Contact precautions (CPs), the use of gloves and gowns, have been widely used for many years to prevent pathogen transmission between patients in health care facilities. The US Centers for Disease Control and Prevention has recommended CPs for patients infected with multidrug-resistant organisms for decades, including in its most recent guidance.1 Despite this recommendation, adherence is often variable, owing largely to perceptions that CPs are burdensome to use. Observational studies also suggest that CPs are costly, reduce health care worker interactions with patients, and increase anxiety and depression among patients.
Use of CPs is not supported by a strong clinical trial evidence base. Large clinical trials, such as the Benefits of Universal Glove and Gown study, have failed to clarify the issue, with results suggesting that CPs may have a beneficial effect for preventing transmission of some organisms (eg, methicillin-resistant Staphylococcus aureus[MRSA]) but not others (eg, vancomycin-resistant Enterococci [VRE]).2,3 As a result, more selective use of CP in settings of endemic MRSA and VRE has been suggested, along with reconsideration of legal mandates requiring active surveillance.4
However, the use of CP has had and continues to have an important role in clinical care and infection control. Patients colonized or infected with MRSA are the most common focus of CP and the target of several US state mandates and national initiatives, including the search and destroy policy implemented in the Netherlands in 1988, the enhanced mandatory surveillance program initiated in the United Kingdom in 2005, and the nationwide MRSA Prevention Initiative implemented at all Veterans Affairs acute-care hospitals in 2007. These initiatives incorporated CP into a bundle of infection-prevention practices that included enhanced emphasis on hand hygiene, active surveillance for MRSA and, in some cases, MRSA decolonization. As more of these initiatives were implemented, a coincident trend was manifest: a decline in health care–associated MRSA infections over the past 10 to 12 years in the United States and Europe. This decline occurred whether the outcome observed was MRSA incidence rate, MRSA hospitalization rate, in-hospital mortality associated with MRSA, MRSA acquisition rate, MRSA admission prevalence, or the percentage of clinical S aureus isolates resistant to methicillin. This decline is a major accomplishment of infection-prevention practice, even if it remains uncertain which specific practices were most responsible.
Why is estimating the effectiveness of CP so challenging? First, CPs rarely are used as a sole infection-prevention intervention, and the attributable effect of CP on outcomes such as infection rates is difficult to isolate in observational studies. The same is true for most clinical trials in which CPs are often used as 1 component of a multifaceted infection-prevention strategy. However, several additional factors limit the interpretation of the literature around the effectiveness of CP to prevent pathogen transmission. These include the following:
Measurement of study outcome: Observational studies and clinical trials typically measure outcomes such as MRSA acquisitions or infections for the facility or hospital ward being studied. Given the potential delays between organism acquisition (deposited on skin or ingested) and colonization, and between colonization and development of clinical infection, studies in hospitals with a typical average length of stay of 4 days will miss acquisitions and infections that are discovered after patient transfer or discharge and therefore underestimate the effect of CP. Similarly, because MRSA infection is much more likely to occur in patients already colonized at admission (which CP will not prevent), hospital-onset infection as a study outcome makes sense only when admission colonization status is measured and patients are monitored well after transfer or discharge.
Insufficient study duration or scope: Similarly, studies rarely account for the long-term flow of patients between health care facilities and their communities, and the downstream effect of infection prevention measures rates of pathogens introduced into facilities due to patient readmissions. These beneficial effects on transmission may be too small to detect statistically during the course of a study but may accumulate over longer periods of time (possibly years) owing to longitudinal changes in importation.
Patient movement between facilities: Studies also typically do not account for the movement of patients between nonstudy health care facilities, either through patient transfers or admission to an alternate regional facility, each of which may have different, overlapping, or nonexistent infection-prevention strategies. The resulting patient flow networks can have a significant effect on facility importation rates, obscuring the interpretation of studies performed at a single facility.
Measurement and interpretation of intervention use: What is interpreted as the effectiveness (or ineffectiveness) of an intervention such as CP is often more accurately the difference in the frequency of CP use between study groups. An example is the STAR*ICU trial, which did not measure the difference between CP use and no CP use, but rather the difference between 51% of patient days receiving CP in the intervention group vs 38% of patient-days receiving CP in the control group.5 The difference in CP use between study groups was thus quite small.
Likewise, studies on the discontinuation of CP also have significant methodological issues that hinder the interpretation of their results. For instance, a recent meta-analysis reviewed 14 studies that evaluated the discontinuation of CP for multidrug-resistant organisms and concluded that there is no evidence that discontinuation increased infection rates.6 However, nearly all of the studies assessed had important limitations and additional methodologic weaknesses such as before-and-after study designs, concurrent changes in infection prevention practice, flawed outcome measures, and small sample sizes. Clearly, more and better evidence is needed before drawing conclusions about the safety of discontinuing CP.
Similar to other established, noncontroversial infection-prevention practices such as hand hygiene, environmental cleaning and disinfection, antimicrobial stewardship, and bundled interventions to prevent central line–associated bloodstream infections, there are compelling, rational reasons that suggest CPs have an important role in preventing pathogen transmission.
First, gloves and gowns are frequently contaminated when health care workers interact with patients known to be infected with multidrug-resistant organisms such as MRSA—contamination that would be on the health care worker’s clothing and hands without gloves and gowns. Without CP use, some proportion of contaminated health care workers will transmit these organisms to subsequent patients, and for MRSA and other multidrug-resistant organisms, a significant proportion of patients in whom colonization has occurred will later progress to having clinical infection.
Second, adherence with hand hygiene among health care workers is lackluster, and despite great effort, is unlikely to ever reach levels that could replace the positive effects of CP on reducing transmission. Furthermore, glove use appears to increase hand hygiene adherence, conferring an additional benefit of CP use.
Third, CP may be burdensome to health care workers and may affect patient flow in health care facilities, but the highest-quality evidence suggests CPs do not cause more adverse events and do not increase depression or anxiety in patients.
Fourth, the cost of gowns and gloves is minimal compared with other medical costs.
It is important to improve the evidence surrounding the use of CP, such as by designing more appropriate studies and using more tools to understand the dynamics of pathogen transmission and prevention in health care settings. However, large, cluster-randomized trials of CP effectiveness may not necessarily be the answer. The very large sample sizes and the strategies required to address complex issues, such as postdischarge consequences of new acquisition, may make such studies difficult and impractical, if not completely infeasible. Examining other surrogate outcomes, such as contamination measured in the environment or on health care worker clothing, might be a more reasonable approach.
Current evidence suggests that more rigorous evaluation of infection-prevention practices is needed before easing the use of CP in settings of endemic MRSA and VRE. Recent success with MRSA control may owe a debt to CP use, and it would be an error to accept the results from a few small, limited studies as a guarantee that these changes to MRSA rates will not reverse over time, absent such interventions as CP. Facilities should increase efforts to prevent the transmission of pathogens that can cause serious infections, not decrease them.