A primary aim of public health surveillance is to direct prevention and control activities and monitor their effectiveness. Collection of surveillance data by itself does not control disease or constitute public health action. As public health agencies operate with finite resources, implementation of surveillance tools should occur in conjunction with a plan to interpret and act on the data collected. Resources should be devoted to MDRO surveillance activities only when resources are also available for specific MDRO infection- or transmission-prevention activities or to build capacity to respond with public health action to the MDRO surveillance data. Moreover, integrating input from local partners and key opinion leaders in infection control and prevention to ensure that any surveillance and response strategy is consistent with regional priorities or concerns is critical. A state is more likely to develop a surveillance system that meets its particular needs and functions well within the constraints of its available resources if the state health department takes an active role in deciding what MDRO surveillance activities are appropriate for its circumstances. Surveillance activities developed in response to mandates created without health department input could lack these characteristics.
Role of the health department
The details of case- or patient-level information necessary for MDRO reporting to public health agencies depend on several factors and vary by the specific MDRO. Very rare MDROs, such as vancomycin-resistant Staphylococcus aureus (VRSA), should include comprehensive reporting (i.e., all potential cases reported) coupled with an outbreak investigation to help contain spread and assist infection-control efforts. Such case-based reporting includes the collection of detailed demographic and clinical information about infected individuals to guide this response and evaluate potential clusters or transmission among health-care communities. This type of case-based reporting is commonly used for diseases included on the list of Nationally Notifiable Diseases. In the case of MDROs, the frequency of occurrence and the necessity of a public health response to the MDRO incident must be considered to avoid requiring ongoing reporting of frequent events unlinked to any coordinated public health or infection-control response. Case-based reporting can be a time-consuming and labor-intensive process because of the need to collect detailed information about each person with the disease or condition.
When an MDRO of interest has become established or endemic in a city, state, or region, obtaining detailed information about patients infected with the MDRO through public health investigation or case reporting may not be needed to implement effective prevention and control measures. The exact incidence or prevalence of an MDRO in a population above which mandatory reporting becomes too burdensome is not simple to determine and must take into account the pathogen's transmissibility, infectivity, and associated morbidity and mortality. Therefore, not all MDROs of epidemiologic importance or public health interest should be added to lists of notifiable diseases or require detailed case reporting to public health departments. The detection and control of MDROs in community and health-care settings can occur without mandatory reporting. For example, data on HAIs voluntarily reported to the NNIS System documented a 68% decrease in the incidence of central line-associated bloodstream infections among intensive care unit patients in Pennsylvania from 2001 to 2005.11
In the community setting, outbreaks of community-associated MRSA have been identified in the absence of mandatory reporting by various public health authorities who led successful infection-prevention interventions.12–15
Regarding MDROs that are somewhat common or are well established in a community, surveillance activities should focus on assessing infection rates and trends in these rates, identifying risk factors for infection, and monitoring the effectiveness of infection-prevention or transmission-interruption interventions. One area where this type of information is critically needed is in evaluating the role that health-care facilities other than short-stay acute care hospitals play in the transmission, propagation, and emergence of MDROs. Periodic evaluations in non-acute care settings may identify potential reservoirs of MDROs and could help guide the design of necessary studies. Such focused studies addressing these issues could be completed within the framework of existing programs (such as the activities of the Emerging Infections Program) in states that already have MDRO surveillance activities in place, and in conjunction with outbreak investigations. These types of focused studies will benefit stakeholders beyond their geographic areas and do not need to be replicated in all states or localities.
Relationship between the health department and health-care facilities
State and local health departments are the only entities with the authority and resources to respond to infections involving MDROs in the community setting, as well as those involving people residing in institutions outside health-care facilities. The role of the health department in responding to MDRO issues occurring within health-care facilities, however, has been less clearly defined because, historically, hospital epidemiologists and infection preventionists took responsibility for monitoring and controlling MDROs within individual acute care facilities. Additionally, the frequent movement of patients (and their colonizing or infecting MDROs) between acute care inpatient settings and non-acute care or outpatient facilities has made understanding the transmission of MDROs difficult. More information is needed about effective and appropriate prevention and infection-control measures that should be used at the interface between acute care hospitals and other health-care settings (such as long-term care facilities, ambulatory surgical centers, and outpatient dialysis centers) and the community; health departments can serve a role in assessing these measures in these settings. Second, health departments could have a role in the research, prevention, and control of community-onset MDRO infections in individuals whose infections may be related to previous exposure to the health-care environment. Working across acute care facilities to develop a communication system to provide accurate information about MDROs of concern among patients moving between facilities is one example.
Health departments often provide assistance for outbreak control within both acute and non-acute health-care facilities. However, not all health departments have infection-control expertise or the resources to provide such assistance. Significant investment in training staff or reliance on local experts in consulting roles would be required for this expertise to be commonplace among health departments. Variability in the level of this expertise and resource commitment also exists because states differ with regard to whether this function is housed under the regulatory or the epidemiology branch of their health departments. The amount of involvement that a state or local health department should take within acute care facilities is not static and varies between jurisdictions, depending on local resources and needs.
Regarding common infections, including MDROs but also for HAIs in general, the group agreed that it would be beneficial if state health departments were involved in the evaluation of data reported from health-care facilities within their jurisdiction. If the health department will have the responsibility for analyzing these data, then it must be able to devote adequate resources to this work. Ideally, the health department should also have a role in determining the surveillance system that collects the data. CDC's NHSN is an established system that provides a standardized method for HAI and MDRO reporting, which some health-care facilities are already using to share their data with state health departments. The specific reporting requirements and dissemination of data collected using NHSN and other mechanisms currently differ by state, and these issues were not addressed by the meeting participants. Some participants expressed reservations regarding public reporting of facility-specific rates, citing the need for appropriate risk adjustment of data, which is an area requiring further research. Other participants viewed the dissemination of these types of data by public health authorities as appropriate and one of several tools in facilitating improvements in patient safety and prevention of HAIs in all types of facilities. Preventing and controlling MDRO infections outside of acute care facilities requires collaboration among all types of health-care facilities and health departments. The health department's key functions should include facilitating the sharing of information, coordinating risk assessments, and working with regulatory agencies.
In summary, health departments and health-care facilities should partner on MDRO surveillance and prevention activities and, ideally, these activities should include information sharing across the continuum of care from both acute care and non-acute care facilities.
Current state- or city-based surveillance activities
Representatives of five state health departments and one city presented summaries of the current HAI- and MDRO-related surveillance activities occurring in their jurisdictions. Five of the six jurisdictions have a requirement mandating reporting of HAI cases to the state health department; four of these mandates included MRSA. The surveillance mechanisms discussed included manual laboratory-based case reporting, electronic laboratory reporting (ELR), utilization of hospital discharge claims data along with laboratory reports, aggregation of cumulative antimicrobial susceptibility data, and the use of CDC's NHSN (). Use of these data by the state health departments varied and included counting cases, tracking trends, identifying disparities in incidence among population groups, informing the public (consumers), educating stakeholders with local-level data, and publicly reporting hospital-specific rates ().
NHSN Patient Safety Component surveillance modules—focus, data elements, and metrics
Summary of key elements for HAI and MDRO surveillance from six health departments, as of September 22, 2008
An ideal surveillance system
Because many MDRO surveillance activities already exist in various states and localities, any new MDRO surveillance system that is developed should utilize and create linkages between existing data sources. The development of any new surveillance system for MDROs should not be limited to one pathogen alone, but should be an adaptable framework that can be used to receive data on various MDROs as circumstances and needs change over time. Due to the changing nature of MDROs and the change of pathogens of concern, the goal of surveillance may evolve over time. One goal is the early detection of emerging organisms in a region or emerging antimicrobial resistance in an established organism. Other goals are to detect outbreaks or guide control and eradication efforts. For endemic MDROs, the goal of surveillance might be incidence reporting and trend analysis. Surveillance systems can also be used to identify patients for inclusion in epidemiologic studies and to collect organism isolates for laboratory investigation. A surveillance system should be designed to achieve its intended purpose in an effective and efficient manner.16
There is not a single best method to address these broad goals for MDRO surveillance; however, there was agreement among participants that investment in the use of electronic reporting of data elements from hospital or laboratory information systems would provide a flexible framework to address most, if not all, of the goals of MDRO surveillance.
Identification of MDROs requires laboratory data—specifically, antimicrobial susceptibility test results. The implementation of automated ELR will potentially allow for an efficient process of population-based reporting. The benefits of ELR include the ability to address both community- and hospital-onset infections; the ability to limit the resources needed for data entry; a reduction in reporting delays, previously due to all manual extraction and entering of laboratory data; the availability of a wide range of data; the ability to detect more than just invasive infections; and the adaptability of an infrastructure that is not organism specific. ELR does, however, have limitations. Specifically, implementation and maintenance require sufficient funding, widespread use does not currently exist in most states, and all laboratories must format data in a standardized manner.
An ideal surveillance system for MDROs would incorporate automated ELR but require more data than ELR alone can provide. For example, ELR alone does not provide information about risk factors and cannot guide interventions. Further work is needed to link laboratory results with claims data or specialized case-based reporting to improve health department understanding of health-care exposures leading to infection. Also needed are evaluations of the utility of algorithms based on electronic data elements to inform surveillance data.
The development of ELR as part of a systematic surveillance effort by public health authorities will require substantial work by health departments, health-care facilities, and their reporting laboratories. Currently, standardized vocabulary for the critical test results is available but is typically not being used in laboratory information systems. Also, standards for the messaging of electronic data continue to evolve. As standards become established and incorporated into the various products offered by vendors supplying health-care information technology for the management of infection-control or patient-care activities, capturing and conveying standard data elements could become an integral part of state-based efforts. ELR for reporting HAI data should be integrated with ELR for reporting notifiable disease data so that common infrastructure and data standards serve both purposes.
Recommended surveillance activities using existing resources
Because a new, integrated MDRO surveillance system that includes automated ELR will take time to develop, state health departments should consider which MDRO surveillance activities should be pursued immediately. As stated previously, it is important to integrate input from local health-care partners and key opinion leaders in infection control to ensure any surveillance and response strategy is consistent with regional priorities or concerns. Additionally, the health department should provide feedback based on the surveillance activities to reporting facilities and other stakeholders. Dedicated support should be aimed at specific MDRO infection- or transmission-prevention activities. For common MDROs that are well established in a community, surveillance activities should focus on assessing infection rates and trends in those rates, identifying risk factors for infection, and monitoring the effectiveness of infection-prevention or transmission-interruption interventions.
For those jurisdictions that do not currently have ELR capability, the meeting participants recommended that health departments planning to implement surveillance for MDROs limit their surveillance to blood culture isolates if the MDRO is common (e.g., MRSA). Isolation of MRSA on blood culture has been shown to correlate with infection, whereas organisms cultured from non-sterile body sites, such as skin, could represent colonization.4
If a jurisdiction does have ELR capability, the state health department could expand the culture results to include both normally sterile and non-sterile body sites to utilize and report several types of metrics that are outlined in the Society for Healthcare Epidemiology of America/Healthcare Infection Control Practices Advisory Committee position paper on measuring MDROs in health-care facilities.17
Consideration should be given to which laboratories will be required to provide information. To achieve the most complete surveillance of both health care- and community-associated MDRO infections, it would be necessary to include all laboratories that serve a state's population, whether located inside or outside the state. Both hospital-based and non-hospital-based laboratories would be included. Depending on the goals of a state's MDRO surveillance system and the resources available, reporting by only a subset of all possible laboratories might be sufficient, such as in a sentinel surveillance system. The minimum data elements that should be collected for MDRO surveillance in any system are presented in .
Minimum data elements to consider in multidrug-resistant organism surveillance