Skip directly to search Skip directly to A to Z list Skip directly to navigation Skip directly to site content Skip directly to page options
CDC Home

Guidance on Public Reporting of Healthcare-Associated Infections: Recommendations of the Healthcare Infection Control Practices Advisory Committee

Guidance on Public Reporting of Healthcare-Associated Infections: Recommendations of the HICPAC [PDF 150 KB]

Linda McKibben, MD,a Teresa Horan, MPH,b Jerome I. Tokars, MD, MPH,b Gabrielle Fowler, MPH,b Denise M. Cardo, MD,a Michele L. Pearson, MD,c Patrick J. Brennan, MD,d and the Healthcare Infection Control Practices Advisory Committee*

Since 2002, 4 states have enacted legislation that requires health care organizations to publicly disclose health care–associated infection (HAI) rates. Similar legislative efforts are underway in several other states. Advocates of mandatory public reporting of HAIs believe that making such information publicly available will enable consumers to make more informed choices about their health care and improve overall health care quality by reducing HAIs. Further, they believe that patients have a right to know this information. However, others have expressed concern that the reliability of public reporting systems may be compromised by institutional variability in the definitions used for HAIs, or in the methods and resources used to identify HAIs. Presently, there is insufficient evidence on the merits and limitations of an HAI public reporting system. Therefore, the Healthcare Infection Control Practices Advisory Committee (HICPAC) has not recommended for or against mandatory public reporting of HAI rates. However, HICPAC has developed this guidance document based on established principles for public health and HAI reporting systems. This document is intended to assist policymakers, program planners, consumer advocacy organizations, and others tasked with designing and implementing public reporting systems for HAIs. The document provides a framework for legislators, but does not provide model legislation. HICPAC recommends that persons who design and implement such systems 1) use established public health surveillance methods when designing and implementing mandatory HAI reporting systems; 2) create multidisciplinary advisory panels, including persons with expertise in the prevention and control of HAIs, to monitor the planning and oversight of HAI public reporting systems; 3) choose appropriate process and outcome measures based on facility type and phase in measures to allow time for facilities to adapt and to permit ongoing evaluation of data validity; and 4) provide regular and confidential feedback of performance data to healthcare providers. Specifically, HICPAC recommends that states establishing public reporting systems for HAIs select one or more of the following process or outcome measures as appropriate for hospitals or long-term care facilities in their jurisdictions: 1) central-line insertion practices; 2) surgical antimicrobial prophylaxis; 3) influenza vaccination coverage among patients and healthcare personnel; 4) central line-associated bloodstream infections; and 5) surgical site infections following selected operations. HICPAC will update these recommendations as more research and experience become available. (Am J Infect Control 2005;33:217-26.)

Consumer demand for health care information, including data about the performance of health care providers, has increased steadily over the past decade. Many state and national initiatives are underway to mandate or induce health care organizations to publicly disclose information regarding institutional and physician performance. Mandatory public reporting of health care performance is intended to enable stakeholders, including consumers, to make more informed choices on health care issues.

Public reporting of health care performance information has taken several forms. Health care performance reports (report cards and honor rolls) typically describe the outcomes of medical care in terms of mortality, selected complications, or medical errors and, to a lesser extent, economic outcomes. Increasingly, process measures (ie, measurement of adherence to recommended health care practices, such as hand hygiene) are being used as an indicator of how well an organization adheres to established standards of practice with the implicit assumption that good processes lead to good health care outcomes. National health care quality improvement initiatives, notably those of the Joint Commission on the Accreditation of Healthcare Organizations (JCAHO), the Centers for Medicare & Medicaid Services (CMS), and the Hospital Quality Alliance, use process measures in their public reporting initiatives.

Health care–associated infections (HAIs) are infections that patients acquire during the course of receiving treatment for other conditions (see Appendix 1 for full definition of this and other terms used in this document). In hospitals alone, HAIs account for an estimated 2 million infections, 90,000 deaths, and $4.5 billion in excess health care costs annually1; however, few of the existing report cards on hospital performance use HAIs as a quality indicator. Since 2002, 4 states (Illinois, Pennsylvania, Missouri, and Florida) have enacted legislation mandating hospitals and health care organizations to publicly disclose HAI rates. Similar legislative efforts are underway in several other states.

Because of the increasing legislative and regulatory interest in this area, the Healthcare Infection Control Practices Advisory Committee (HICPAC) conducted a scientific literature review to evaluate the merits and limitations of HAI reporting systems.We found no published information on the effectiveness of public reporting systems in reducing HAIs. Therefore, HICPAC has concluded that there is insufficient evidence at this time to recommend for or against public reporting of HAIs.

However, to assist those who will be tasked with designing and implementing such reporting systems, HICPAC presents the following framework for an HAI reporting system and recommendations for process and outcome measures to be included in the system. The framework and recommendations are based on established principles for public health and HAI surveillance. This document is intended primarily for policymakers, program planners, consumer advocacy organizations, and others who will be developing and maintaining public reporting systems for HAI. The document does not provide model legislation.

This document represents the consensus opinion of HICPAC. HICPAC is a federal advisory committee that was established in 1991 to provide advice and guidance to the Department of Health and Human Services and CDC regarding surveillance, prevention, and control of HAIs and related events in healthcare settings (www.cdc.gov/HICPAC/).These recommendations also have been endorsed by the Association for Professionals in Infection Control and Epidemiology, the Council of State and Territorial Epidemiologists, and the Society for Healthcare Epidemiology of America. These recommendations also have been endorsed by the Association for Professionals in Infection Control and Epidemiology, the Council of State and Territorial Epidemiologists, and the Society for Healthcare Epidemiology of America. These recommendations will be updated as new information becomes available.

Top of Page

Essential Elements Of A Public Reporting System For HAIs

As a first step, the goals, objectives, and priorities of a public reporting system should be clearly specified and the information to be monitored should be measurable to ensure that the system can be held accountable by stakeholders. The reporting system should collect and report healthcare data that are useful not only to the public, but also to the facility for its quality improvement efforts. This can be achieved by selection of appropriate measures and patient populations to monitor; use of standardized case-finding methods and data validity checks; adequate support for infrastructure, resources, and infection control professionals; adjustment for underlying infection risk; and production of useful and accessible reports for stakeholders, with feedback to healthcare providers. The planning and oversight of the system should be monitored by a multidisciplinary group composed of public health officials, consumers, health care providers, and health care infection control professionals.

Top of Page

Identifying Appropriate Measures of Health Care Performance

Monitoring both process and outcome measures and assessing their correlation is a comprehensive approach to quality improvement. Standardized process and outcome measures for national health care performance for hospitals, nursing homes, and other settings have been endorsed through the National Quality Forum (NQF) voluntary consensus process.2-4 NQF also has developed a model policy on the endorsement of proprietary performance measures.5 Several other agencies and organizations, including CDC, CMS, the Agency for Healthcare Quality and Research, JCAHO, the Leapfrog organization, and the National Committee for Quality Assurance, also have developed health care quality measures. Health care performance reports should identify the sources and endorsers of the measures and the sources of the data used (eg, administrative or clinical).

Process measures. Process measures are desirable for inclusion in a public reporting system because the target adherence rate of 100% to these practices is unambiguous. Furthermore, process measures do not require adjustment for the patient's underlying risk of infection. Process measures that are selected for inclusion in a public reporting system should be those that measure common practices, are valid for a variety of health care settings (eg, small, rural versus large, urban hospitals); and can be clearly specified (eg, appropriate exclusion and inclusion criteria). Process measures meeting these criteria include adherence rates of central line insertion practices and surgical antimicrobial prophylaxis and coverage rates of influenza vaccination for health care personnel and patients/residents (Table 1). Collection of data on one or more of these process measures already is recommended by the NQF and required by CMS and JCAHO for their purposes. Outcome measures. Outcome measures should be chosen for reporting based on the frequency, severity, and preventability of the outcomes and the likelihood that they can be detected and reported accurately.14 Outcome measures meeting these criteria include central line–associated, laboratory-confirmed primary bloodstream infections (CLA-LCBI) in intensive care units (ICU) and surgical site infections (SSI) following selected operations (Table 2). Although CLA-LCBIs and SSIs occur at relatively low rates, they are associated with substantial morbidity and mortality and excess health care costs. Also, there are well-established prevention strategies for CLA-LCBIs and SSIs.6,10 Therefore, highest priority should be given to monitoring these two HAIs and providers' adherence to the related processes of care (ie, central-line insertion practices for CLA-LCBI and surgical antimicrobial prophylaxis for SSIs).

Use of other HAIs in public reporting systems may be more difficult. For example, catheter-associated urinary tract infections, though they may occur more frequently than CLA-LCBIs or SSIs, are associated with a lower morbidity and mortality; therefore, monitoring these infections likely has less prevention effectiveness relative to the burden of data collection and reporting. On the other hand, HAIs such as ventilator-associated pneumonia, which occur relatively infrequently but have substantial morbidity and mortality, are difficult to detect accurately. Including such HAIs in a reporting system may result in invalid comparisons of infection rates and be misleading to consumers.

Monitoring of process and outcome measures should be phased in gradually to allow time for facilities to adapt and to permit ongoing evaluation of data validity.

Top of Page

Identifying Patient Populations for Monitoring

CDC16 and other authorities17 no longer recommend collection or reporting of hospital-wide overall HAI rates because 1) HAI rates are low in many hospital locations (which makes routine inclusion of these units unhelpful), 2) collecting hospital-wide data is labor intensive and may divert resources from prevention activities, and 3) methods for hospital-wide risk adjustment have not been developed. Rather than hospitalwide rates, reporting rates of specific HAI for specific hospital units or operation-specific rates of SSIs is recommended.16 This practice can help ensure that data collection is concentrated in populations where HAIs are more frequent and that rates are calculated that are more useful for targeting prevention and making comparisons among facilities or within facilities over time.

Top of Page

Case-Finding

Once the population at risk for HAIs has been identified, standardized methods for case-finding should be adopted. Such methods help to reduce surveillance bias (ie, the finding of higher rates at institutions that do a more complete job of casefinding). Incentives to find cases of HAI may be helpful. Conversely, punitive measures for hospitals that report high rates may encourage underreporting.

Traditional case-finding methods for HAIs include review of medical records, laboratory reports, and antibiotic administration records. However, these standard case-finding methods can be enhanced. For example, substantially more SSIs are found when administrative data sources (eg, International Classification of Diseases, 9th Revision [ICD-9], discharge codes) are used in combination with antimicrobial receipt to flag charts for careful review.18,19 However, the accuracy of case-finding using ICD-9 codes alone likely varies by HAI type and by hospital. Therefore, ICD-9 discharge codes should not be relied upon as the sole source of case finding for HAI monitoring systems.

Traditional HAI case-finding methods were developed in an era when patients' lengths of hospitalization were much longer than they are today, allowing most HAIs to be detected during the hospital stay. However, for SSIs in particular, the current climate of short stays and rapid transfers to other facilities makes accurate detection difficult because as many as 50% of SSIs do not become evident until after hospital discharge or transfer.20 Since there is no consensus on which postdischarge surveillance methods are the most accurate and practical for detection of SSIs,10 the limitations of current case-finding methods should be recognized if SSIs are selected for inclusion in mandatory reporting systems.

Top of Page

Validation of Data

A method to validate data should be considered in any mandatory reporting system to ensure that HAIs are being accurately and completely reported and that rates are comparable from hospital to hospital or among all hospitals in the reporting system. The importance of validation was emphasized by a CDC study of the accuracy of reporting to the NNIS system, which found that although hospitals identified and reported most of the HAIs that occurred, the accuracy varied by infection site.15

Top of Page

Resources and Infrastructure Needed for a Reporting System

A reporting system can not produce quality data without adequate resources. At the institution level, trained personnel with dedicated time are required, eg, infection control professionals to conduct HAI surveillance. At the system level, key infrastructure includes instruction manuals, training materials, data collection forms, methods for data entry and submission, databases to receive and aggregate the data, appropriate quality checks, computer programs for data analysis, and standardized reports for dissemination of results. Computer resources within reporting systems must include both hardware and software and a standard user interface. In order to collect detailed data on factors such as use of invasive devises (eg, central lines), patient care location within the facility, type of operation, and extensive data dictionaries and coding schema must be developed and maintained.

Top of Page

HAI Rates and Risk Adjustment

For optimal comparison purposes, HAI rates should be adjusted for the potential differences in risk factors.

For example, in the NNIS system, device-associated infections are risk adjusted by calculating rates per 1,000 device-days (eg, CLA-LCBI per 1,000 central line–days) and stratifying by unit type.21-23 For that system, risk adjustment of SSIs is done by calculating of operation-specific rates stratified by a standardized risk index.23-25 Although these methods do not incorporate all potential confounding variables, they provide an acceptable level of risk adjustment that avoids the data collection burden thatwould be required to adjust for all variables.

Risk adjustment is labor intensive because data must be collected on the entire population at risk (the denominator) rather than only the fraction with HAIs (the numerator). Risk adjustment can not correct for variability among data collectors in the accuracy of finding and reporting events. Further, current riskadjustment methods improve but do not guarantee the validity of inter-hospital comparisons, especially comparisons involving facilities with diverse patient populations (eg, community versus tertiary-care hospitals).

Valid event rates are facilitated by selecting events that occur frequently enough and at-risk populations that are large enough to produce adequate sample sizes. Unfortunately, use of stratification (eg, calculation of rates separately in multiple categories) for risk adjustment may lead to small numbers of HAIs in any one category and thereby yield unstable rates, as is the case of a small hospital with low surgical volume.

Top of Page

Producing Useful Reports and Feedback

Publicly released reports must convey scientific meaning in a manner that is useful and interpretable to a diverse audience. Collaboration between subject matter experts, statisticians, and communicators is necessary in developing these reports. The reports should provide useful information to the various users and highlight potential limitations of both the data and the methods used for risk adjustment. In a new reporting system, data should be examined and validated before initial release; in addition, sufficient sample size should be accumulated so that rates are stable at the time of public release. Lastly, feedback of performance data should be given to health care providers regularly so that interventions to improve performance can be implemented as quickly as possible. For example, feedback of SSI rates to surgeons has been shown to be an important component of strategies to reduce SSI risk.26

Top of Page

Adapting Established Methods for Use In Mandatory Reporting Systems

Where appropriate, developers of reporting systems should avail themselves of established and proven methods of collecting and reporting surveillance data. For example, many of the methods, attributes, and protocols of CDC's NNIS system may be applicable for public reporting systems. A detailed description of the NNIS methodologies has been described elsewhere, 23 and additional information on NNIS is available at www.cdc.gov/ncidod/hip/surveill/nnis.htm.

Most reporting systems, such as NNIS, use manual data collection methods. In most instances, information in computer databases, when available, can be substituted for manually collected data.27,28 However, when manual data collection is necessary, alternate approaches include limiting reporting to well-defined and readily identifiable events, using simpler and more objective event definitions,29 and sampling to obtain denominators.30 These approaches could decrease the burden of data collection and improve the consistency of reporting among facilities. If data collection were simplified, expanding the number of infection types and locations in which they are monitored may become more feasible.

Top of Page

Potential Consequences of Mandatory Public Reporting Systems

Mandatory reporting of HAIs will provide consumers and stakeholders with additional information for making informed health care choices. Further, reports from private systems suggest that participation in an organized, ongoing system for monitoring and reporting of HAIs may reduce HAI rates.31,32 This same beneficial consequence may apply to mandatory public reporting systems. Conversely, as with voluntary private reporting, mandatory public reporting that doesn't incorporate sound surveillance principles and reasonable goals may divert resources to reporting infections and collecting data for risk adjustment and away from patient care and prevention; such reporting also could result in unintended disincentives to treat patients at higher risk for HAI. In addition, current standard methods for HAI surveillance were developed for voluntary use and may need to be modified for mandatory reporting. Lastly, publicly reported HAI rates can mislead stakeholders if inaccurate information is disseminated. Therefore, in a mandatory public report of HAI information, the limitations of current methods should be clearly communicated within the publicly released report.

Top of Page

Research and Evaluation Needs

Research and evaluation of existing and future HAI reporting systems will be needed to answer questions about 1) the comparative effectiveness and efficiency of public and private reporting systems and 2) the incidence and prevention of unintended consequences. Ongoing evaluation of each system will be needed to confirm the appropriateness of the methods used and the validity of the results.

Top of Page

Recommendations

The Healthcare Infection Control Practices Advisory Committee proposes four overarching recommendations regarding the mandatory public reporting of HAIs. These recommendations are intended to guide policymakers in the creation of statewide reporting systems for health care facilities in their jurisdictions.

  1. Use established public health surveillance methods when designing and implementing mandatory HAI reporting systems. This process involves:
    1. selection of appropriate process and outcome measures to monitor;
    2. selection of appropriate patient populations to monitor;
    3. use of standardized case-finding methods and data validity checks;
    4. provision of adequate support and resources;
    5. adjustment for underlying infection risk; and
    6. production of useful and accessible reports to stakeholders.

      Do not use hospital discharge diagnostic codes as the sole data source for HAI public reporting systems.
  2. Create a multidisciplinary advisory panel to monitor the planning and oversight of the operations and products of HAI public reporting systems. This team should include persons with expertise in the prevention and control of HAIs.
  3. Choose appropriate process and outcome measures based on facility type, and phase in measures gradually to allow time for facilities to adapt and to permit ongoing evaluation of data validity. States can select from the following measures as appropriate for hospitals or long term care facilities in their jurisdictions.
    1. Three process measures are appropriate for hospitals and one (iii below) is appropriate for long term care facilities participating in a mandatory HAI reporting system (Table 1).
      1. Central line insertion practices (with the goal of targeting ICU-specific CLA-LCBIs can be measured by all hospitals that have the type of ICUs selected formonitoring (eg, medical or surgical).
      2. Surgical antimicrobial prophylaxis (with the goal of targeting SSI rates) can be measured by all hospitals that conduct the operations selected for monitoring.
      3. Influenza vaccination coverage rates for health care personnel and patients can be measured by all hospitals and long term care facilities. For example:
        • Coverage rates for health care personnel can be measured in all hospitals and long term care facilities.
        • Coverage rates for high-risk patients can be measured in all hospitals.
        • Coverage rates for all residents can be measured in all long term care facilities.
    2. Two outcome measures are appropriate for some hospitals participating in a mandatory HAI reporting system (Table 2).
      1. CLA-LCBIs.
      2. SSIs following selected operations.

        Hospitals for which these measures are appropriate are those in which the frequency of the HAI is sufficient to achieve statistically stable rates. To foster performance improvement, the HAI rate to be reported should be coupled with a process measure of adherence to the prevention practice known to lower the rate (see 3ai and 3aii). For example, hospitals in states where reporting of SSIs is mandated should monitor and report adherence to recommended standards for surgical prophylaxis (see 3aii).
  4. Provide regular and confidential feedback of performance data to health care providers. This practice may encourage low performers to implement targeted prevention activities and increase the acceptability of the public reporting systems within the health care sector.

HICPAC thanks the following subject-matter experts for reviewing preliminary drafts of this guidance document: Victoria Fraser, MD, Washington University School of Medicine, St Louis, MO; Lisa McGiffert, Consumers Union; Richard Platt, MD, Harvard-Pilgrim Health, Boston, MA; Robert A Weinstein, MD, John J Stroger, Jr, Hospital of Cook County, Chicago, IL; and Richard P Wenzel, MD, Richmond, VA. HICPAC also thanks J Shaw and Patricia Simone, MD, for exceptional editorial guidance during the development of this document. The opinions of all the reviewers may not be reflected in all the recommendations contained in this document.

Top of Page

 
Contact Us:
  • Centers for Disease Control and Prevention
    1600 Clifton Rd
    Atlanta, GA 30333
  • 800-CDC-INFO
    (800-232-4636)
    TTY: (888) 232-6348
  • Contact CDC–INFO
USA.gov: The U.S. Government's Official Web PortalDepartment of Health and Human Services
Centers for Disease Control and Prevention   1600 Clifton Rd. Atlanta, GA 30333, USA
800-CDC-INFO (800-232-4636) TTY: (888) 232-6348 - Contact CDC–INFO
A-Z Index
  1. A
  2. B
  3. C
  4. D
  5. E
  6. F
  7. G
  8. H
  9. I
  10. J
  11. K
  12. L
  13. M
  14. N
  15. O
  16. P
  17. Q
  18. R
  19. S
  20. T
  21. U
  22. V
  23. W
  24. X
  25. Y
  26. Z
  27. #