June 13, 2008 - ASM Submits Comments to CMS on Proposed Rule on Hospital Acquired Infections

Centers for Medicare & Medicaid Services
Department of Health and Human Services
Attn: CMS-1390-P
7500 Security Boulevard
Baltimore, MD 21244-1850

To Whom It May Concern:

The American Society for Microbiology (ASM) appreciates the opportunity to review and comment on specific components of CMS-1390-P, Medicare Program; Proposed Changes to the Hospital Inpatient Prospective Payment Systems and Fiscal Year 2009 Rates, published in the Federal Register, Volume 73, Number 84 on April 30, 2008.

The ASM is the largest, single life sciences society dedicated to the advancement of the microbiological sciences and their application for the common good. The Society represents approximately 42,000 microbiologists, including scientists and science administrators working in a variety of areas, including biomedical, environmental, and clinical laboratory fields.

Many of our members have primary involvement in clinical laboratory medicine including individuals directing clinical microbiology or immunology laboratories, individuals licensed or accredited to perform such testing, industry representatives marketing products for use, and researchers involved in developing and evaluating the performance of new technologies. Clinical microbiologists, in particular, are involved on a day-to-day basis with the etiologic diagnosis of infectious diseases, including hospital acquired infections. Therefore, ASM members have a significant interest in ensuring that any revisions to the Inpatient Prospective Payment System acknowledge the importance and the complexities of establishing a microbiologic diagnosis, particularly if such data are to be used as a basis for payment adjustment.

The ASM will focus its comments on Section II.F. Preventable Hospital-Acquired Conditions (HACs), including infections:

General Comments

ASM acknowledges the statutory obligation under Public Law 109-171, the Deficit Reduction Act of 2005 (DRA) to select at least two conditions that are (1) high cost and/or high volume, (2) result in assignment to a Diagnosis-Related Group (DRG) that has a higher payment when present as a secondary diagnosis, and (3) could have been reasonably prevented through the application of evidence-based guidelines. Payment for such conditions will not be made unless documented as “Present on Admission” (POA). However, the ASM has significant concerns with the selection of hospital acquired infections (HAIs), including several identified in CMS-1533-P, Medicare Program; Proposed Changes to the Hospital Inpatient Prospective Payment Systems and Fiscal Year 2008 Rates, published in the Federal Register, Volume 72, Number 85 on May 3, 2007 as well as several proposed for consideration in CMS-1390-P.

First, it must be acknowledged that the documentation of a hospital acquired infection (HAI) is an imperfect process. Definitions for HAIs vary considerably based on guidelines issued from differing groups and include a combination of clinical and laboratory findings. In general, by definition an HAI will not have been present or incubating at the time of admission, but will manifest clinically at some variable point during treatment for another condition. Documenting the presence of an infection at admission or as a consequence of an intervention is inherently subject to variability based on clinical parameters used and laboratory procedural differences. Therefore, it is often difficult to reliably distinguish a healthcare associated infection from one that was community acquired but delayed in presentation or diagnosis.

There is also extreme variability in patient populations being served which influence rates of HAIs. Patients compromised by co-morbidities or underlying conditions that increase susceptibility to infection are necessarily at increased risk for HAIs. It is entirely possible that every generally accepted guideline for prevention of infection may be in place, yet the patient’s condition creates a situation whereby infection is inevitable.

It must also be acknowledged that during a hospital course, if the medical condition of the patient dictates that they require an intervention where benefits outweigh risks (e.g. use of a urinary catheter or a vascular access device, or surgical procedures and use of therapeutic agents), then it is a standard of care that the ordering physician has discussed the use of these interventions with the patient or authorized patient representative. Use of appropriate interventions would have to meet the requirements of being both reasonable and necessary, but they may also carry a risk of secondary infection. Likewise, use of antimicrobial agents may be indicated to treat an infection POA, but may predispose the patient to secondary Clostridium difficile associated-disease (CDAD). ASM strongly believes that measures to reduce these risks can and should be taken and are described in numerous infection control guidelines. In addition, guidelines should be in place to prevent transmission of infectious agents in healthcare settings, and are well described in Centers for Disease Control and Prevention (CDC) publications. However, we do not believe non-payment for often unavoidable infectious complications is the best option to insure compliance.

It is equally important to acknowledge that the laboratory diagnosis of an infection using a variety of microbiologic procedures is also subject to variability. In particular, microbiologic methods show a marked degree of heterogeneity with regard to both methods employed as well as determination of clinical significance of isolates, both factors which can affect the reporting of a definitive microbiologic diagnosis for a given infection. Further, recent trends towards decentralization of laboratories introduce additional preanalytical variables which may be a significant confounding factor in establishing an accurate etiologic diagnosis. With this is mind, we urge the Centers for Medicare & Medicaid Services (CMS) to consider the need to identify specific microbiologic parameters based on best practices prior to implementing any hospital acquired infection as a condition for DRG non-payment.

Therefore, it is the strong opinion of the ASM that while all reasonable precautions should be taken to mitigate risks of HAIs, it is virtually impossible to reduce the risk to zero or near-zero, methods for establishing a diagnosis are non-standard and highly variable, and any financial penalties for having used a necessary intervention leading to a secondary infection are unreasonable. Further, it is our opinion that a more reasonable approach is to develop Infection Control Quality Indicators as a component of the “Reporting Hospital Quality Data for Annual Payment Updates” (RHQDAPU). Ultimately, it is anticipated that a report-for-payment system would be replaced by a pay-for-performance system as data become available. However, it will be necessary to meet the conditions of GAO-08-283, Healthcare-Associated Infections in Hospitals (March 2008) to increase HHS leadership in prioritizing prevention practices and to improve data collection through processes such as the National Healthcare Safety Network (NHSN) prior to doing so. It is strongly recommended that an Advisory Group comprised of representatives of professional organization identified as stakeholders be established to assist in defining and implementing such Infection Control and Prevention Quality Indicators.

With regard to the specific infectious conditions proposed for implementation in 2008:

Catheter Associated Urinary Tract Infections

ASM continues to withhold support for the proposal that this be selected as an HAC. While we agree that the condition meets the established criteria, there are currently no widely accepted guidelines that define microbiologic diagnosis for this condition. The prevention guidelines cited from 1981 do not include laboratory parameters necessary for establishing diagnosis. However, at least two professional society documents are forthcoming that may help in establishing such criteria. The ASM Cumitech series has an updated edition of the document on Diagnosis of Urinary Tract Infections under review, and the Infectious Diseases Society of America has a planned document on Complicated Urinary Tract Infections in progress for Fall 2008 release. Until a standard microbiologic definition with a defined threshold for clinical significance and delineation of expected uropathogens is established, we believe there exists significant risk for variability in establishing and coding for a catheter associated urinary tract infection. Our recommendation is to withhold this from consideration until such criteria are developed. In addition, we have significant concerns that there is a risk of unanticipated consequences through unnecessary treatment of asymptomatic bacteriuria which may result in adverse antibiotic effects including emergence of resistance as well as CDAD.

Vascular catheter-associated infection

ASM also does not support this HAC for reasons similar to those stated for Catheter Associated Urinary Tract Infections. There is significant variability in risk for infection based on patient factors beyond the control of the healthcare entity. In addition, there is not a clearly established definition for a “catheter associated infection.” The CDC document “Guidelines for the Prevention of Intravascular Catheter-Related Infections” (MMWR, Vol. 51, RR-10, 2002) lacks definitional clarity, and a recent meta analysis of methods for diagnosing intravascular device related bloodstream infections demonstrates lack of a consensus procedure (Ann Int Med 2005; 142:451-466). Until there is clinical and microbiological definitional clarity, there exists significant risk in for variability in establishing and coding for such infections.

With regard to the specific infectious conditions proposed for implementation in 2009:

Surgical Site Infections Following Elective Surgeries

ASM does not support this as an HAC, again for reasons similar to those for the aforementioned HAIs. The 1999 HICPAC Guideline for Prevention of Surgical Site Infections fails to clearly define the clinical and microbiological basis for establishing an infection in the surgical situations delineated in this proposal. Further, there are significant differences in the infection risks associated with each of these surgery types, and even within each category there are distinctive differences based on the specific clinical situation. Moreover, post-surgical infections may occur sometime after the procedure and fail to be identified as an HAI, particularly if the patient presents to an alternate facility for follow up. Until additional definitional clarity is developed, there exists significant risk for variability in establishing and coding for these surgical infections. Finally, the number of cases reported does not seem compatible with the intent of the DRG adjustment to address high occurrence events.

Legionnaire’s Disease

ASM does not support the inclusion of Legionnaire’s Disease as an HAC. The diagnosis of Legionellosis is a difficult one to make both clinically and microbiologically. The numbers of reported cases and often clinical suspicion are low, and therefore the diagnosis is often delayed. Further, available methods are imperfect with generally low sensitivities when considering all serotypes and species, and diagnosis is frequently one of exclusion based on clinical grounds. Because of delays and imperfections in diagnosis, it is virtually impossible to identify a case as community acquired versus hospital acquired. Even clusters of cases may represent a common community outbreak as opposed to a healthcare facility associated outbreak. Finally, definitive association with a healthcare facility would require that isolates from patients be molecularly typed and matched to isolates obtained from a healthcare environmental source. Given that CDC does not recommend routine environmental sampling, and the reality that patient isolates are uncommonly available, making this association is both extremely difficult and highly impractical, and best left to public health officials well versed in best practices for this agent. The low numbers of cases and the extensive resources required to document Legionellosis render this an unsuitable HAC to focus on.

Ventilator-Associated Pneumonia

While Ventilator-Associated Pneumonia (VAP) is a common and highly significant clinical consequence of mechanical ventilation, ASM does not support its use as an HAC. Even with implementation of well defined best practices to reduce infection rates, there remain a substantial proportion of patients who will develop this serious condition. As for the other HAIs discussed, there is also significant variability in published clinical and microbiological criteria for the condition. Clinical features overlap significantly with other inflammatory lung conditions (e.g. ARDS), and microbiologic results are subject to many procedural variables. For these reasons, at least two articles have recently reached the conclusion that that VAP is NOT an appropriate quality measure for benchmarking hospital performance (Clin Infect Dis 2008. Vol 46: 557-563; Ann Int Med 2007: Vol 147: 803-805).

Staphylococcus aureus septicemia

ASM does not support the proposal that this be selected as an HAC.  Any attempt to single out a specific bloodstream infection with any particular pathogen as a preventable condition is fraught with confounding variables as bloodstream invasion can accompany virtually any serious infection, including those caused by S. aureus. The timing of the bloodstream invasion may suggest hospital acquisition, but in fact, may be simply a manifestation of the time course of the disease. With the recent emergence of serious skin and soft tissue infections as well as influenza-associated pneumonia due to methicillin resistant S. aureus, as well as the propensity of the organism to disseminate via the vasculature to establish secondary abscesses at other locations, we believe it is impossible to clearly define a S. aureus bloodstream infection as strictly hospital acquired. Further, as a subset of individuals are known to be stably colonized with S. aureus, absent any mandatory effort to perform preadmission testing and possibly decolonization for this organism, hospital acquired infection with this organism, including septicemia, may not be entirely preventable. In other words, colonization with S. aureus may actually represent a condition POA that is a risk factor for hospital acquired infection, and bloodstream infection may be due, in part, to patient related factors beyond the control of the health care worker. Consensus guidelines for managing MRSA or S. aureus colonization do not currently exist.

Clostridium difficile-associated disease (CDAD)

ASM does not support the inclusion of CDAD as an HAC. While the numbers of serious infections have reached an alarming rate (Emerg Infect Dis 2008, Vol 14: 929-931), it is virtually impossible to practically establish an infection as community acquired or hospital acquired. This is due in large part to the pathogenesis of the disease, whereby toxigenic organisms present in low numbers in many normal non-diseased individuals overgrow under the influence of antimicrobial or other chemotherapy for infectious and other diseases and result in significant pathology manifest as colitis. The inciting therapeutic agents may have been given prior to admission or as an appropriate intervention for another infectious process necessitating admission. As for Legionella, to establish an isolate of C. difficile as part of a hospital outbreak as opposed to an isolated virtually unavoidable case would require isolation of the organism pre-admission, and upon initiation of symptoms from all affected patients for molecular typing to establish an association. As diagnosis is primarily based on toxin detection and not culture, this is highly impractical and resource intensive for a condition that is not “reasonably preventable” in our opinion.

Concluding Statement

For all of the HAIs proposed for both 2008 and 2009 that we have specifically addressed, we strongly reiterate our concern that establishing these infections as HACs for which additional reimbursement will be denied is not a reasonable undertaking. We believe that HACs selected should have a high probability of prevention based primarily upon select interventions and not subject to extreme variability based on inadequate clinical and non-standardized microbiologic definitions. These criteria are not met for HAIs. Until consensus guidelines are developed and accepted for the accurate etiologic diagnosis of these conditions, we have significant concerns that accurate diagnosis of the conditions may be compromised and variability in obtaining appropriate reimbursement may be significant.

We also wish to emphasize that the approach of non-payment for specific HACs, particularly HAIs may have serious unintended consequences. There may be a disincentive to establish an etiologic diagnosis that could be harmful to public health efforts to control and prevent infectious diseases. Not only may significant outbreaks be missed, but data on emerging clinically significant antimicrobial resistance may be lost. Decreasing incentives to diagnose infectious diseases has dangerous public health implications and weakens our national monitoring capabilities for infectious diseases.

There also exists a significant risk for inappropriate and unnecessary utilization of already inadequate resources through extensive efforts to establish agents as POA. Surveillance for colonization in the absence of infection, with few exceptions, is not a strategy endorsed by the CDC in its guideline, “Management of Multidrug-Resistant Organisms in Healthcare Settings, 2006.” Attempts to establish surveillance programs in this manner will almost certainly lead not only to unnecessary testing, but may result in inappropriate antimicrobial therapy which may facilitate emergence of antibiotic resistant organisms.

CMS-1390-P is a highly complex document, and the time frame for developing comments was limited. Therefore, should you require any additional supporting information with regard to the ASM’s recommendations, please do not hesitate to contact the Society. In addition, should you require assistance or input into developing consensus microbiologic procedural guidelines, we stand ready to cooperate in any way possible. Similarly, should you decide to move forward with the development of more reasonable quality indicators for infection prevention rather than using HAIs as designated HACs, ASM would stand ready to cooperate in an advisory group designed to evaluate such indicators.

Thank you for the opportunity to comment.


Vickie S. Baselski, Ph.D., Chair, Committee on Professional Affairs, Public and Scientific Affairs Board
Susan E. Sharp, Ph.D., Chair, Committee on Laboratory Practices, Public and Scientific Affairs Board
Ruth L. Berkelman, M.D., Chair, Public and Scientific Affairs Board