Preventing Health Care–Associated Infections

Amy S. Collins.

Amy S. Collins, B.S., B.S.N., M.P.H., Centers for Disease Control and Prevention; Atlanta, Georgia. E-mail: acollins@cdc.gov

The occurrence and undesirable complications from health care–associated infections (HAIs) have been well recognized in the literature for the last several decades. The occurrence of HAIs continues to escalate at an alarming rate. HAIs originally referred to those infections associated with admission in an acute-care hospital (formerly called a nosocomial infection), but the term now applies to infections acquired in the continuum of settings where persons receive health care (e.g., long-term care, home care, ambulatory care). These unanticipated infections develop during the course of health care treatment and result in significant patient illnesses and deaths (morbidity and mortality); prolong the duration of hospital stays; and necessitate additional diagnostic and therapeutic interventions, which generate added costs to those already incurred by the patient’s underlying disease. HAIs are considered an undesirable outcome, and as some are preventable, they are considered an indicator of the quality of patient care, an adverse event, and a patient safety issue.

Patient safety studies published in 1991 reveal the most frequent types of adverse events affecting hospitalized patients are adverse drug events, nosocomial infections, and surgical complications.^{1, 2} From these and other studies, the Institute of Medicine reported that adverse events affect approximately 2 million patients each year in the United States, resulting in 90,000 deaths and an estimated $4.5–5.7 billion per year in additional costs for patient care.³ Recent changes in medical management settings have shifted more medical treatment and services to outpatient settings; fewer patients are admitted to hospitals. The disturbing fact is that the average duration of inpatient admissions has decreased while the frequency of HAIs has increased.^{4, 5} The true incidence of HAIs is likely to be underestimated as hospital stays may be shorter than the incubation period of the infecting microorganism (a developing infection), and symptoms may not manifest until days after patient discharge. For example, between 12 percent and 84 percent of surgical site infections are detected after patients are discharged from the hospital, and most become evident within 21 days after the surgical operation.^{6, 7} Patients receiving followup care or routine care after a hospitalization may seek care in a nonacute care facility. The reporting systems are not as well networked as those in acute care facilities, and reporting mechanisms are not directly linked back to the acute care setting to document the suspected origin of some infections.

Since the early 1980s HAI surveillance has monitored ongoing trends of infection in health care facilities.⁸ With the application of published evidence-based infection control strategies, a decreasing trend in certain intensive care unit (ICU) health care-associated infections has been reported through national infection control surveillance⁹ over the last 10 years, although there has also been an alarming increase of microorganism isolates with antimicrobial resistance. These changing trends can be influenced by factors such as increasing inpatient acuity of illness, inadequate nurse-patient staffing ratios, unavailability of system resources, and other demands that have challenged health care providers to consistently apply evidence-based recommendations to maximize prevention efforts. Despite these demands on health care workers and resources, reducing preventable HAIs remains an imperative mission and is a continuous opportunity to improve and maximize patient safety.

Another factor emerging to motivate health care facilities to maximize HAI prevention efforts is the growing public pressure on State legislators to enact laws requiring hospitals to disclose hospital-specific morbidity and mortality rates. A recent Institute of Medicine report identified HAIs as a patient safety concern and recommended immediate and strong mandatory reporting of other adverse health events, suggesting that public monitoring may hold health care facilities more accountable to improve the quality of medical care and to reduce the incidence of infections.³ Since 2002, four States (Florida, Illinois, Missouri, and Pennsylvania) set legislation mandating health care organizations to publicly disclose HAIs.^{10, 11} In 2006, the Association for Professionals in Infection Control and Epidemiology (APIC) reported that 14 States have mandatory public reporting, and 27 States have other related legislation under consideration.¹² Participation in public reporting has not been regulated by the Federal sector at this time. Some hospital reporting is intended for use solely by the State health department for generating confidential reports that are returned to each facility for their internal quality improvement efforts. Other intentions to utilize public reporting may be aimed at comparing rates of HAI and subsequent morbidity and mortality outcomes between different hospitals. This approach is problematic as there is currently a lack of scientifically validated methods for risk adjusting multiple variations (e.g., differences in severity of illnesses in each population being treated) in patients’ intrinsic and extrinsic risks for HAIs.^13–15 Moreover, data on whether public reporting systems have an effective role in reducing HAIs are lacking.

To assist with generating meaningful data, process and outcome measures for patient safety practices have been proposed.^{13, 14, 16} Monitoring both process and outcome measures and assessing their correlation is a model approach to establish that good processes lead to good health care outcomes. Process measures should reflect common practices, apply to a variety of health care settings, and have appropriate inclusion and exclusion criteria. Examples include insertion practices for central intravenous catheters, appropriate timing of antibiotic prophylaxis in surgical patients, and rates of influenza vaccination for health care workers and patients. Outcome measures should be chosen based on the frequency, severity, and preventability of the outcome events. Examples include intravascular catheter-related blood stream infection rates and surgical-site infections in selected operations. Although these occur at relatively low frequency, the severity is high—these infections are associated with substantial morbidity, mortality, and excess health care costs—and there are evidence-based prevention strategies available.^{17, 18}