectious Diseases of the Federal University tion control nurse, but, among respondents from larger ˜of Sao Paulo with funds from a collaborative project hospitals (more than 400 beds), 68.6% reported 3 orbetween the Pan American Health Organization, the more full-time employed infection control nurses. TheNational Health Surveillance Agency, and the Ministry others characteristics of ICC are summarized in Table 1.of Health in Brazil. The target audiences for this WBCwere physicians, public health personnel, nurses, phar-macists, and other HCW involved with IC in health care ICP essential activities and policiessettings. All Brazilian hospitals were contacted by e-mail We asked professionals about the use of an indicatoror letter to encourage the participation of at least 1 rep- measurement system for monitoring nosocomial infec-resentative from each hospital. The course was offered tions, and 15.1% said they used no indicator, whereasat no cost to all selected professionals, and online regis- 66.8% reported using the Centers for Disease Controltration was required. To complete the registration pro- and Prevention-National Nosocomial Infections Sur-cess, applicants were required to ﬁll out a pre-course veillance System.10 This practice was found to be sig-survey. They were assured total conﬁdentiality of their niﬁcantly more common in the Southeast and lessidentities and answers and informed that the survey common in the North region (P , .001).had no relation to the approval of their application for Sixty-seven percent of the professionals said thatthe course. This study has been approved by the Ethics HAI rates in their institutions were reported to the ˜Committee of the Federal University of Sao Paulo. hospital administration and discussed in regular The questionnaire consisted of 13 multiple choice ICC meetings (at least 1 meeting each 2 months).questions. The ﬁrst part (Table 1) focused on the char- Each professional was also asked whether their insti-acteristics of the health care unit, including geographic tution had deﬁned written protocols related to IClocation. The next set of questions addressed essential policies. Six hundred seventy-one respondentsICP activities, including monitoring systems and anti- (33.6%) said their institutions have no written proto-microbial use control (Table 1). In the last question, col (Table 1).respondents were asked to choose the closest alterna- We asked the professionals whether their hospitalstive to the estimated prevalence of AR of speciﬁc path- had an established antimicrobial stewardship pro-ogen-drug combinations and indicate the measures gram or any kind of antimicrobial use policy. Thetaken to control the transmission of these pathogens most commonly reported strategy for antimicrobialin their institutions (Tables 2 and 3). use control was formulary restriction (68.4%), The data were analyzed with the SPSS software, ver- whereas 247 (12.4%) professionals said their institu-sion 10.0 (SPSS, Inc, Chicago, IL). The x2 test, Fisher tion had no kind of antimicrobial use control systemexact test, or Pearson correlation coefﬁcient was calcu- (Table 1).lated, as appropriate. A P value less than .05 was con- Most professionals said their hospitals perform mi-sidered as statistically signiﬁcant. crobiologic surveillance targeted at epidemiologicallyRESULTS important multidrug-resistant organisms, more often vancomycin-resistant Enterococci and methicillin- Of 6256 Brazilian HCW registered for the WBC, 1998 resistant Staphylococcus aureus and implement isola-were members of ICCs from more than 1000 different tion procedures for patients infected or colonizedBrazilian institutions, and all answered the survey. with these epidemiologically important multidrug-Most respondents (43.2%) were from institutions resistant organisms, at least in some speciﬁc situationslocated in the Southeastern region. Nearly half of the (Table 2). The answers obtained concerning theICC members were nurses (42.4%), 27.8% were physi- estimated prevalence of AR for the speciﬁc pathogen-cians, 23.8% were pharmacists, and 6.0% were other drug combinations matching with the data fromtypes of health care providers. The others characteris- the Brazilian Antimicrobial Resistance Surveillancetics are summarized in Table 1. Network11 are shown in Table 3.
www.ajicjournal.org Guerra et al. 61Vol. 38 No. 1Table 1. Overall characteristics of institutions, ICP, and (approximately 100 million) and economic resources inICC activities reported by surveyed professionals the Southeast and in the South. The distribution of health care services also varies widely, and, of the 7155 hospitalsTopics Number % in Brazil, 48.6% are in the Southeast and in the South.12Overall characteristics Despite the recommendation from the Ministry of Geographic area Health in Brazil, since the 1980s, that every hospital Southeast 864 43.2 should maintain an ICP, 34.9% of respondents said Northeast 443 22.2 South 329 16.4 that ICP in their institutions had been implemented Midwest 229 11.5 less than 5 years ago. On the other hand, our survey North 133 6.7 showed that, on average, there is 1 infection control Type of institution nurse for every 200 beds, a ratio that complies with Public 1108 55.5 Brazilian regulations. Almost all the respondents said Private 644 32.2 Teaching hospital 246 12.3 that their institution’s ICC includes a physician, but Number of beds this can be explained by the fact that most hospital ,100 814 40.7 heads in Brazil are physicians; these professionals are 101-200 649 32.5 in charge of hospital administration and probably 201-400 395 19.8 supervise the ICC, but they do not perform epidemio- .401 140 7.0 Number of ICU beds logic surveillance functions. Surveys conducted in the None 488 24.4 United States reveal that the proportion of time spent 1-10 477 23.9 in IC activities by physicians is small.6 11-30 724 36.2 Written protocols for infection prevention and anti- .31 309 15.5 microbial stewardship programs were consideredTime in activity and human resources Time in activity, y important strategies to reduce HAI in the SENIC study,2 ,5 698 34.9 but, in our survey, approximately one third of the pro- 5-10 493 24.7 fessionals said their institutions have no written proto- .10 806 40.4 cols, and 12.4.0% reported no antimicrobial use Number of nurses control system. Such scenarios was more often 1 1058 52.9 2 565 28.4 reported in Midwest hospitals and less often reported 3 or more 374 18.7 by professionals from the Southeast and the South Number of physicians (P , .001) reinforcing the disparities between regions. None 48 2.4 We found that most professionals underestimated 1 1134 56.8 the prevalence of AR when their answers were con- 2 500 24.9 3 or more 315 15.9 fronted with data from the Brazilian AntimicrobialEssential activities of ICP Resistance Surveillance Network.11 This network was Surveillance system established in 2006 when selected hospitals in Brazil None 302 15.1 routinely began reporting the microbiologic proﬁle of NNISS 1335 66.8 selected pathogens identiﬁed in intensive care unit pa- Other than NNISS 361 18.1 Antimicrobial use control tients with bloodstream infections for inclusion into a None 247 12.4 national database. Data have been collected using stan- Formulary restriction 1366 68.4 dardized protocols from the National Nosocomial Clinical guidelines for antimicrobial use 406 20.3 Surveillance System and the Clinical Laboratory Stan- Discussion with infection control team 320 16.0 dards Institute. Maybe this misperception of the AR Computer-based decision support system 155 7.8 Written prevention protocols issue could explain the low adherence to antimicrobial None 671 33.6 stewardship programs found in our study. These ﬁnd- Bloodstream infections 771 38.6 ings are similar to those of a recent survey conducted Urinary tract infections 764 38.2 with 310 physicians in a Brazilian teaching hospital Ventilator-associated lung infections 807 40.4 in which most respondents underestimated the preva- Surgical antimicrobial prophylaxis 1204 60.3 lence of AR for 3 bug-drug combinations.13NOTE. N51998 individuals. Our survey showed the importance of developingNNISS, National Nosocomial Infections Surveillance System. written protocols for infection control policies, particu-DISCUSSION larly in Midwest hospitals, and standardized monitor- ing methods, particularly in the North region. Brazil has a large territory—more than 8 million Furthermore, it showed the need for improving thesquare kilometers—and a population of 185 million peo- access of the infection control staff to the data fromple. There is a high concentration of inhabitants local antimicrobial susceptibility tests. This could be
62 Guerra et al. American Journal of Infection Control February 2010Table 2. Microbiologic surveillance and isolation measures reported by professionals as used in their institutions Microbiologic surveillanceMicroorganisms No instance, n (%) Some cases, n (%) Regularly found, n (%) Isolation measures, n (%)Staphylococcus aureus and methicillin 290 (22.2) 761 (58.1) 258 (19.7) 1049 (80.1)ESBL-producing Klebsiella spp 319 (34.2) 466 (50) 147 (15.8) 772 (82.8)Pseudomonas spp and carbapenem 253 (28.3) 474 (53) 168 (18.8) 817 (91.3)Acinetobacter spp and carbapenem 1409 (28.4) 278 (52.8) 100 (19) 493 (93.5)Enterococcus spp and vancomycin 49 (14.6) 185 (55.2) 101 (30.1) 298 (89)ESBL, Extended-spectrum b-lactamase.Table 3. Prevalence of antimicrobial resistance estimated by each professional for some speciﬁc pathogen-drugcombinations in his/her institutionPrevalence of antimicrobial resistance ,20%, n (%) 20%-40%, n (%) .40%, n (%) Unknown, n (%) National data (%)*Staphylococcus aureus and methicillin 516 (39.4) 243 (18.6) 234 (17.9) 316 (24.1) 61.4ESBL-producing Klebsiella spp 434 (46.6) 192 (20.6) 68 (7.3) 238 (25.5) 66.8Pseudomonas spp and carbapenem 444 (49.6) 155 (17.3) 86 (9.6) 210 (23.5) 47.6Acinetobacter spp and carbapenem 268 (50.9) 88 (16.7) 52 (9.9) 119 (22.6) 48.9Enterococcus spp and vancomycin 203 (60.6) 33 (9.8) 5 (1.5) 94 (28.1) 17.5ESBL, Extended-spectrum b-lactamase.*Brazilian Antimicrobial Resistance Surveillance Network: rates of antimicrobial resistance in selected pathogens identiﬁed from ICU patients with bloodstream infections. For eachantimicrobial/pathogen pair, the pooled mean rate of resistance for January through December 2007 is displayed.achieved through the development of an antimicrobial 2. Haley RW, Culver DH, White JW, Morgan WM, Emori TG, Munn VP,surveillance program and the extensive communica- et al. The efﬁcacy of infection surveillance and control programs in preventing nosocomial infections in US hospitals. Am J Epidemioltion of data on local prevalence, particularly of multi- 1985;121:182-205.drug-resistant organisms. 3. Struelens MJ, Wagner D, Bruce J, MacKenzie FM, Cookson BD, Voss A, This study has some limitations. First, it is based on et al. Status of infection control policies and organization in Europeanself-reported data and focused on the perceptions of hospitals, 2001: the ARPAC study. Clin Microbiol Infect 2006;12:729-37.ICC members and their knowledge about ICP in their 4. Oh HS, Chung HW, Kim JS, Cho S. National survey of the status of infection surveillance and control programs in acute care hospitalsinstitutions; therefore, we may draw a picture of the with more than 300 beds in the Republic of Korea. Am J Infect Con-current status of infection control in different Brazilian trol 2006;34:223-33.regions, but further studies are warranted to establish 5. Zoutman DE, Ford BD, Bryce E, Gourdeau M, Hebert G. Canadianthe reliability and validity of the survey. In addition, Hospital Epidemiology Committee. The state of infection surveillancethis study included only hospitals that have an ICC, and control in Canadian acute care hospitals. Am J Infect Control 2003;31:266-73.and it might be important for Brazilian authorities to 6. Nguyen GT, Proctor SE, Sinkowitz-Cochran RL, Garrett DO, Jarvishave information on institutions that still lack an WR. Association for Professionals in Infection Control and Epidemiol-active ICC. ogy. Status of infection surveillance and control programs in the Our study highlights some important information United States, 1992-1996. Am J Infect Control 2000;28:392-400.about the perceptions and attitudes of ICC members 7. Pannuti CS, Grinbaum RS. An overview of nosocomial infection con- trol in Brazil. Infect Control Hosp Epidemiol 1995;16:170-4.that may be used to tailor key interventions for the sur- 8. Wey SB. Infection control in a country with annual inﬂation of 3600%.veillance and effective control of multidrug-resistant Infect Control Hosp Epidemiol 1995;16:175-8.organism infections. It suggests, additionally, that, to x˜ ` ˆ ` ´ 9. Infeccoes relacionadas a assistencia a saude (IRAS); 2005. Availableachieve countrywide standardized IC mechanisms in from: http://www.iras.org.br. Accessed August 30, 2008.a developing country, authorities should consider the 10. Emori TG, Culver DH, Horan TC, Jarvis WR, White JW, Olson DR, et al. National Nosocomial Infections Surveillance System (NNIS). De-social, cultural, and economical disparities among re- scription of surveillance methods. Am J Infect Control 1991;19:19-35.gions and identify speciﬁc regional needs to make ˆ 11. Rede Nacional de Monitoramento da Resistencia Microbiana emavailable the resources required to minimize such x ´ Servicos de Saude (Rede RM). Available from: http://www.anvisa.disparities. gov.br/servicosaude/controle/rede_rm/index.htm. Accessed October 25, 2008. 12. Brazilian Institute of Geography and Statistics (IBGE). Available from:References http://www.ibge.gov.br. Accessed October 25, 2008. 13. Guerra CM, Pereira CAP, Neves Neto AR, Cardo DM, Correa L. Phy- ˆ 1. Scheckler WE, Brimhall D, Buck AS, Farr BM, Friedman C, Garibaldi sicians’ perceptions, beliefs, attitudes, and knowledge concerning anti- RA, et al. Requirements for infrastructure and essential activities of in- microbial resistance in a Brazilian teaching hospital. Infect Control fection control and epidemiology in hospitals: a consensus panel re- Hosp Epidemiol 2007;28:1411-4. port. Infect Control Hosp Epidemiol 1998;19:114-24.