Health Care–Associated Bloodstream Infections: A Change in Thinking

Editorials19 November 2002Health Care–Associated Bloodstream Infections: A Change in ThinkingFREERobert Gaynes, MDRobert Gaynes, MDCenters for Disease Control and Prevention; Emory University School of Medicine; Atlanta, GA 30333Author, Article, and Disclosure Informationhttps://doi.org/10.7326/0003-4819-137-10-200211190-00015 SectionsAboutVisual AbstractPDF ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinkedInRedditEmail It is not necessary to change. Survival is not mandatory.—W. Edwards DemingNosocomial, or hospital-acquired, bloodstream infections (BSIs) are an important cause of morbidity and mortality, affecting more than 200 000 patients per year in the United States (1). These infections are often associated with the use of catheters. Catheter-related infections make up about 14% of all hospital-acquired infections (2) and are responsible for more hospital days and death than any other hospital-acquired infection (3). Bloodstream infections can be primary or secondary. A primary infection is an infection directly into the vascular system; a secondary infection originates at a site other than the vascular system, such as the urinary tract, respiratory tract, or wounds, and spreads to the vascular system. According to the definitions from the Centers for Disease Control and Prevention (CDC) National Nosocomial Infection Surveillance (NNIS) System, primary BSIs account for 64% of nosocomial BSIs (2, 4)."Nosocomial," "community-acquired," "bloodstream infections," and "primary/secondary" are surveillance terms that have stood, essentially unaltered, for almost two decades (4). Yet the health care system has been anything but static during this period. Routine care for patients with serious underlying illnesses and with invasive devices, such as intravascular catheters, has shifted from exclusively acute care settings to nursing homes, rehabilitation centers, free-standing dialysis centers, and homes. Complications of caring for patients with serious illness occur in all of these settings.In this issue, Friedman and colleagues propose a new classification scheme for BSIs (5). This scheme does not change the fundamental classification of what we call BSIs. It only changes the number of categories of BSIs from two (nosocomial and community-acquired) to three (nosocomial, health care–associated, and community-acquired). Clinicians have recognized that the pathogens and pathogenesis of nosocomial BSIs (for example, those resulting from complications related to placement of an intravenous catheter) and community-acquired BSIs (that is, those not related to a complication from a medical device or procedure) differ. The authors take issue with the term "community-acquired," stating that it is too broad and no longer useful in the new world of serious illness in diverse venues.The authors reviewed daily microbiology laboratory reports from adults in three hospitals. After excluding patients with contaminated blood cultures, the authors classified patients with BSIs into one of the three new categories. A nosocomial BSI was an infection from patients hospitalized 48 hours or longer. A health care–associated BSI was an infection from a patient who had recent contact with some aspect of health care, such as intravenous therapy via a home health care agency, residence in a nursing home, recent hospitalization, or dialysis. A community-acquired BSI was an infection in a patient with no discernible recent contact with the health care community. After the authors classified the BSIs, they used the CDC definitions to classify whether each patient had a primary or secondary BSI; they then performed a descriptive analysis of the patients.The authors found that patients with health care–associated BSIs are similar to patients with nosocomial BSIs but different from patients with community-acquired BSIs. On the surface, this distinction seems rather obvious. The invasive devices that might be associated with a BSI in a hospitalized patient are the same as or similar to those used in nonhospitalized patients; in addition, nonhospitalized patients often have the same comorbid factors. The authors' important observation is not that patients with health care–associated BSIs are similar to patients with hospital-acquired BSIs. Rather, it is that they are different from patients with community-acquired BSIs.The frequency of bacteremia resulting from the use of intravascular devices and from gastrointestinal tract infection was similar in patients with health care–associated and nosocomial BSIs. The frequency of BSIs secondary to urinary tract infection was higher in patients with community-acquired BSIs than in patients with nosocomial and health care–associated BSIs. The pathogens also differed among these groups. Staphylococcus aureus was the most frequently isolated pathogen in health care–associated and nosocomial BSIs; Escherichia coli and Streptococcus pneumoniae were most frequently isolated in community-acquired BSIs. The authors assert that patients with suspected BSIs who have had recent contact with the health care community (health care–associated BSIs) should receive empirical treatment that is similar to the treatment of hospitalized patients with suspected BSIs.These results should be viewed cautiously. First, the authors studied 504 patients, whereas an estimated 200 000 patients have BSIs annually in the United States. Although some of their findings may not apply to the population as a whole, their basic contribution—the distinction between a health care–associated and a community-acquired BSI—is likely to endure because it reflects fundamental changes in health care. Second, the authors did not measure the risk for health care–associated BSIs. They examined only patients with BSIs (the numerator in a calculation of risk) and did not attempt to determine the size of the population at risk for a health care–associated BSI (the denominator for a risk calculation).Measuring the risk for BSI can be important for several reasons. It could identify the patients at highest risk for an event so that resources could be used more efficiently for monitoring and intervention. For example, when compared with other hospitalized patients, those patients in critical care areas are disproportionately at risk for BSIs and other nosocomial infections, largely as a result of their exposure to invasive devices (6). Measuring the risk for health care–associated BSI could also be the first step in setting benchmarks of expected performance to avoid BSIs. Benchmarking has resulted in substantial reductions in the rate of BSIs in intensive care units in the NNIS System (7). Benchmarking for health care–associated BSIs, particularly with home health care agencies that provide intravenous therapy in the home, would be an important step toward improving community-based care of serious illness.Monitoring for BSIs will be more difficult in communities than in hospitals, where it is relatively easy to determine the denominator for a risk calculation. Friedman and colleagues provide some guidance for defining the numerator for a health care–associated BSI rate. However, after patients leave the hospital, data collection for the denominator for a risk calculation becomes exceedingly difficult. An integrated electronic medical record across all venues of health care may eventually solve this problem (8).As the care of serious illness shifts away from acute care facilities, other complications of care are likely to occur outside the hospital (9, 10). Therefore, the study by Friedman and colleagues should prompt efforts to find other possible adverse events associated with the care of serious illness in the community. Undoubtedly, events such as thrombotic complications, which we have learned to monitor and expect in the hospital setting, are occurring in patients in other settings. Quantifying these events in the community is an attractive target for research.The study by Friedman and colleagues presents an important clinical message—that clinicians should reconsider therapy for patients with community-acquired BSIs if their recent care falls into the health care–associated category. Empirical treatment with antibiotics that are effective against the commonly isolated Staphylococcus species (often with oxacillin resistance) and not only E. coli or Streptococcus pneumoniae, which are more characteristic of community-acquired BSIs, should be selected. We must change our thinking such that we not only care for serious illness but also look at the setting in which that care is taking place.Robert Gaynes, MDCenters for Disease Control and Prevention; Emory University School of Medicine Atlanta, GA 30333References1. Maki DG. Infections due to infusion therapy.. In: Bennett JV, Brachman PS, eds. Hospital Infections. 4th ed. Boston: Little, Brown; 1998. Google Scholar2. Martone WJ, Gaynes RP, Horan TC, Danzig L, Emori TG, Monnet D, et al. . National Nosocomial Infections Surveillance (NNIS) semiannual report, May 1995. A report from the National Nosocomial Infections Surveillance (NNIS) System. Am J Infect Control. 1995;23:377-85. [PMID: 8821115] CrossrefMedlineGoogle Scholar3. Public health focus: surveillance, prevention, and control of nosocomial infections. MMWR Morb Mortal Wkly Rep. 1992;41:783-7. [PMID: 1406572] MedlineGoogle Scholar4. Garner JS, Jarvis WR, Emori TG, Horan TC, Hughes JM. CDC definitions for nosocomial infections, 1988. Am J Infect Control. 1988;16:128-40. [PMID: 2841893] CrossrefMedlineGoogle Scholar5. Friedman ND, Kaye KS, Stout JE, McGarry SA, Trivette SL, Briggs JP, et al. . Health care–associated bloodstream infections in adults: a reason to change the accepted definition of community-acquired infections. Ann Intern Med. 2002;137:791-7. LinkGoogle Scholar6. 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;27:887-92. [PMID: 10362409] CrossrefMedlineGoogle Scholar7. Gaynes R, Richards C, Edwards J, Emori TG, Horan T, Alonso-Echanove J, et al. . Feeding back surveillance data to prevent hospital-acquired infections. Emerg Infect Dis. 2001;7:295-8. [PMID: 11294727] CrossrefMedlineGoogle Scholar8. Tang PC. AMIA advocates national health information system in fight against national health threats. J Am Med Inform Assoc. 2002;9:123-4. [PMID: 11861625] CrossrefMedlineGoogle Scholar9. Field TS, Gurwitz JH, Avorn J, McCormick D, Jain S, Eckler M, et al. . Risk factors for adverse drug events among nursing home residents. Arch Intern Med. 2001;161:1629-34. [PMID: 11434795] CrossrefMedlineGoogle Scholar10. Sands K, Vineyard G, Livingston J, Christiansen C, Platt R. Efficient identification of postdischarge surgical site infections: use of automated pharmacy dispensing information, administrative data, and medical record information. J Infect Dis. 1999;179:434-41. [PMID: 9878028] CrossrefMedlineGoogle Scholar Comments0 CommentsSign In to Submit A Comment Author, Article, and Disclosure InformationAuthors: Robert Gaynes, MDAffiliations: Centers for Disease Control and Prevention; Emory University School of Medicine; Atlanta, GA 30333Corresponding Author: Robert Gaynes, MD, Healthcare Outcomes Branch, Division of Healthcare Quality Promotion, Centers for Disease Control and Prevention, 1600 Clifton Road NE, E-55, Atlanta, GA 30333. PreviousarticleNextarticle Advertisement FiguresReferencesRelatedDetailsSee AlsoHealth Care–Associated Bloodstream Infections in Adults: A Reason To Change the Accepted Definition of Community-Acquired Infections N. Deborah Friedman , Keith S. Kaye , Jason E. Stout , Sarah A. McGarry , Sharon L. Trivette , Jane P. Briggs , Wanda Lamm , Connie Clark , Jennifer MacFarquhar , Aaron L. Walton , L. Barth Reller , and Daniel J. Sexton Metrics Cited byIsolation of ESBL-producing Bacteria from Sputum in Community-acquired Pneumonia or Healthcare-associated Pneumonia Does Not Indicate the Need for Antibiotics with Activity against This ClassProspective Evaluation of a Novel Treatment Algorithm for Health Care-Associated PneumoniaNosocomial PneumoniaIs the Present Definition of Health Care–Associated Pneumonia the Best Way to Define Risk of Infection with Antibiotic-Resistant Pathogens?Community-Associated Extended-Spectrum β-Lactamase–Producing Escherichia coli Infection in the United StatesProspective randomized comparison study of piperacillin/tazobactam and meropenem for healthcare-associated pneumonia in JapanRelationship between the pathogens of postoperative pneumonia after an esophagectomy for thoracic esophageal cancer and the aggregate length of preoperative hospital stayDiversificación y evolución clonal de patógenos grampositivos: nuevas perspectivas y retos en el siglo xxiAnforderungen an die Hygiene bei der medizinischen Versorgung von immunsupprimierten PatientenÜbertragung von Problemkeimen im ambulanten BereichNosocomial Pneumonia in Critical CareInfective Endocarditis and Its Mimics inCriticalCareNosocomial Pneumonia in Critical CareInfective Endocarditis and Its Mimics inCriticalCareBloodstream infections and setting of onset in persons with spinal cord injury and disorderGemeinsame Stellungnahme zur Erfassung nosokomialer und gesundheitssystemassoziierter Infektionen in der PädiatrieWhat is healthcare-associated pneumonia and how is it managed?Healthcare associated infective endocarditis: A distinct entityEpidemiologyNosocomial and Health Care-Associated Infective Endocarditis (Iatrogenic Infective Endocarditis)Providing appropriate antimicrobial therapy in the intensive care unit: Surveillance vs. de-escalation*Optimizing Antibiotic Treatment for Ventilator-Associated PneumoniaStaphylococcus Aureus PneumoniaGram-Negative Bacterial Resistance: Evolving Patterns and Treatment ParadigmsGuidelines for the Management of Adults with Hospital-acquired, Ventilator-associated, and Healthcare-associated PneumoniaThe importance of antimicrobial resistance in hospital-acquired and ventilator-associated pneumoniaHospital Mortality for Patients With Bacteremia Due to Staphylococcus aureus or Pseudomonas aeruginosaEditorial Commentary: An Empirical Approach to the Treatment of Multidrug‐Resistant Ventilator‐Associated PneumoniaAppropriate Empirical Antibacterial Therapy for Nosocomial Infections 19 November 2002Volume 137, Issue 10Page: 850-851KeywordsBacteremiaCathetersHealth careHealth promotionNosocomial infectionsPneumoniaPrevention, policy, and public healthStreptococcusUpper respiratory tract infectionsUrinary tract infections ePublished: 19 November 2002 Issue Published: 19 November 2002 PDF downloadLoading ...