Skip Navigation

The Use of Active Surveillance Cultures in Adult Intensive Care Units to Reduce Methicillin-Resistant Staphylococcus aureus-Related Morbidity, Mortality, and Costs: A Systematic Review

  1. Katharine L. McGinigle1,2,
  2. Margaret L. Gourlay1,2,3, and
  3. Ian B. Buchanan2,4
  1. 1School of Medicine, Chapel Hill, North Carolina
  2. 2Public Health Leadership Program, School of Public Health, Chapel Hill, North Carolina
  3. 3Department of Family Medicine, School of Medicine, University of North Carolina, Chapel Hill, North Carolina
  4. 4Department of Performance Improvement, UNC Health Care, Chapel Hill, North Carolina
  1. Reprints or correspondence: Katharine McGinigle, Public Health Leadership Program, Health Care and Prevention Division, University of North Carolina-Chapel Hill, Rosenau Hall, CB7469, Chapel Hill, NC 27599 (kmcginig{at}unch.unc.edu)

Abstract

Active surveillance cultures (ASCs) are universal or targeted microbiological screening cultures for patients admitted to a hospital. ASCs have been proposed to control the increasing numbers of infections due to multidrug-resistant organisms, but their efficacy and cost-effectiveness are unproven. We conducted a systematic review of the literature pertaining to the use of ASCs and control of methicillin-resistant Staphylococcus aureus (MRSA). We searched revelant journals and the PubMed Medline, Web of Science, CINAHL, and Cochrane Library databases. No randomized, controlled trials were identified. Sixteen observational studies and 4 economic analyses were reviewed. Only 2 of the observational studies had a control group. None of the studies were of good quality. Thus, we identified important gaps in the literature, including a need for a clear definition of ASCs, a clear implementation protocol, and rigorous economic evaluations. Existing evidence may favor the use of ASCs, but the evidence is of poor quality, and definitive recommendations cannot be made.

Hospital-acquired infections are a common and serious public health problem, and their management and control are essential to minimize hospital-related morbidity and mortality [1]. In American hospitals, there are ∼2 million nosocomial infections per year, resulting in an estimated 99,000 deaths [2]. The microorganisms responsible for these infections are often resistant to multiple antibiotics, and the resistance problem continues to grow [3, 4]. Currently, >60% of patients with Staphylococcus aureus infection in intensive care units (ICUs) have methicillin-resistant S. aureus (MRSA) isolated [5]. MRSA bacteremia and surgical site infection have been associated with higher mortality and higher health care costs than is the case for hospital-acquired, methicillin-susceptible staphylococcal infections [68]. Although it is difficult to precisely quantify infection-attributable complications because they often underlie other disorders or procedures, it is clear that hospital-acquired infections do contribute to increased morbidity, mortality, and health care costs [9].

In 2003, the Society for Healthcare Epidemiology of America (SHEA) published a special report titled “SHEA Guideline for Preventing Nosocomial Transmission of Multidrug-Resistant Strains of Staphylococcus aureus and Enterococcus” [3]. The recommendations in the SHEA guideline included strong support for the use of active surveillance cultures (ASCs). These ASC recommendations were not based on a systematic review, but stemmed from evidence accessed through the PubMed Medline database (1966–2002; unreported search strategy) and unpublished literature in the authors' libraries. The 2003 SHEA guideline was the first recommendation for ASCs from a major infection control organization, and it created controversy within the field [10]. Professional organizations and infection control officers continue to question whether ASCs (compared with no screening cultures or usual care) can decrease MRSA-related mortality, the incidence of MRSA infection, and the cost of MRSA-related care [11, 12].

Methods

The following search strategy combined 2 searches using keyword and text word combinations in the PubMed Medline database (1966 through September 2007), the Web of Science database (1955 through September 2007), CINAHL database (1982 through September 2007), and the Cochrane Library database. The first search identified screening-related articles: “Methicillin Resistance” OR “methicillin resistant” OR “methicillin resistance” as a group was combined with “staphylococcal infections” OR “Staphylococcus aureus” OR “MRSA” and also combined with a group of screening-related terms “sentinel surveillance” OR “surveillance” OR “mass screening” OR “screening” OR “active screening” OR “active surveillance.” The second search identified cost-related articles by adding relevant keywords and text words to the same base of antibiotic and bacteria keywords and text words used in the first search: “Methicillin Resistance” OR “methicillin resistant” OR “methicillin resistance” as a group was combined with “staphylococcal infections” OR “Staphylococcus aureus” OR “MRSA” and also combined with a group of cost-related terms “economics” OR “cost and cost analysis” OR “cost.”

A hand search for guidelines and cited original articles was conducted on the Centers for Disease Control and Prevention and Institute for Healthcare Improvement Web sites. Five major journals in the field of infection control were also hand-searched for relevant articles published during 2000–2007; these journals were Journal of Infectious Disease, Clinical Infectious Diseases, Journal of Hospital Infection, Infection Control and Hospital Epidemiology, and the American Journal of Infection Control. Finally, the reference lists of all included articles were hand searched.

On the basis of an exploratory search of the literature, inclusion and exclusion criteria were established (table 1). The study population was limited to patients in ICUs who were screened for MRSA at the time of admission and at least weekly thereafter [1315]. The primary outcomes chosen were either patient centered (i.e., MRSA transmission rates, MRSA infection rates, MRSA-related mortality, or all-cause mortality) or related to cost-benefit. The body of literature on ASCs includes few studies with rigorous study designs; therefore, broad inclusion criteria for study design were set. More emphasis was placed on studies that were well designed with appropriate control groups, but uncontrolled before-and-after studies and ecological studies were also reviewed to cover the literature on which current recommendations have been made.

Figure 1

QUORUM tree of articles selected for systematic review

Table 1

Inclusion and exclusion criteria for a literature search regarding active screening for methicillin-resistant Staphylococcus aureus.

The articles identified in the search were independently reviewed by one author (K.L.M.) and a research assistant to identify those that met the inclusion and exclusion criteria. Any disagreements were resolved by a third reviewer (M.L.G.). The initial search yielded 2578 articles. After review of the title plus abstract or full text, 2558 articles did not meet all inclusion criteria and were excluded from the review. Twenty articles were included in the review. A QUORUM tree illustrates the process used to select the studies in the final review (figure 1).

The quality criteria used to assess the internal validity of each of the included observational studies were based on guidelines set forth by the United Kingdom's National Health Service Centre of Reviews and Dissemination [16]. The quality of the economic analyses was rated based on guidelines from the US Preventive Services Task Force [17]. Studies that met all eligibility criteria were summarized and organized into an evidence table and assigned a good/fair/poor evidence rating. All data were organized according to methodological quality rating and publication year, with the economic analyses presented separately (table 2).

Table 2

Summary of studies meeting inclusion criteria according to evidence rating.

Results

Twenty studies that examined active surveillance cultures and incidence of hospital-acquired MRSA or MRSA-associated costs fulfilled the eligibility criteria for this systematic review. There were no randomized, controlled trials. Sixteen observational studies were identified; only 2 of these studies (12.5%) included a control group. In all of the studies, the background rate of endemic MRSA at each institution was either known or measured at the onset of the study period. None of the studies included in this systematic review were of “good” quality. The maximum rating that uncontrolled studies could obtain was “fair,” because of the inherent weaknesses in an uncontrolled study design. All of the before-and-after studies and ecological studies were of poor quality (table 3).

Table 3

Methodological quality by study design.

Effectiveness of ASCs. Sixteen observational studies provided information on the effectiveness of ASCs as a method of reducing hospital-acquired MRSA infection [1833]. Six of the studies defined the primary outcome as MRSA transmission and measured incident MRSA colonization, as well as infection [18, 27, 28, 3032]. The other 10 studies used incident MRSA infection as the primary outcome. Nine of the studies were conducted in Europe, 5 were conducted in the United States, 1 was conducted in Brazil, and 1 was conducted in Israel. Three of the studies were conducted at multiple institutions [18, 25, 34].

All 16 of the studies used microbiological samples from the nares as part of the ASC intervention, but one-half also took samples from other sites, including the throat, groin, rectum, and open wounds. The studies that reported culturing methods (69%) all used standard methods and did not use rapid PCR for MRSA detection. In all of the studies, isolation precautions were used only after MRSA-colonized patients were detected. Approximately one-half of the studies also reported using topical decolonization therapy, most commonly nasal mupirocin ointment and chlorhexidine baths, for MRSA-colonized patients [20, 21, 23, 24, 2933].

Thirteen of the 16 observational studies, including both controlled studies, reported a decrease in the incidence of hospital-acquired MRSA infections in association with the use of ASCs [1824, 26, 27, 29, 30, 32, 33]. Huang et al. [19] reported the most striking results: a 75% decrease in MRSA infections in the ICU where the ASC program was being used and a 40% decrease in MRSA infections in the remainder of the hospital that was not receiving the intervention. The majority of the other studies (66%) showed a 40%–60% decrease in hospital-acquired MRSA incidence associated with ASCs in an ICU setting [21, 23, 24, 26, 27, 29, 32, 33]. However, these articles were limited by study design (e.g., none had control groups, few measured baseline characteristics), possible confounding, and failure to monitor or report intervention integrity and compliance.

Three studies reported negative findings and did not find an association between ASCs and MRSA incidence. Two of the studies reported no change in the MRSA infection rate [25, 31]. The third study specifically separated the surveillance cultures from other subsequent patient management, such as contact isolation, and did not report the results to hospital staff [28]. Over a 3-month period, the mean daily prevalence of MRSA was no different from that during a surrounding 3-year period. Further, there was no cross-contamination of methicillin-susceptible S. aureus or MRSA between patients in the ICU, and the authors conclude that had the ASC program been followed up with isolation precautions, ASCs would have appeared falsely successful and were not a useful infection control strategy [28].

Cost analyses of ASCs. Four economic analyses provided information regarding the costs of ASCs and MRSA infections [3437]. The studies were performed in Canada, France, Germany, and Spain. Two of the 4 economic evaluations were true cost-benefit analyses with good internal validity and appropriate statistical analysis, including the use of a base year to account for differential timing and sensitivity analyses [34, 35]. The primary weakness in both cost-benefit analyses was that they were based on the effectiveness levels found in poor- to fair-quality studies conducted by the authors rather than the best available evidence of effectiveness of ASCs.

In all 4 studies, a variety of costs related to ASCs were examined, including swabs, gowns, gloves, masks, disinfectant hand wash, culturing materials, laboratory personnel, nursing time, and environmental cleaning. Costs of hospital-acquired MRSA infections were based on the number of hospital days attributable to MRSA and daily charges in hospital accounting records. All of the analyses were reported from an institutional perspective. One cost-benefit analysis reported that universal—not just high-risk—screening is cost-beneficial if the importation rate of MRSA into an ICU is >1% [34]. Two studies reported that ASC programs are economically justified if they can prevent 3 or 4 cases of MRSA infection [35, 36].

Discussion

Our review of the 5 fair-quality observational studies and the 2 cost-benefit analyses showed fair- to poor-quality evidence that ASCs were associated with a decreased incidence of hospital-acquired MRSA infections. Although the 2 cost-benefit analyses showed favorable results, they are based on poor-quality evidence and do not provide adequate justification for routine implementation of ASCs in hospital ICUs.

Only 2 of the articles reported using a control group [18, 22], and none of the studies used randomly assigned controls, so there was a high possibility of confounding. Potential confounders include compliance with taking swabs for culture; compliance with isolation precautions, including hand hygiene and glove and gown use; length of time before reported culture results; patient length of stay; patient case mix; staffing levels; use of prophylactic antibiotics; and the endemic level of community-acquired MRSA.

Many infection control specialists, epidemiologists, and methodologists have called for a randomized, controlled trial to obtain reliable results and stronger guidance [38]. A large, multicenter, randomized, controlled trial that uses hospitals as the unit of analysis could be useful to clarify the efficacy of ASCs. However, negative findings might not halt implementation of ASCs in ICUs under administrative pressure to demonstrate some attempt to control MRSA. In fact, some state legislatures have already considered passing laws requiring ASCs in hospitals [3942].

Many Scandinavian countries have been using ASCs combined with patient isolation for years on the basis of the belief that such methods aid in controlling MRSA infections, despite the absence of high-quality studies or randomized, controlled trials demonstrating efficacy. The Dutch Working Party on Infection Prevention has set strict guidelines for the control of MRSA that are largely centered on “search and destroy” and include isolation of patients in single rooms until proven to be negative for MRSA, daily screening of health care workers, and unit closures when 2 patients are MRSA positive [43]. The prevalence of MRSA in The Netherlands is <1% [43, 44].

Many other European countries use ASCs and subsequent isolation to control MRSA cross-contamination; however, the aforementioned method from The Netherlands is the most aggressive [43]. It is unclear whether such strict treatment of patients and health care workers in the United States would be acceptable or affordable. In addition to varied health care practices, the marked difference in levels of endemic MRSA in hospitals within the United States, Europe, and other parts of the world makes comparisons among countries and institutions difficult and limits the external validity of the studies in this review.

Within the United States, there is no standard definition of what constitutes an “active surveillance culture,” and it is unclear what population should be screened for MRSA colonization. Possible options would be to screen all patients admitted to the hospital or only “high-risk” patients, which could include patients being readmitted, admitted to an ICU, or transferred from another facility, as well as patients with high-risk medical conditions, a history of MRSA infection, or frequent medical facility contact (e.g., hemodialysis) [5, 13]. The 2003 Society for Healthcare Epidemiology of America guideline states that ASC samples should be taken from the nares and areas of skin breakdown and that these screening cultures should be performed at admission and weekly thereafter. However, the guideline also states that the frequency of culturing may need to be adjusted based on the patient population [3]. There is minimal evidence regarding how patient-specific risk factors and endemic MRSA rates may affect the time to MRSA acquisition and the frequency of screening.

The high level of variability in culturing method, time, and cost makes it difficult to assess each hospital's ability to accommodate increased demand for culturing if ASCs are implemented. Kunori et al. [45] report that the most cost-effective screening method is to inoculate a single nasal sample directly onto ciprofloxacin Baird-Parker agar without the use of broth and then confirm the result with a staphylococcal latex test (e.g., Pastorex Staph-Plus; Bio-Rad), but not with any methicillin resistance test. However, different strains of MRSA may require different culturing methods. In the future, rapid screening with PCR may be advantageous because the lag time to identifying MRSA-positive patients who need to be isolated will be shortened, thereby increasing the effectiveness of the program. At this time, the value of PCR remains unclear, because there are concerns about test sensitivity, high cost, and the need for technically skilled laboratory workers. Notably, none of the studies included in this review reported using PCR for MRSA detection.

Although the literature—including many randomized, controlled trials reviewed by the Cochrane Collaboration [46]—does not support the routine use of topical or systemic medications for MRSA decolonization, almost one-half of the studies included in this review used decolonization as part of their intervention program. Despite the existence of sufficient literature with clearly reviewed results, there is still variation in the use of decolonization as a part of an ASC program, further illustrating the lack of standardized ASC definitions.

It is also unclear how isolation precautions should be used in conjunction with ASCs. One option is that all patients in the screening population have isolation precautions implemented until it is proven that they are negative for MRSA, but another option is that precautions are implemented only after it is shown that a patient is positive for MRSA. It could be logistically difficult in many ICUs to use presumptive precautions for all patients until the MRSA screen is available. Furthermore, contact isolation may negatively affect the quality of care received by patients.

In addition to the increased financial burden to the hospital that comes with isolation precautions for higher numbers of patients, there are also patient-centered “costs” to isolation. Kirkland and Weinstein [47] conducted a prospective, observational study and reported that health care workers treating ICU patients who are under contact precautions were ∼2 times less likely to enter the room than were health care workers taking care of ICU patients without contact precautions. Evans et al. [48] report similar results and found that isolated patients were visited 5.3 times/h, compared with 10.9 visits/h for nonisolated patients (P<.001). In addition, despite more severe disease as designated by higher mean APACHE II scores (±SDs) among isolated patients (10.1±1.0 vs. 7.6±0.8; P=.05), the overall contact time with health care workers was significantly lower (29±5 vs. 37±3 min/h; P=.008) [48].

Few of the studies included in this review reported levels of compliance with performing nasal swabs for the active surveillance of MRSA. The success of ASCs also hinges on compliance with subsequent isolation, contact precautions, and increased hygiene and cleaning that try to reduce the possibility of cross-contamination. Observational studies have reported compliance rates of ∼40% for hand hygiene recommendations [49], 60%–75% for isolation precaution guidelines [50], and 28% for compliance with all MRSA precautions [51]. With such poor levels of hand hygiene and contact precaution compliance, it is not certain that ASCs and contact precautions will be effective as their use becomes widespread across a variety of types of hospitals. It is unclear what level of compliance with other infection control practices is required for ASCs to be useful.

Our review has some limitations. Publication bias is of particular concern in this review because of the large number of poor-quality studies. Studies with negative findings, especially those of fair or poor quality, are not as likely to be published, and it is possible that the preponderance of studies supporting ASCs present a falsely inflated positive evidence base. All of the components of the search were limited to English; however, this bias is likely minimal because a number of studies (60%) published in English by researchers from non-English speaking countries have been included in this review.

Evidence from multiple observation-based studies suggests that the use of ASCs reduces the incidence of MRSA infection, but the overall quality of the evidence is poor; thus, definitive, evidence-based clinical recommendations cannot be made. An unambiguous definition of ASCs, a clear implementation protocol including defined screening groups and laboratory methods, and rigorous economic evaluations are all lacking. In their most recent guidelines, highly respected professional organizations, such as the Society for Healthcare Epidemiology of America and Institute for Healthcare Improvement, have strongly recommended using ASCs; however, there is no consensus among infection control experts on this issue. A randomized, controlled trial to prove the efficacy of ASCs would be helpful but may not be feasible because of high cost. Instead, it may be better to make the best use of available evidence that demonstrates the usefulness of ASCs in reducing MRSA infections and MRSA-related costs and to focus further research on defining the details (risk assessment, lower-cost rapid culture methods or PCR, and treatment for colonized patients) for effective ASC implementation.

Acknowledgments

We acknowledge Sarah Brier for her contribution as the second reviewer.

Potential conflicts of interest. All authors: no conflicts.

  • Received October 10, 2007.
  • Accepted January 9, 2008.

References

| Table of Contents