The role of health care facilities
Download this section
Chapter 4

Ramanan Laxminarayan

Rapid improvements in medical technology have made possible lifesaving interventions that keep hospitalized patients alive for longer. However, the downside of these interventions is that patients tend to be sicker, spend longer periods of time in the hospital, and are more in need of intensive medical care than before, leading to an increased prevalence of many nosocomial infections.1 Also known as a hospital-acquired infection (HAI), a nosocomial infection is acquired in a hospital by a patient who was admitted for a reason other than that infection. Moreover, protracted illness and time on life support for these patients, many of whom are immuno-compromised, have increased reliance on antibiotics to help stave off infection, which in turn has resulted in increasing drug resistance among common, previously treatable HAIs.

According to the Centers for Disease Control and Prevention (CDC), HAIs account for an estimated 2 million infections and 90,000 deaths each year. Common HAIs include infections of surgical wounds, urinary tract infections, and lower respiratory tract infections. Infections acquired in health care institutions are among the top 10 causes of death in the United States: they are the primary cause of 1 percent of all deaths and are major contributors to an additional 2 percent of all deaths (Harrison and Lederberg 1998). Many of the endemic bacteria causing these infections are resistant to one or more classes of antibiotics’ pose a major challenge to inpatient health, and significantly increase the costs of hospital stays. In fact, the United States has among the highest rates of drug-resistant hospital infections in the world, as described in Chapter 1. Vancomycin-resistant enterococci (VRE) and methicillin-resistant Staphylococcus aureus (MRSA) are among the most important HAIs because they now account for a large fraction of nosocomial infections, but they are not the only problematic pathogens: increasingly, resistant Gramnegative bacteria such as Pseudomonas aeruginosa and Klebsiella pneumoniae are causing serious infections in hospital patients. Hospitals and long-term care facilities like nursing homes and hospice care tend to use large quantities of antibiotics and are consequently significant reservoirs of resistant pathogens. The ability of these pathogens to persist may be due to multiple interacting factors, including excessive antibiotic use, poor hygiene by health care workers, high susceptibility of patients, establishment in long-term care facilities (as  well as in prisons and in the community), and colonization of hospital staff or the hospital environment. Each of these factors contributes to the emergence and establishment of endemicity within a clinical setting. In addition to the impact of endemic antibiotic-resistant bacteria on their own patients, hospitals are significant reservoirs of resistant pathogens that can be transported to other facilities.

Strategies for lowering the resistance levels in hospitals fall into three categories.2 First is lowering antibiotic use by requiring preapproval for certain antibiotic prescriptions. Second is using creative antibiotic restriction strategies, such as cycling and treatment heterogeneity. Third is better infection control, which is applicable not just to resistant pathogens, but to all HAIs. Studies suggest that the economic and health benefits of many common interventions to lower the prevalence of HAIs exceed the costs. In this chapter we explore the incentives for hospitals3 to invest in hospital infection control (HIC) and other measures to lower the prevalence of resistant bacteria in their facilities, as well as potential regulatory solutions to encourage greater reporting and improved infection control.

Economic costs and benefits

HAIs cost between $17 billion and $29 billion each year, and older studies have shown that a third of this burden can be lowered by adequate infection control programs (Haley, Culver et al. 1985). Numerous studies show that HAIs, especially resistant infections, cause longer hospital stays, greater risk of death, and much higher rates of hospitalization. There is also strong evidence that the overall economic benefits of infection control programs exceed costs by a wide margin and that “an effective infection control programme is one of the most cost-beneficial medical interventions available in modern public health” (Wenzel 1995). However, there is considerable disagreement over who bears the principal economic burden of these infections, and this influences incentives for health care facilities to engage in better infection control. In this section, we review existing evidence on the economic benefits of hospital infection control and incentives for hospitals to engage in it.

Cost of hospital-acquired infections

Numerous studies have documented the increased costs of nosocomial bloodstream infections, stretching back into the 1970s and 1980s. Pittet, Tarara et al. (1994) and Pittet and Wenzel (1995) found that during the 1980s, the incidence and risk of death from nosocomial bloodstream infections had risen markedly and that a patient with a nosocomial bloodstream infection was 35 percent more likely to die; for a patient who survived, extra costs attributable to the infection were approximately $40,000, and extra hospital costs, $6,000. Haley (1986) looked at all nosocomial infection costs and found that the average cost was about $1,800 per infection, with a maximum cost of about $42,000.

It is important to recognize the significant economic costs that nosocomial infections impose on both the hospital and the patient. The congressional Office of Technology Assessment has estimated the minimal hospital cost associated with nosocomial infections caused by antibiotic-resistant bacteria to be $1.3 billion per year (in 1992) (OTA 1995). This does not include the increased cost to patients, both monetarily and through the indirect and long-term morbidity and mortality consequences of resistant infections. In addition, most published studies have shown increased mortality risk on the order of 1.3 to 2 times, which may also have significant effects on indirect costs, such as long-term lost productivity. It is also important to understand that antibiotic resistance has an effect on many patients who do not become infected: they have to use stronger drugs, which may be more expensive, have more dangerous side effects, or be more toxic or possibly less effective than older or mainline drugs.

Those indirect costs aside, the cost of an antibiotic-resistant infection is still significant. According to Cosgrove, Qi et al. (2005), a nosocomial MRSA bacteremia significantly increases the length of hospital stays, the charges per patient, and hospital costs per case. They estimate that the excess cost of an MRSA bacteremia is $26,424 in patient charges and $14,655 in excess hospitals costs (total, $41,079 in excess charges) versus a control population. They also calculated costs for patients with methicillin-sensitive Staphylococcus infection (MSSA); these averaged $19,212 in excess patient charges and $10,655 in excess hospital costs (total, $29,867). McHugh and Riley (2004), similarly, estimated total per patient costs (as opposed to excess costs) of $9,699 for an MSSA infection versus $45,920 for an MRSA infection (an excess cost of $36,221).

Another important problem is surgical site infections, which are responsible for increased morbidity and mortality and cost hospitals more than $1.6 billion in extra charges each year (Martone and Nichols 2001). Engemann, Carmeli et al. (2003) studied MRSA in surgical site infections in a large cohort at the Duke University Medical Center. MRSA in a surgical wound was found to result in more than a 12-fold increase in mortality versus non-infected patients and more than a 3-fold increase versus patients infected with MSSA. MRSA infections also cost patients about $40,000 more than an MSSA infection and about $84,000 more than an uninfected patient.

Vancomycin-resistant enterococci (VRE) are also associated with higher morbidity, mortality, and costs. Carmeli, Eliopoulos et al. (2002) found that a VRE infection led to longer hospital stays, a 2-fold increase in the rate of mortality, increased odds that a patient would require major surgery or be placed in the intensive care unit, and a 1.4-fold increase in hospital costs, which over the length of the study translated to excess costs of $2,974,478 (233 patients at an excess cost of $12,766 each). In addition, the authors found an increase in the likelihood that a patient would end up being discharged to a long-term care facility, meaning that the additional costs of a VRE infection are significantly understated in the study and that they continue for many patients. These estimates are lower than in another study (Stosor, Peterson et al. 1998), which found that VRE bacteremia was associated with $27,190 in excess costs; yet another study (Song, Srinivasan et al. 2003) found mean excess costs of VRE to be $81,208.

According to the Pennsylvania Health Care Cost Containment Council (PHC4) (PHC4 2005), the average charge for Pennsylvania Medicare patients with HAIs was about $160,000, five times the $32,000 average charge for Medicare patients who did not contract infections. Among Medicaid patients, the average charge was approximately $391,000 for patients who contracted infections while hospitalized, compared with an average of $29,700 when infections did not occur. Private commercial insurers of businesses and labor unions that provide health insurance were billed for almost 23 percent, or 2,633, of the reported hospital-acquired infections, which added $604 million in extra hospital charges. The average charge for a hospital admission in which a commercially insured patient contracted an infection was almost $258,000, compared with an average of $28,000 for admissions when infections did not occur. The average charge for stays in which uninsured patients contracted infections reached almost $230,000, compared with $21,000 for an uninsured patient without an infection.

Benefits of hospital infection control

There has been relatively little evaluation of the impact of programs to lower antibiotic use within hospitals, but greater attention has been paid to the benefits and costs of infection control programs. For example, a program of intensive surveillance and interventions targeted at reducing the risk of hospital-acquired ventilator-associated pneumonia at the University of Massachusetts Medical Center in 1997–1998 lowered the incidence of this pneumonia and resulted in a cost savings greater than $350,000 (Lai, Baker et al. 2003).

Similarly, a 1994 VRE outbreak at the University of Virginia Hospital prompted an active surveillance program and contact isolation of colonized patients. The costs of the program, including time spent collecting samples, additional length of hospital stays in isolation, and laboratory fees, were estimated at $253,099 during the two-year study, during which time only one primary VRE bacteremia occurred (Muto, Giannetta et al. 2002). At a control hospital, where no such program was in place, there were 29 cases of VRE bacteremia during the corresponding period, and these resulted in an estimated cost of $761,320, based on an estimate of excess costs of $27,190 per case of VRE (Stosor, Peterson et al. 1998). Other percase VRE cost estimates would value the program benefits at $357,448 (Carmeli, Eliopoulos et al. 2002) to $2,273,824 (Song, Srinivasan et al. 2003), but even the lower end of these benefits far exceeded the costs of the program.

Two Charleston, S.C., hospitals implemented an active surveillance program and a contact isolation protocol as recommended by the Society for Health care Epidemiology of America (SHEA). Based on prior rates of nosocomial infections, the new programs and protocols prevented an estimated 13 MRSA bacteremias and 9 surgical site infections for a cost savings of about $596,960 for the prevented bacteremias ($45,920 per case, based on McHugh and Riley 2004) and $756,000 for the prevented surgical site infections ($84,000 in excess costs per case, based on Engemann, Carmeli et al. 2003). The cost of implementing the program was $113,955, comprising $54,381 for surveillance and $59,574 for contact isolation (West, Guerry et al. 2006).

Quality control in U.S. hospitals

This section provides an overview of how hospital quality, in general, and in particular with respect to infections, is currently measured and how hospitals are currently regulated or accredited.

Accreditation process

Hospital accreditation organizations such as the Joint Commission on Accreditation of Health care Organizations (JCAHO) currently do not require standards for antibiotic use, resistance, or nosocomial infections.4 Hospitals are required to report only whether they follow a certain set of best practices for infection control, and not infection revalence rates or resistance levels. JCAHO uses an onsite evaluation as the basis of accreditation. No long-term reporting is required for continued accreditation. Standards alone may not be able to solve the problem; a change in attitudes about hospital infections would come from a combination of education about the benefits of infection control and stronger incentives for hospitals to invest in control programs. Moreover, JCAHO clears more than 99 percent of all hospitals it inspects, which suggests that the current system is set up more to catch egregious violators of medical practices than to address pervasive problems like hospital-acquired infections and resistance (Gaul 2005).5

Health care quality organizations

The Leapfrog Consortium and other organizations that represent the interests of large purchasers of health care (such as automobile manufacturers) work with hospitals to encourage lpublic reporting of health care quality and outcomes. They use a carrot-and-stick approach by rewarding hospitals that perform well and by leveraging consumer and health care purchaser choice to improve poor performers. Information on hospital infection control practices—including safety measures, hand washing, and vaccination of health care staff—is collected using self-reported surveys by hospitals. However, like JCAHO, Leapfrog may be better at separating good institutions from bad ones than at discerning finer indicators of performance, such as the prevalence of hospitalacquired infection.

In general, hospital-acquired infections and resistance are not a focus for existing organizations like JCAHO and Leapfrog. Although drug resistance can be seen as a quality issue, the current system for determining hospital quality may not work well to improve reporting or compliance with better infection control practices.

HICPAC and SHEA guidelines

Existing initiatives to improve hospital infection control—such as by CDC’s Health care Infection Control Practices Advisory Committee (HICPAC) (McKibben, Horan et al. 2005) and SHEA (Muto, Jernigan et al. 2003)—provide guidance to hospitals to engage in greater infection control and thereby help prevent the spread of resistance. Both sets of guidelines are based on clinical evidence that the vast majority of MRSA and VRE infections are the result of transmission from patient to patient and not from de novo mutations, and thus they suggest that stringent infection control practices are probably the most important factor in limiting the spread of MRSA and VRE.

However, they differ in some important respects. In the context of MRSA and VRE, the SHEA guidelines call for active surveillance cultures to identify colonized patients, with barrier precautions for patients colonized or infected with MRSA and VRE. CDC guidelines, on the other hand, reject the need for active surveillance cultures on the grounds that they may impose unnecessary costs on hospitals. Nevertheless, the voluntary nature of these guidelines indicates that many hospitals are not likely to apply them unless they have a strong financial motivation for doing so.6

Reporting of infections and resistance in hospitals

Since 1970, data on hospital infections and prevalence of MRSA and VRE (based on passive surveillance) have been voluntarily reported confidentially by hospitals participating in CDC’s National Nosocomial Infection Surveillance (NNIS) program. These hospitals provide general medicalsurgical inpatient services to adults or children requiring acute care. With a few exceptions, most current understanding of the extent of HAIs and drug resistance comes from the NNIS surveillance. However, there are important problems with this system that restrict its usefulness in delivering a comprehensive, nationwide picture of hospital infections and resistance. First, the nearly 300 hospitals that participate in the program are self-selected and represent only about 2 percent of hospitals, mainly academic centers—raising strong concerns about selection bias. Second, reporting within hospitals can change significantly. For instance, hospitals do not necessarily report data from the same intensive care unit each year, making comparisons across years problematic. Third, NNIS data are generally not available to researchers outside CDC because of confidentiality agreements signed with hospitals. This has restricted wider use of these data.

In recent years, under strong pressure from consumer advocates, some states have moved to require public reporting of hospital infections. In 2006, Consumers Union reported that six states (Illinois, Pennsylvania, Missouri, Florida, Virginia, and New York) had hospital infection disclosure laws, and 30 states had introduced similar legislation requiring hospitals to report their infection levels to state monitoring bodies (CU 2006). By providing more transparency to consumers, better reporting of infection and resistance levels may give hospitals greater incentives to engage in infection control.

Incentives and disincentives to control resistant HAIs

Hospital incentives

Despite some awareness of the problem and new measures to tackle the growing threat, the overall trend of infections, both susceptible and resistant, appears to be upward, as seen in Chapter 1. Antibiotic restrictions and better infection control are the two main tools available to hospitals. Currently, antibiotic restrictions are the main strategy reported by hospitals. Cost containment had been the original reason for implementing these restrictions (to divert physicians from expensive antibiotics to cheaper generics), but these reasons have been reborn in the form of concerns about drug resistance.

Programs intended to control antibiotic-resistant infections associated with health care have been around for a long time; however, implementation of these programs has been highly variable across facilities. Moreover, the guidelines have mostly focused on contact precautions that require staff hand washing, staff cohorting,7 and use of protective equipment to prevent the spread of infection from patients identified as carrying an infection. Guidelines issued by SHEA in 2003, focused mainly on the spread of MRSA and VRE within the hospital setting, called existing measures insufficient and recommended active surveillance cultures to identify patients colonized but not infected with resistant pathogens (Muto, Jernigan et al. 2003).

Next we consider important reasons why hospitals may not
invest heavily in infection control programs on their own.

Hospital disincentives

The extent to which hospitals bear the cost of resistant HAIs is a subject of disagreement, as is the extent to which these costs are passed on to Medicare, Medicaid, and private insurers. If  eimbursement to the facility is tied to the number of days of hospitalization rather than by diagnosis-related group or episode of illness, the hospital may not bear any of the financial burden of extended hospital stays and may have few financial incentives for investing in HAIs even if the overall benefits of such investments exceed the costs.

A 1987 study that looked at charges associated with 9,423 nosocomial infections identified in 169,526 admissions, selected randomly from adult admissions to a random sample of U.S. hospitals, found that at least 95 percent of the cost savings obtained from preventing nosocomial infections represented financial gains to the hospital (Haley, White et al. 1987).

However, a series of recent reports from PHC4 find that Medicare and Medicaid bear the greatest burden of the additional cost of HAIs.8 Pennsylvania hospitals billed the federal Medicare program and Pennsylvania’s Medicaid program for 76 percent of the 11,668 hospital-acquired infections in 2004, with Medicare taking up much of the burden. The economic burden on government resources imposed by the additional hospital charges was estimated at $1 billion for Medicare and $372 million for Medicaid. Extrapolating from the figures in Pennsylvania to the entire country, PHC4 estimated that at least $20 billion was charged to Medicare to pay for HAIs during 2004. These figures indicate that hospitals may actually benefit by extending the length of stay and may have fewer incentives to control infection levels within the hospital.

Medicare is currently in the process of revising its rules on reimbursing for hospital-acquired infections, and these changes could have a significant impact on hospital incentives to invest in infection control. Some payers, such as Blue Cross–Blue Shield, have already made some payments contingent on lower rates of certain HAIs, and anecdotal evidence suggests that this has  owered the prevalence of those HAIs.

Impact of lawsuits

Some hospitals have faced lawsuits from individual patients for HAIs, based on plaintiffs’ claims that defendants (hospitals) failed to adhere to the standard of care for infection control.9 A study from Philadelphia found that 72 percent of HAI malpractice cases in Philadelphia were either withdrawn or settled; when brought to trial, the plaintiff was more likely to prevail (Guinan, McGuckin et al. 2005). MRSA infections were the most common reason for lawsuits. Moreover, MRSA in class I surgical sites were more likely to result in a victory for plaintiffs because national data show lower rates of infection for these surgeries, with the implication that these infections were preventable. The impact of lawsuits on infection control is unclear, but they may have made hospitals wary of reporting infection and resistance rates.

Short-term financial considerations

Even if most of the costs of HAIs can be passed on to payers, hospitals and long-term care facilities may bear at least some of the burden associated with the high cost of treating resistant infections. However, even for these limited costs, short-term cost considerations may trump the long-term gains of lower levels of resistance and infection for facilities in financial trouble. Are financially troubled institutions more likely to cut back on infection control? Do hospitals and long-term care facilities really behave optimally, or do they tend to be myopic because they fail to recognize the effect of resistance management and infection control on future costs? Also, to what extent are hospitals prompted by the threat of lawsuits to do a better job of controlling nosocomial infections and resistant pathogens? Answering these questions is pivotal to making policy decisions on how best to incentivize hospitals to invest in stronger infection control programs.

Issues of agency
Although the hospital as a whole may have an incentive to restrict the use of antibiotics and drug resistance, individual clinicians may not share the same incentives. Also, many physicians are not employees but consultants of hospitals and may therefore have a smaller incentive to care about costs imposed by resistance on hospitals. Conversely, the problem of resistance may be evident to infection control committees and clinicians, but they may not be able to convince senior management of the long-term financial benefits of lower levels of resistance. Management and operational structures of hospitals have implications both for investment in infection control and for implementation of control measures, but little is currently known about the influence of organizational culture and structure on infection and resistance levels.10

Incentives to free-ride

Hospital infection control is expensive and becomes more difficult and less effective when patients enter the hospital already carrying the resistant pathogens. Recent research on incentives for hospitals to control HAIs suggests that the large spillovers of antibiotic-resistant bacteria among medical care facilities may be one factor that explains the lack of response (Smith, Levin et al. 2005). When institutions share patients, a person colonized in one facility may be responsible for introducing or increasing the prevalence of resistance in another facility.

Since any single hospital (especially in the current era of cost cutting and short-term financial pressures) may not see the benefits of its HIC program outside its own walls, hospitals may not benefit from decreasing the overall level of resistance in the catchment area when those patients are admitted later to other hospitals. Instead, hospitals may prefer to free-ride on the infection control investments of other hospitals. This results in an overall higher level of resistance.

Modeling shows that the selfishly “optimal” level of HIC that any hospital would undertake is lower the greater the number of hospitals that share a catchment area. In fact, it is in the interests of the hospital to spend less and free-ride on the efforts of other hospitals. When everyone free-rides, all hospitals will spend less on HIC, leading to epidemics that develop earlier and faster. A much better outcome can be achieved through regulation and the resulting coordination between facilities.

A good example comes from the Siouxland experience. An epidemic of VRE in the Siouxland region of Iowa, Nebraska, and South Dakota was first detected in late 1996. Within a short time, VRE had quickly spread to nearly half of the health care facilities in the region. In response, a VRE task force was constituted with representatives from acute care and long-term care facilities and public health departments in the region (Ostrowsky, Tricke et al. 2001). Following a comprehensive two-year intervention (including aggressive culturing to identify VRE-colonized patients, isolation of patients, improved antibiotic use, sterile device measures, improved staff hand hygiene, and sharing of information among institutions), VRE was eliminated from all acute care facilities and significantly reduced in long-term care facilities in the region. This could not have happened without coordination. When hospitals are unwilling to coordinate on their own, regulation will ensure that no single hospital free-rides on the efforts of others. Regulations that require portability of patient records (which could show which patients are colonized) could help hospitals in identifying high-risk carriers of resistant pathogens.

The similarly successful experience of Dutch hospitals in lowering the prevalence of MRSA is described in Box 4.1.

Incentives to report infection levels

Hospitals have a clear incentive to downplay infection levels in their facilities, since accurate reporting could decrease demand for their services. “Report cards” that provide patients with information on hospital quality, including nosocomial infection rates, may encourage hospitals to discriminate against sicker patients or those coming from long-term care facilities because they might be more likely to carry a resistant pathogen.11 To address this problem, Florida and some other states that publicly report outcome indicators by hospital risk adjust the data to account for the fact that some hospitals admit more patients who are sicker and require more resources than the average patient. An alternative strategy would be to monitor and subsidize inputs for hospital infection control rather than monitoring the outputs—that is, infection levels. Educational efforts to get hospitals to recognize the long-term gains of infection control may also be part of the solution.

Hospital report cards also should be issued by an independent agency that is less susceptible to political pressure. These reports, if issued by government agencies, can be influenced or quashed by interference from the governor or state senators, who in turn are influenced by campaign contributions from wealthy doctors. Governmental policy can also influence the timing of the release of reports.

Some degree of enforcement is required, via periodic external surveillance cultures, withdrawal of approval for state Medicare reimbursement, or fines. Because reporting requirements can create perverse incentives—for example, hospitals that suspect high levels of resistance may cut back on surveillance expenditures—any reporting program needs to be designed to take these factors into consideration.


Hospitals are an important reservoir for resistant pathogens, and the problem of resistant infections is emblematic of broader problems with ensuring health care quality. The issue is not knowing how to address resistant infections in hospitals12—good examples exist, from both the United States and abroad, of how to maintain low levels of resistance in health care settings—but rather, understanding why some facilities have an incentive to invest in these programs while others do not.

Regulatory agencies play two important roles in the antibiotic resistance problem. One is to enable cooperative outcomes better than those attained if hospitals behave in their own self-interest. Regional coordination in infection control efforts may be one of several solutions to this dilemma (Kaye, Engemann et al. 2006). Another is to make public the data on resistance and infection levels so that hospitals have an incentive to invest in addressing the problem. Here we propose ways to encourage reporting and control of resistant infections and improve surveillance, and we also recommend additional research.




Bergstrom, C. T., M. Lo, et al. (2004). “Ecological Theory Suggests that Antimicrobial Cycling Will Not Reduce Antimicrobial Resistance in Hospitals.” Proceedings of the National Academy of Sciences of the United States of America 101(36): 13285-90.

Carmeli, Y., G. Eliopoulos, et al. (2002). ”Health and Economic Outcomes of Vancomycin-Resistant Enterococci.” Archives of Internal Medicine 162(19): 2223-2228.

Cosgrove, S. E., Y. Qi, et al. (2005). “The Impact of Methicillin Resistance in Staphylococcus aureus Bacteremia on Patient Outcomes: Mortality, Length of Stay, and Hospital Charges.” Infection Control and Hospital Epidemiology 26: 166-174.

CU. (2006). (accessed October 18, 2006). Consumers Union.

Dranove, D., D. P. Kessler, et al. (2002). “Is More Information Better? The Effects of ’Report Cards’ on Health Care Providers.” Working paper W8697. Cambridge, MA: NBER (National Bureau of Economic Research).

Engemann, J. J., Y. Carmeli, et al. (2003). ”Adverse Clinical and Economic Outcomes Attributable to Methicillin Resistance among Patients with Staphylococcus aureus Surgical Site Infection.” Clinical Infectious Diseases 36: 592-598.

Gaul, G. M. (2005). “Accreditors Blamed for Overlooking Problems: Conflict of Interest Cited Between Health Facilities, Group That Assesses Conditions.” Washington Post. Washington, DC, July 25, A01.

Guinan, J. L., M. McGuckin, et al. (2005). “A Descriptive Review of Malpractice Claims for Health Care-Acquired Infections in Philadelphia.” American Journal of Infection Control 33(5): 310-2.

Haley, R.W. (1986). Managing Hospital Infection Control for Cost Effectiveness: A Strategy for Reducing Infectious Complications. Chicago: American Hospital Publishing, Inc.

Haley, R. W., D. H. Culver, et al. (1985). “The Efficacy of Infection Surveillance and Control Programs in Preventing Nosocomial Infections in US Hospitals.” American Journal of Epidemiology 121(2): 182-205.

Haley, R. W., J. W. White, et al. (1987). “The Financial Incentive for Hospitals to Prevent Nosocomial Infections under the Prospective Payment System. An Empirical Determination from a Nationally Representative Sample.” JAMA 257(12): 1611-4.

Harrison, P. F. and J. Lederberg (eds.). (1998). Antimicrobial Resistance: Issues and Options, Workshop Report. Forum on Emerging Infections. Washington, DC: Institute of Medicine.

Jevons, M. (1961). “Celbenin-Resistant Staphylococci.” BMJ 1: 124-5. Kaye, K. S., J. J. Engemann, et al. (2006). “Favorable Impact of an Infection Control Network on Nosocomial Infection Rates in Community Hospitals.” Infection Control and Hospital Epidemiology 27(3): 228-32.

Lai, K. K., S. P. Baker, et al. (2003). “Impact of a Program of Intensive Surveillance and Interventions Targeting Ventilated Patients in the Reduction of Ventilator-Associated Pneumonia and Its Cost-Effectiveness.” Infection Control and Hospital Epidemiology 24(11): 859-863.

Laxminarayan, R., D. L. Smith, et al. (2005). ”On the Importance of Incentives in Hospital Infection Control Spending.” Discovery Medicine 5(27): 303-308.

Martone, W. J. and R. L. Nichols. (2001). “Recognition, Prevention, Surveillance, and Management of Surgical Site Infections: Introduction to the Problem and Symposium Overview.” Clinical Infectious Diseases 33: S67-S68.

McHugh, C. G. and L. W. Riley. (2004). “Risk Factors and Costs Associated With Methicillin-Resistant Staphylococcus aureus Bloodstream Infections.” Infection Control and Hospital Epidemiology 25(5): 425-430.

McKibben, L., T. Horan, et al. (2005). “Guidance on Public Reporting of Health care-Associated Infections: Recommendations of the Health care Infection Control Practices Advisory Committee.” American Journal of Infection Control 33(4): 217-26.

Muto, C. A., E. T. Giannetta, et al. (2002). ”Cost-Effectiveness of Perirectal Surveillance Cultures for Controlling Vancomycin-Resistant Enterococcus.” Infection Control and Hospital Epidemiology 23(8): 429-435.

Muto, C. A., J. A. Jernigan, et al. (2003). “SHEA Guideline for Preventing Nosocomial Transmission of Multidrug-Resistant Strains of Staphylococcus aureus and Enterococcus.” Infection Control and Hospital Epidemiology 24(5): 362-386.

Ostrowsky, B. E., W. E. Trick, et al. (2001). “Control of Vancomycin-Resistant Enterococcus in Health Care Facilities in a Region.” The New England Journal of Medicine 344(19): 1427-1433.

OTA. (1995). Impact of Antibiotic-Resistant Bacteria: A Report to the U.S. Congress. Washington, DC: Government Printing Office. Office of Technology Assessment.

PHC4. (2005). “Reducing Hospital-Acquired Infections: The
Business Case.” Research Brief No. 8. (accessed March 5, 2007). Pennsylvania Health Care Cost Containment Council.

PHC4. (2006). Hospital-acquired Infections in Pennsylvania 2005. (accessed December 11, 2006). Pennsylvania Health Care Cost Containment Council.

Pittet, D., D. Tarara, et al. (1994). “Nosocomial Bloodstream Infection in Critically Ill Patients. Excess Length of Stay, Extra Costs, and Attributable Mortality.” JAMA 271(20): 1598-1601.

Pittet, D. and R. P. Wenzel. (1995). “Nosocomial Bloodstream Infections. Secular Trends in Rates, Mortality, and Contribution to Total Hospital Deaths.” Archives of Internal Medicine 155(11): 1177-1184.

RIVM. (2005). “European Antimicrobial Resistance Surveillance System (EARSS ).” (accessed on January 18, 2005). The National Institute for Public Health and the Environment.

Sepkowitz, K.A. (1996a). “Occupationally Acquired Infections in
Health Care Workers: Part I.” Annals of Internal Medicine 125(10): 826-834.

Sepkowitz, K.A. (1996b). “Occupationally Acquired Infections in
Health Care Workers: Part II .” Annals of Internal Medicine 125(11): 917-928.

Smith, D. L., S. A. Levin, et al. (2005). “Strategic Interactions in Multi-Institutional Epidemics of Antibiotic Resistance.” Proceedings of the National Academy of Sciences of the United States of America 102(8): 3153-8.

Sohn, A. H., B. E. Ostrowsky, et al. (2001). “Evaluation of a Successful Vancomycin-Resistant Enterococcus Prevention Intervention in a Community of Health Care Facilities.” American Journal of Infection Control 29(1): 53-57.

Song, X., A. Srinivasan, et al. (2003). ”Effect of Nosocomial Vancomycin-Resistant Enterococcal Bacteremia on Mortality, Length of Stay, and Costs.” Infection Control and Hospital Epidemiology 24(4): 251-256.

Stosor, V., L. R. Peterson, et al. (1998). ”Enterococcus faecium Bacteremia: Does Vancomycin Resistance Make a Difference?” Archives of Internal Medicine 158(5): 522-527.

Vos, M. C., A. Ott, et al. (2005). ”Successful Search-and-Destroy Policy for Methicillin-Resistant Staphylococcus aureus in The Netherlands.” Journal of Clinical Microbiology 43(4): 2034-2035.

Vriens, M., H. Blok, et al. (2002). “Costs Associated with a Strict Policy to Eradicate Methicillin-Resistant Staphylococcus aureus in a Dutch University Medical Center: A 10-Year Survey.” European Journal of Clinical Microbiology and Infectious Diseases 21(11): 782-786.

Wagenvoort, J. H. (2000). “Dutch Measures to Control MRSA and the Expanding European Union.” Euro Surveillance 5(3): 26-28.

Wenzel, R. P. (1995). “The Lowbury Lecture. The Economics of Nosocomial Infections.” The Journal of Hospital Infection 31(2): 79-87.

West, T. E., C. Guerry, et al. (2006). “Effect of Targeted Surveillance for Control of Methicillin-Resistant Staphylococcus aureus in a Community Hospital System.” Infection Control and Hospital Epidemiology 27(3): 233-238.

WIP. (2005). Policy for Methicillian-resistant Staphylococcus aureus. (accessed May 31, 2006). Dutch Workingparty Infection Prevention.




1. Hospital reimbursement policies for HAIs could be linked to levels of infection and drug resistance. Tying Medicare and private insurance payments to a hospital to its levels of infection control may be one approach.

2. Subsidizing inputs for infection control and surveillance programs would provide a greater incentive for hospitals to invest in them. Chapter 6, on health insurance and Medicare, describes such a program.

3. State requirements for reporting of hospital infections should adjust the data for risk so that hospitals that admit sicker patients are not penalized for having higher levels of antibiotic use and infection.

4. The national hospital infection and resistance surveillance system should be more comprehensive. Ideally, it would be separate from JCAHO and other accreditation groups and would take the approach used by several states: it would collect nationwide data not just on outcomes (infections and resistance) but also on inputs, such as antibiotic use, number of infection control nurses, and physical inputs for HIC. Given the incentive problems with reporting outcomes, independent monitoring and reporting of infections should be complemented with reports on infection control inputs.

5. Legal avenues for responding to resistance should be examined, perhaps involving a combination of workplace safety and labor laws (e.g., penalizing hospitals for a failure to protect nursing staff if they are at risk). Studies indicate that nurses are at-risk for infections caused by C. difficile and E. coli, however this risk is believed to be low (Sepkowitz 1996a; 1996b).

6. Research needs to address the important policy-relevant questions. Little is known about the institutional characteristics (ownership structure,13 proximity to other hospitals and facilities) that predict resistance. We also know little about the costs of surveillance and infection control for a typical hospital and how these compare with other hospital expenses. Additional data will help determine the burden of infection control on hospital budgets and inform the design of taxes and subsidies for specific inputs for infection control.

7. A policy research program is needed to explore how to create incentives for hospitals to conduct surveillance and reporting, not just of infections but also of other important health care quality measures.






1 Many of the procedures commonly performed on the seriously ill today, such as central venous catheterization and mechanical ventilation, predispose the patient to colonization with hospital-associated bacteria and an enhanced susceptibility to invasive infection with these agents.

2 There are others, such as physician education, that are discussed in Chapter 3.

3 Although the problem of MRSA (and other HAI s) in nursing homes and prisons is not addressed in this chapter, a number of the recommendations made here are applicable to those situations as well.

4 Based on a conversation with Dennis O’Leary, vice president, JCAHO, November 28, 2005.

5 In fact, concerns have been raised about the rigor of JCAHO ’s hospital surveys and its ability to catch even gross violations that have seriously compromised patient health.

6 And, some hospitals do have such an incentive, as seen in studies reviewed earlier in this chapter.

7 This refers to assigning hospital staff to a limited number of patients rather than allowing for unlimited contact between health care workers and patients, which increases the likelihood of infection spread.

8 The PHC 4 reports were based on a state law that required hospitals to submit data on some categories of HAI s to PHC 4 starting January 2004. Starting January 1, 2006, nearly all hospital-acquired infections are reportable to PHC 4 (PHC 4 2006).

9 Media reports of MRSA -related lawsuits are growing. In one recent example, the families of two women who died from MRSA infections while incarcerated at the jail in Allegheny County, Pennsylvania, sued the warden and other county officials for failing to provide medical care.

10 Hospital objectives may be multifaceted. Many participants at our consultations agreed that although hospital managers care about reducing infection-related mortality, they are less adept at seeing the long-term health and economic benefits of infection control. Some of the shortsightedness is reflected in compensation of infectious disease clinicians and nurses: an infection control nurse typically earns less than a bedside nurse and consequently there is a shortage in supply.

11 A related study in the context of cardiac surgery found that the use of hospital report cards in New York and Pennsylvania led to improved matching of patients with hospitals, but also gave doctors and hospitals an incentive to turn away severely ill patients who were more difficult to treat. This resulted in higher levels of resource use and worse health outcomes, particularly for sicker patients (Dranove, Kessler et al. 2002).

12 Although much of the problem with drug resistance in hospitals is related to lack of sufficient infection control rather than to excessive antibiotic use, hospitals have tended to focus on the antibiotic use issue to a greater extent. Some hospitals have pursued cycling and other antibiotic restriction policies even though ecologists have questioned the soundness of these strategies (Bergstrom, Lo et al. 2004).

13 Categories include government hospitals, for-profit hospitals, nonprofit teaching hospitals, and nonprofit nonacademic hospitals.