AGENCY:

Centers for Medicare & Medicaid Services (CMS), HHS; Centers for Disease Control and Prevention (CDC), HHS.

ACTION:

Final rule.

SUMMARY:

This final rule updates proficiency testing (PT) regulations under the Clinical Laboratory Improvement Amendments of 1988 (CLIA) to address current analytes (that is, substances or constituents for which the laboratory conducts testing) and newer technologies. This final rule also makes technical changes to PT referral regulations to better align them with the CLIA statute.

DATES:

Effective August 10, 2022, except for the amendments to §§ 493.2 and 493.801 through 493.959 (amendatory instructions 2 and 5 through 21), which are effective July 11, 2024.

FOR FURTHER INFORMATION CONTACT:

Sarah Bennett, CMS, (410) 786-3531; or Nancy Anderson, CDC, (404) 498-2741.

SUPPLEMENTARY INFORMATION:

I. Background

On October 31, 1988, Congress enacted the Clinical Laboratory Improvement Amendments of 1988 (Pub. L. 100-578) (CLIA '88), codified at 42 U.S.C. 263a, to ensure the accuracy and reliability of testing in all laboratories, including, but not limited to, those that participate in Medicare and Medicaid, that test human specimens for the purpose of providing information for the diagnosis, prevention, or treatment of any disease or impairment, or the assessment of health, of human beings. The Secretary established the initial regulations implementing CLIA on February 28, 1992 at 42 CFR part 493 (57 FR 7002). Those regulations required laboratories conducting moderate or high-complexity testing to enroll in an approved proficiency testing (PT) program for each specialty, subspecialty, and analyte or test for which the laboratory is certified under CLIA. PT referral was further addressed by enactment of the Taking Essential Steps for Testing Act of 2012 (Pub. L. 112-202, December 4, 2012) (TEST Act) and our implementing regulations (79 FR 25435 and 79 FR 27105). As of January 2020, approximately 35,967 CLIA-certified laboratories were required to enroll in a U.S. Department of Health and Human Services (HHS)-approved PT program and comply with the PT regulations.

Participation in PT is required under the CLIA statute for laboratories that perform moderate or high complexity testing. PT evaluates a laboratory's performance by testing unknown samples just as it would test patient samples. An HHS-approved PT program sends unknown samples to a laboratory for analysis. After testing, the laboratory reports its results to the PT program. The program grades the results using the CLIA grading criteria and provides the laboratory with its scores. PT is crucial to maintaining the quality of laboratory testing because it independently verifies the accuracy and reliability of laboratory testing, including the competency of testing personnel.

Testing has evolved significantly since 1992, and today's technology is more accurate and precise than the methods used when the PT regulations became effective in 1994. In addition, many tests for analytes for which PT was not initially required are now in routine clinical use. For example, tests for troponins, which are used to diagnose myocardial infarction, and the hemoglobin A1c test commonly used to monitor glycemic control in persons with diabetes were not routinely performed prior to 1992. Recognizing these changes, we proposed revisions to update the existing PT regulations in a proposed rule entitled, “Clinical Laboratory Improvement Amendments of 1988 (CLIA) Proficiency Testing Regulations Related to Analytes and Acceptable Performance”, published in the February 4, 2019 Federal Register (84 FR 1536) (hereinafter the proposed rule).

Generally, a final rule must be issued within 3 years of publishing a proposed rule, except under exceptional circumstances. As discussed in a notice entitled, “Clinical Laboratory Improvement Amendments of 1988 (CLIA) Proficiency Testing Regulations Related to Analytes and Acceptable Performance; Extension of Timeline for Publication of Final Rule”, published in the January 19, 2022, Federal Register (87 FR 2736) (hereinafter the notice of extension), we could not meet the February 4, 2022 deadline due to the necessary reallocation of resources to respond to the COVID-19 public health emergency. Therefore, in the notice of extension, we announced an extension of the timeline to publish the final rule by 1 year until February 4, 2023.

As part of the process for developing the proposed rule, HHS solicited input from the Clinical Laboratory Improvement Advisory Committee (CLIAC), the official Federal advisory committee charged with advising HHS regarding appropriate regulatory standards for ensuring accuracy, reliability, and timeliness of laboratory testing. Taking CLIAC's recommendations into account, CMS and CDC collaborated to develop a process to revise the list of required PT analytes listed in subpart I to determine which analytes should be retained, which should be deleted, and which analytes not currently listed in subpart I should be added to the regulations. Following the data-driven process and step-wise criteria used to select the candidate analytes to be included in the proposed rule, CMS and CDC sought feedback from PT programs on the following topics: current PT program practices using “peer grouping” to determine target values; the potential to include new analytes as required PT; the mechanism for grading current analytes; possible changes to the criteria for acceptable performance; and potential changes to microbiology subspecialties, including the replacement of the types of service as outlined currently at §§ 493.911(a), 493.913(a), 493.915(a), 493.917(a) and 493.919(a), with the proposed categories of required PT for each microbiology subspecialty at the above citations and the replacement of the list of specific organisms for each subspecialty with a proposed list of types of microorganisms.

Based on empirical data and clinical relevance, CMS and CDC next worked to determine or revise the acceptance limits (ALs) (as defined in § 493.2) for new and existing required analytes, respectively. Whenever possible, we proposed ALs as percentages. For each analyte, PT programs voluntarily provided data simulations using real PT data as a means of pilot testing our potential ALs. As stated in the proposed rule, ALs are intended to be used for scoring PT performance by PT programs and are not intended to be used by individual laboratories to satisfy the requirement at § 493.1253(b) to establish performance specifications.

II. Provisions of the Proposed Regulations

The proposed rule, if finalized, would amend the definitions and PT Start Printed Page 41195 requirements in subpart A—General Provisions, § 493.2 Definitions; subpart H—Participation in Proficiency Testing for Laboratories Performing Nonwaived Testing; and subpart I—Proficiency Testing Programs for Nonwaived Testing in the CLIA regulations.

A. Proposed Changes to Microbiology PT

1. Categories of Testing

Subpart I of the CLIA regulations includes PT requirements for each subspecialty of microbiology, §§ 493.911 through 493.919, which describe “Types of services offered by laboratories” for each subspecialty. In addition, since the regulations do not specify required analytes for microbiology as they do for other specialties, they include descriptions of levels or extents (for example, identification to the genus level only, identification to the genus and species level) used to determine the type of laboratory for PT purposes. CLIAC discussed the usefulness and limitations of the types of services listed in subpart I in helping laboratories enroll properly or in helping surveyors conduct laboratory inspections. It was noted that the types of services listed in subpart I do not allow for reporting growth or no growth, presence or absence, or presumptive identification of microorganisms on PT samples, which are common ways that physician office laboratories report patient results. CLIAC suggested revision of the regulations to include broad categories for the types of PT required for each microbiology subspecialty to allow flexibility for the inclusion of new technologies.

After deliberation, CLIAC made the following recommendations:

• A system for categorizing types of service should be maintained in the regulations to help laboratories determine what PT they need to perform and assist surveyors in monitoring PT performance and patient testing.
• The regulations should include four categories of testing for each microbiology subspecialty, as applicable: stain(s), susceptibility and resistance testing, antigen and/or toxin detection, and microbial identification or detection.

Based on these recommendations, we conducted a review of the PT modules offered by HHS-approved PT programs and consulted with CDC microbiology subject matter experts, who concurred that not all four recommended categories above are applicable to each microbiology subspecialty nor do PT programs have PT available for each category. If at some point in the future PT becomes available, we may propose to include additional categories of testing for microbiology subspecialties in future rulemaking. Based on these recommendations and our review, we proposed to modify §§ 493.911 through 493.919 to remove the types of services listed for each microbiology subspecialty and to add the recommended categories of testing (that is, replace the list with broader categories of organisms) for each microbiology subspecialty as described in the bullets below. We believe that the revised microbiology PT regulations would better reflect current practices in microbiology.

++ Section 493.911(a): For bacteriology, we proposed that the categories required include, as applicable: Gram stain including bacterial morphology; direct bacterial antigen detection; bacterial toxin detection; detection and identification of bacteria which includes either: detection of growth or no growth in culture media or identification of bacteria to the highest level that the laboratory reports results on patient specimens; and antimicrobial susceptibility or resistance testing on select bacteria.

++ Section 493.911(a)(3): We proposed that the bacteriology annual PT program content described must include representatives of the following major groups of medically important aerobic and anaerobic bacteria if appropriate for the sample sources: Gram-negative bacilli; Gram-positive bacilli; Gram-negative cocci; and Gram-positive cocci.

++ Section 493.913(a): For mycobacteriology, we proposed that the categories for which PT is required include, as applicable: acid-fast stain; detection and identification of mycobacteria which includes one of the following: detection of growth or no growth in culture media or identification of mycobacteria; and antimycobacterial susceptibility or resistance testing.

++ Section 493.913(a)(3): For mycobacteriology, we proposed that the annual program content must include Mycobacterium tuberculosis complex and Mycobacterium other than tuberculosis (MOTT), if appropriate for the sample sources.

++ Section 493.915(a): For mycology, we proposed the categories for which PT is required include, as applicable: direct fungal antigen detection; detection and identification of fungi and aerobic actinomycetes which included one of the following: detection of growth or no growth in culture media or identification of fungi and aerobic actinomycetes; and antifungal susceptibility or resistance testing.

++ Section 493.915(a)(3): We proposed that annual program content must include the following major groups of medically important fungi and aerobic actinomycetes if appropriate for the sample sources: yeast or yeast-like organisms; molds that include dematiaceous fungi, dermatophytes, dimorphic fungi, hyaline hyphomycetes, and mucormycetes; and aerobic actinomycetes.

++ Section 493.917(a): For parasitology, we proposed requiring PT for direct parasite antigen detection and detection and identification of parasites.

++ Section 493.917(a)(3): We proposed that the annual program content must include intestinal parasites and blood and tissue parasites, if appropriate for the sample source.

++ Section 493.919(a): For virology, we proposed requiring PT, as applicable, for viral antigen detection; detection and identification of viruses; and antiviral susceptibility or resistance testing.

++ Section 493.919(a)(3): We proposed that the annual program content must include respiratory viruses, herpes viruses, enterovirus, and intestinal viruses, if appropriate for the sample source.

We proposed revising the requirements for evaluating a laboratory's performance at §§ 493.911(b) through 493.919(b) to be consistent with these categories. We did not propose to include antigen and toxin detection in the mycobacteriology subspecialty because no PT program currently offers applicable PT modules. We did not propose to include stains and antiparasitic susceptibility or resistance testing in the subspecialty of parasitology because no PT program offers applicable PT modules. We invited the public to comment on these proposals and specifically on the proposed categories of testing for the subspecialties listed above. We stated that if public comments indicate that applicable PT modules are available for antigen and toxin detection or stains and antiparasitic susceptibility or resistance testing, we may finalize their inclusion in the final rule, as applicable. If PT becomes available at some point in the future for mycobacteriology antigen and toxin detection testing, and stains and antiparasitic susceptibility or resistance testing, we may propose to include this category of testing for PT in future rulemaking. We summarize and respond to the public comments on these proposals and summarize our final policies in section III.E. of this final rule. Start Printed Page 41196

++ Sections 493.911(b)(1), 493.913(b)(1), 493.915(b)(1), 493.917(b)(1), and 493.919(b)(1): We proposed amending these provisions to clarify that to achieve consensus, PT programs must attempt to grade using both participant and referee laboratories ^[1] before determining that the sample is ungradable. We believe that this change will enhance consistency among the PT programs when grading samples. The current regulations noted above allow for scoring either with participants or with referees before calling a sample ungradable. We summarize and respond to the public comments we received on these proposals and summarize our final policies in section III.D. of this final rule.

2. Major Groups of Microorganisms

In the proposed rule (84 FR 1536, 1538), we proposed to remove the lists of specific example organisms from each microbiology subspecialty and add a more general list of organisms. This change clarifies that PT programs are able to be flexible in selecting which samples to provide to laboratories for PT, especially as new organisms are identified as being clinically important.

Each subspecialty of microbiology, §§ 493.911 through 493.919, currently includes a list of the types of microorganisms that might be included in an HHS-approved PT program over time. Several PT programs have suggested to HHS that the regulations should include a more general list of types of organisms that must be included in required PT instead of a specific list. CLIAC considered whether there needs to be a more general list of organisms in the regulations to ensure a variety of challenges are offered over the course of the year. Following their deliberation, CLIAC made the following recommendation:

• Require PT for a general list of types of organisms in each subspecialty. For example, in bacteriology, the groups listed should include Gram-negative bacilli, Gram-positive bacilli, Gram-negative cocci, and Gram-positive cocci.

Generally, we have found that PT programs include only those organisms listed in the current regulations, and do not include additional organisms outside the current regulatory list. By restructuring to a more general list of organisms, it will be more apparent that PT programs are able to be flexible in selecting which samples to provide to laboratories for PT, especially as new organisms are identified as being clinically important. Therefore, we proposed to remove the lists of specific example organisms from each microbiology subspecialty, §§ 493.911 through 493.919, and to add the following list of types of organisms to each.

++ Section 493.911(a)(3): For bacteriology, we proposed that the annual program content must include representatives of the following major groups of medically important aerobic and anaerobic bacteria if appropriate for the sample sources: Gram-negative bacilli; Gram-positive bacilli; Gram-negative cocci; and Gram-positive cocci. The more general list of types of organisms will continue to cover the six major groups of bacteria currently listed in the regulations.

++ Section 493.913(a)(3): For mycobacteriology, we proposed that the annual program content must include Mycobacterium tuberculosis complex and Mycobacterium other than tuberculosis (MOTT), if appropriate for the sample sources.

++ Section 493.915(a)(3): For mycology, we proposed that the annual program content must include the following major groups of medically important fungi and aerobic actinomycetes if appropriate for the sample sources: yeast or yeast-like organisms; molds that include dematiaceous fungi, dermatophytes, dimorphic fungi, hyaline hyphomycetes, and mucormycetes; and aerobic actinomycetes.

++ Section 493.917(a)(3): For parasitology, we proposed that the annual program content must include intestinal parasites and blood and tissue parasites, if appropriate for the sample sources.

++ Section 493.919(a)(3): For virology, we proposed that the annual program content must include respiratory viruses, herpes viruses, enterovirus, and intestinal viruses, if appropriate for the sample sources.

We summarize and respond to the public comments we received on these proposals and summarize our final policies in section III.E. of this final rule.

3. Declaration of Patient Reporting Practices

The PT requirements at § 493.801(b) specify that laboratories must examine or test, as applicable, the proficiency testing samples it receives from the proficiency testing program in the same manner as it tests patient specimens. CLIAC considered this requirement as applied to microbiology and agreed that PT programs should instruct laboratories to perform all testing as they normally would on patient specimens, including reporting PT results for microorganism identification to the same level reported on patient specimens. CLIAC deliberated on this issue and made the following recommendation:

• Laboratories should declare their patient reporting practices for organisms included in each PT challenge. However, PT programs should only gather this information as the inspecting agency is responsible for reviewing and taking action if necessary.

We believe that laboratories should be instructed to report PT results for microbiology organism identification to the “highest” level that they report results on patient specimens to ensure that they do so to the “same” level that they report results on patient specimens. As a result, we proposed to amend §§ 493.801(b), 493.911(b), 493.913(b), 493.915(b), 493.917(b), and 493.919(b), to state that laboratories must report PT results for microbiology organism identification to the highest level that they report results on patient specimens. If finalized, this proposal should address an issue we identified during the PT program reapproval process in which we found laboratories inappropriately deciding whether to participate in a PT event based on the reporting criteria required by the PT program. We believe that this change will enhance consistency among the PT programs when grading samples.

We summarize and respond to the public comments we received on these proposals and summarize our final policies in sections III.C. and III.E. of this final rule.

4. Gram Stain PT

CLIAC considered whether the required PT for Gram stains should include both stain reaction and morphology. CLIAC concluded it should and recommended:

• PT results for Gram stains should include both stain reaction and morphology.

We agree with this recommendation because knowing the bacterial morphology is essential for accurate identification of specific groups of bacteria. Therefore, we proposed the following in § 493.911:

++ Section 493.911(a): The addition of required morphology for Gram stains.

++ Section 493.911(b): The evaluation of a laboratory's performance would be modified to include bacterial morphology as one part of the performance criterion for scoring the Gram stain.

We summarize and respond to the public comments on these proposals Start Printed Page 41197 and summarize our final policies in section III.E. of this final rule.

5. Mixed Culture Requirement

The current CLIA requirements for bacteriology §§ 493.911(b)(1), mycobacteriology 493.913(b)(1), and mycology 493.915(b)(1) specify that at least 50 percent of the PT samples in an annual program must be mixtures of the principal organism and appropriate normal flora. This requirement aims to simulate the findings that would occur with actual patient specimens. In bacteriology, this 50 percent mixed culture requirement must be met for two required sample types, those that require laboratories to report only organisms that the testing laboratory considers to be a principal pathogen that is clearly responsible for a described illness (excluding immunocompromised patients) and those that require laboratories to report all organisms present. The CLIA requirements for mycobacteriology and mycology PT do not specify two sample types. Still, they include the 50 percent requirement for cultures containing a mixture of the principal organism and appropriate normal flora. None of the 50 percent mixed culture requirements in these subspecialties applies to samples that would only contain normal flora and no reportable organisms.

CLIAC considered whether PT should include mixed cultures and discussed the difficulties of having mixed cultures in challenges for antimicrobial susceptibility testing. CLIAC considered lowering the mixed culture requirement to 25 percent for all subspecialties in microbiology. Upon deliberation, CLIAC made the following recommendation:

• Lower the mixed culture requirement from 50 percent to 25 percent for PT challenges of both sample types (those that require laboratories to report only the principal pathogen and those that require laboratories to report all organisms present).

We agree it is appropriate to lower the mixed culture requirement from 50 percent to 25 percent for bacteriology, mycobacteriology, and mycology to better reflect actual patient samples. As a result, we proposed the following changes:

++ Section 493.911(a)(2): In bacteriology, we proposed to decrease the required mixed cultures from 50 percent to 25 percent for culture challenges that require laboratories to report only the principal pathogen and those that require laboratories to report all organisms present.

++ Sections 493.913(a)(2) and 493.915(a)(2): In mycobacteriology and mycology, respectively, we proposed to decrease the mixed culture requirement from 50 percent to 25 percent.

Since the requirements for parasitology and virology do not currently include requirements for mixed cultures (or mixed PT challenges), we did not propose to make any changes to these subspecialties. We summarize and respond to the public comments we received on these proposals and summarize our final policies in section III.E. of this final rule.

6. Antimicrobial Susceptibility Testing

PT for antimicrobial susceptibility testing is currently required for bacteriology at § 493.911(b)(1) and mycobacteriology at § 493.913(b)(1), but it is not required for mycology, parasitology, or virology. For antimicrobial susceptibility testing in bacteriology at § 493.911(b)(3), at least one sample per testing event must include one Gram-positive or Gram-negative sample, and for mycobacteriology at § 493.913(b)(3), at least one sample per testing event must include a strain of Mycobacterium tuberculosis with a predetermined pattern of susceptibility or resistance to the common antimycobacterial agents. In some instances, laboratories appreciate the opportunity to participate in additional susceptibility testing challenges as educational tools. Under the current regulations, some laboratories may perform the minimum required susceptibility testing on some organisms, such as Gram-positive cocci. When CLIAC discussed this issue, the point was made that by increasing the frequency and number of required susceptibility testing PT challenges for different groups of organisms, potential issues with patient testing in a laboratory may be detected sooner. CLIAC considered recommending increasing the susceptibility testing challenges to two per event and requiring one Gram-positive and one Gram-negative organism in each bacteriology testing event. CLIAC also considered whether PT should be required for resistance as well as susceptibility testing and whether these requirements should be extended to other microbiology subspecialties. Following this deliberation, CLIAC made the following recommendations:

• Required PT for antimicrobial susceptibility and/or resistance testing should be increased to two challenges per event for a total of six challenges per year in bacteriology and should include one Gram-positive and one Gram-negative organism in each event.
• PT should be required for laboratories that perform susceptibility and/or resistance testing in all microbiology subspecialties. It should include two challenges per event and should include resistant organisms.

In considering these recommendations, we reviewed the modules currently offered by PT programs that include susceptibility testing and noted that there is a limited number of applicable PT modules currently available for resistance testing. Also, no PT program currently offers applicable PT modules for antiparasitic susceptibility or resistance testing in the subspecialty of parasitology. We believe it could be beneficial to increase the number of challenges per event from one to two for each microbiology subspecialty to increase the likelihood of detecting a problem in a laboratory. Antiparasitic susceptibility or resistance testing is not included in the subspecialty of parasitology because no PT program currently offers applicable PT modules. Therefore, we proposed the following:

++ Section 493.911(a)(4): For bacteriology, we proposed requiring at least two PT samples per event for susceptibility or resistance testing, including one Gram-positive and one Gram-negative organism with a predetermined pattern of susceptibility or resistance to common antimicrobial agents.

++ Section 493.913(a)(5): For mycobacteriology, we proposed requiring at least two PT samples per event for susceptibility or resistance testing, including mycobacteria that have a predetermined pattern of susceptibility or resistance to common antimycobacterial agents.

++ Section 493.915(a)(4): For mycology, we proposed requiring at least two PT samples per event for susceptibility or resistance testing, including fungi that have a predetermined pattern of susceptibility or resistance to common antifungal agents.

++ Section 493.919(a)(4): For virology, we proposed requiring at least two PT samples per event for susceptibility or resistance testing, including viruses that have a predetermined pattern of susceptibility or resistance to common antiviral agents.

In each of these subspecialties, we also proposed to revise the requirements for the evaluation of a laboratory's performance at §§ 493.911(b), 493.913(b), 493.915(b), and 493.919(b) to account for the fact that PT would be required for susceptibility or resistance Start Printed Page 41198 testing and that the scoring should be consistent with the testing performed.

We summarize and respond to the public comments we received on these proposals and summarize our final policies in section III.E. of this final rule.

7. Direct Antigen Testing

PT for direct antigen testing is only required for bacteriology and virology under §§ 493.911(a) and 493.919(a), respectively, not for the other microbiology subspecialties of mycobacteriology, mycology, and parasitology. Since this type of testing is commonly used for testing patient specimens, especially in mycology and parasitology, CLIAC considered whether PT for direct antigen testing should be part of all of the microbiology subspecialty requirements. CLIAC indicated that direct antigen PT should be required in subspecialties where these methods are used, and PT is available and made the following recommendation:

• PT for direct antigen testing should be required for all microbiology subspecialties.

We reviewed the modules currently offered by PT programs and determined that several modules include direct antigen testing for all microbiology subspecialties except mycobacteriology, for which this technology is not commonly used for testing patient specimens. In addition, we recognized that in bacteriology, PT for direct antigen testing to detect toxins produced by organisms such as Clostridioides (formerly Clostridium ) difficile is also commonly available. Based on the information collected from the PT programs, availability of the modules, and importance to the health and safety of the public, we proposed to:

++ Retain the requirement for direct antigen detection for:

—Section 493.911(a)(1)(ii): Bacteriology.

—Section 493.919(a)(1)(i): Virology.

++ Add the requirement for direct antigen testing detection for:

—Section 493.915(a)(1)(i): Mycology.

—Section 493.917(a)(1)(i): Parasitology.

++ Require PT for bacterial toxin detection under § 493.911(a)(1)(iii). No changes were proposed for mycobacteriology.

++ Add the evaluation criteria of a laboratory's performance for two of the affected subspecialties under §§ 493.911(b) and 493.917(b) to include performance and scoring criteria that address direct antigen and toxin detection. Evaluation of a laboratory's performance for direct antigen testing at § 493.917(b) would align with the other microbiology subspecialties and reflect current microbiology practices in reporting patient results. Evaluation of a laboratory's performance for bacterial toxin detection at § 493.911(b) would reflect the current practice of reporting patient test results (that is, absence or presence of bacterial toxin).

We summarize and respond to the public comments we received on these proposals and summarize our final policies in section III.E. of this final rule.

B. Proposed Changes to PT for Non-Microbiology Specialties and Subspecialties

In addition to determining which analytes should be added or deleted, CMS and CDC proposed to establish or change, if necessary, the criteria for acceptable performance, which include the target value and ALs, for the analytes. Currently, the CLIA regulations at §§ 493.927(c)(2), 493.931(c)(2), 493.933(c)(2), 493.937(c)(2), and 493.941(c)(2) prescribe a variety of ALs, including: a multiple of the standard deviation (SD) of results from the mean of all laboratories in the peer group; fixed limit as a percentage of the assigned value; fixed limit in concentration units; and a mixture of percentage and concentration units, depending on the concentration of the analyte. As discussed in section II.B. of the proposed rule, for all new and currently required non-microbiology analytes, we proposed to amend certain analytes in §§ 493.927, 493.931, 493.933, 493.937, and 493.941 to include percentages with or without fixed ALs. Additionally, we proposed to tighten ALs for certain current analytes in §§ 493.927, 493.931, 493.933, 493.937, 493.941, and 493.959.

We summarize and respond to the public comments we received on these proposals and summarize our final policies in section III.F. of this final rule.

1. Analytes Proposed for Addition to Subpart I

The CLIA statute requires the PT standards established by the Secretary to require PT for each examination and procedure for which the laboratory is certified “except for examinations and procedures for which the Secretary has determined that a proficiency test cannot reasonably be developed” (42 U.S.C. 263a(f)(3)(A)). In determining whether PT can reasonably be developed for a given analyte, we considered whether the estimated cost of PT is reasonable in comparison to the expected benefit. We attempted to maximize improvements to the effectiveness of PT to improve accuracy, reliability and timeliness of testing while minimizing costs to the laboratories. In addition, we recognize that requiring PT for every analyte to derive benefits generalizable to all test methods is unnecessary. For example, systematic analytical problems on a multichannel analyzer might be detected by participation in PT for any of the analytes tested. Further, laboratories are already required under § 493.1236(c)(1) to verify the accuracy of any test or procedure they perform that is not included in subpart I at least twice annually. Also, based on the results of the national PT survey conducted by CDC and the Association of Public Health Laboratories (APHL) in 2013, many laboratories voluntarily purchased PT materials for many nonrequired analytes. Keeping this in mind, as discussed in section II.B.2. of the proposed rule, we proposed adding the most crucial analytes based upon the following criteria:

(1) Current availability of PT materials and the number of PT programs offering PT.

(2) Volume of patient testing performed nationwide.

(3) Impact on patient health and/or public health.

(4) Cost and feasibility of implementation.

2. Process for Ranking Analytes Proposed for Addition to Subpart I

We used a sequential process to narrow the list of eligible analytes for addition based on each of the four criteria listed above.

a. Current Availability of PT Materials and the Number of PT Programs Already Offering PT

We believe that the availability of these PT samples for a particular analyte is an appropriate criterion for narrowing the list of eligible analytes and that scaling up a program would be relatively less difficult than creating a PT sample for a particular analyte that had not previously been offered. For the reasons noted below, we believe that at least three PT programs offering PT samples for a particular analyte under consideration would provide a sufficient number of programs to offer immediate access to PT by laboratories and a reasonable starting point for the analytes under consideration. CMS and CDC want to ensure that the laboratories could choose the best PT program for the services that their laboratories offered as well as not create a market advantage for a small number of PT programs. To evaluate the current availability of PT materials and PT Start Printed Page 41199 programs offering PT samples for a particular analyte, we analyzed the distribution of available PT programs for analytes for which PT is currently not required by subpart I of the CLIA regulations. The supporting data were collected from available sources, including data from PT program catalogs and data routinely reported by PT programs, including enrollment data. We examined the number of PT programs offering these analytes at any number of events per year and any number of challenges per event. We initially determined the number of analytes under consideration for which PT was offered by at least two, three, or four of the 11 existing PT programs. We determined that limiting the analytes under consideration to those for which PT was offered by at least three PT programs allowed a sufficient number of programs to offer immediate access to PT by laboratories and provided a reasonable starting point of 199 for the number of analytes under consideration (96 in routine chemistry, 27 in endocrinology, 28 in toxicology, 25 in general immunology, 21 in hematology, two for antibody identification). The expected impact on laboratories and PT programs was also considered (for example, minimizing the cost of purchasing and providing samples) when determining the minimum number of PT programs. Decreasing the minimum PT programs to two rather than three would increase the number of analytes under consideration to 303 but presumably decrease PT program availability and access for a given analyte. Conversely, increasing the minimum number of PT programs to four while presumably increasing PT program availability and access for a given analyte decreased the number of analytes under consideration to 164. This was the first cut based upon available PT modules.

b. Volume of Patient Testing Being Performed Nationwide

For the second cut, we prioritized the remaining 199 analytes under consideration based upon estimated national testing volumes. We decided that an estimated national test volume of 500,000 per analyte annually was an appropriate threshold as it was based upon testing volumes of the majority (68 out of 81) of analytes currently listed in subpart I. For comparison, of the analytes currently required under subpart I, 63 had a total national test volume above 1,000,000; five had national test volumes between 500,000 and 1,000,000, and 13 had national test volumes below 500,000. We used 500,000 annual tests as a preliminary cut-off for retention on the list of analytes under consideration. We also retained analytes below the 500,000 threshold that we determined to be clinically important based on literature already footnoted in section II.B.2.b. of the proposed rule and consultation with CDC health experts. The following analytes with test volumes less than 500,000 that were retained are: carbamazepine, alpha-1-antitrypsin, phenobarbital, hepatitis Be antigen, antibody identification, theophylline, gentamicin, and tobramycin.

In estimating national testing volumes to rank the remaining 199 analytes under consideration in the proposed rule, we were unable to identify a single source of available data for all patient testing being performed nationwide. We had complete data for Medicare payment, as well as the most current MarketScan Commercial Claims and Encounters (CCAE) and MarketScan Medicaid Multi-state data sets ^[2] and extrapolated accordingly. We used data provided by an HHS-approved accreditation organization, specifically a list of the number of their accredited laboratories offering each test we considered for addition to, or deletion from, subpart I to determine how many laboratories were performing testing for the proposed analytes. We also considered smaller representative data sets, including data sets obtained from a large healthcare network, a large reference laboratory, and a university hospital network, to evaluate the testing trends for the proposed analytes. We analyzed national testing trends based upon Medicare Part B payment data ^[3] to determine the analytes in each specialty that are increasingly used for patient diagnosis and/or management. We concluded that the trends revealed in the data could continue to show increases in payment for the proposed analytes.

We estimated the 2009 national test volumes based upon two data sets: (1) Medicare Part B payment statistics (excluding waived testing); and (2) CCAE. For all analytes under consideration for the addition to subpart I, we used Current Procedural Terminology (CPT) codes from claims data. We identified all possible occurrences of a particular analyte and combined them into one count. For example, if bicarbonate could be performed in a panel and by itself, we included all possible occurrences.

A complete count was available for the Medicare Part B data, and no estimation of total counts was necessary for this sector. MarketScan data, a sample of approximately 40 million covered individuals, was necessary to estimate CCAE data and approximately 6.5 million covered individuals for Medicaid data. Therefore, we estimated the total number of tests in both categories for the entire United States. The Agency for Healthcare Research and Quality (AHRQ) data showed that an estimated total of 181.5 million covered individuals enrolled in CCAE healthcare insurance; from this we derived a factor of 4.5 (181.5 million individuals/40 million individuals) by which to multiply the MarketScan CCAE estimates to extrapolate estimates for the entire United States. Similarly, for the Medicaid estimates, we knew from CMS data that there were approximately 52.5 million individuals covered by Medicaid, so we derived a factor of 8.0 (52.5 million individuals/6.5 million individuals) by which to multiply the MarketScan Medicaid estimates to extrapolate estimates for the entire United States.

We note that these estimates did not account for some inpatient testing that was paid through capitation arrangements for inpatient testing. Testing paid directly by patients was also not counted because, in these cases, CPT codes would not be captured in the data because there was no request for reimbursement. Even with this limitation, we believe that these estimates provide a relative sense of the number of tests being performed annually per analyte. No other accurate data were available to us.

As noted previously in this section, for the second cut, based upon our estimates of national testing volumes, we decided that an estimated national test volume of 500,000 per analyte annually was an appropriate threshold as most of the analytes listed in subpart I had national testing volumes above this threshold. Together with the above-described analytes below the 500,000 threshold that we determined to be clinically important, this narrowed our list of potential analytes under consideration for addition to subpart I to 73, representing analytes in five specialties or subspecialties

c. Impact on Patient and/or Public Health

For the third cut, we considered the evidence available related to each analyte under consideration to assess patient and public health impact of Start Printed Page 41200 testing. Because there was no standardized, generally accepted way to assess this impact on clinical care and public health, we used the following to get a relative sense of the importance of the analytes under consideration: a review of published laboratory practice guidelines (LPGs); a review of critical values; and a review of the analyte's classification by the Food and Drug Administration (FDA).^[4] We accessed several data sources, including tests listed in the CDC Guide to Community Preventive Services; ^[5] National Healthcare Priorities/Disparities reports; ^[6] clinical practice guidelines including the National Guideline Clearinghouse (NGC) database available from AHRQ ( https://www.guideline.gov/​ ); critical values available in publications; and (CAP) Q-Probes.

In reviewing published LPGs, we hypothesized that if there were a relatively large number of LPGs available for a particular analyte, that analyte would be important for health testing. To estimate the number of LPGs, we used the AHRQ's NGC database. For example, there were 60 LPGs listed in the NGC for LDL cholesterol, 31 for hemoglobin A1c, and 27 for troponin, all of which are proposed for addition in Table 1. However, this approach did not differentiate analytes for which there were conflicting recommendations. For example, there are controversies about the value of screening men with prostate specific antigen (PSA) testing, and there is an ongoing debate about the prudence of testing vitamin D in asymptomatic adults (Kopes-Kerr, 2013).

To review critical values, which are pre-determined limits for specific analytes that, when exceeded, may suggest that immediate clinical intervention is required, we assessed analytes included in published on “critical values” lists. This approach allowed us to gauge the importance of an accurate result because an incorrect result could lead to a life-threatening intervention or a failure to intervene. We reviewed published literature and critical values posted online from 16 institutions, including small hospitals, university hospitals, and reference laboratories.

As mentioned earlier in this proposed rule, we also assessed the clinical impact of an analyte by reviewing its medical device classification (Class I, II, or III) as categorized by the Food and Drug Administration's risk classification list. Similarly, we assessed the public health importance of the eligible analytes by counting the number of recommendations for testing the analytes from CDC's Morbidity and Mortality Weekly Report, the Infectious Disease Society of America, and the Council of State and Territorial Epidemiologists for surveillance of health conditions related to the particular analyte under consideration. We found supporting evidence for national prioritization in some of the following: the U.S. Preventive Services Task Force,^[7] the National Healthcare Quality and Disparities Report,^[8] and the CDC Hormone Standardization Program.^[9] For some analytes that are important to measure towards addressing health disparities and have public health impact, such as blood lead, we consulted with subject matter experts in CDC's National Center for Environmental Health, which promotes national testing and/or has standardization programs for some priority analytes, specifically estradiol and testosterone. CMS and CDC used this information to help determine which analytes should be included in the proposed rule.

After assessing patient and public health impact on a case-by-case basis for the third cut, we narrowed the analytes down to 34 for consideration of addition to the proposed list of analytes in subpart I.

d. Cost and Feasibility of Implementation

For the final analysis to determine whether an analyte would be proposed for inclusion in subpart I of the CLIA regulations, we focused on feasibility and costs of conducting PT for each of the remaining 34 analytes under consideration. We provided each of the HHS-approved PT programs the opportunity to submit comments in writing related to: inclusion/deletion of analytes, grading schemes, method(s) for determining target values, evaluating data using peer groups, cost of including new analytes, and structure of microbiology PT. Analytes for which it would be difficult for the PT programs to scale up production to meet the CLIA required frequency of three events per year with five challenges per event were eliminated from consideration because we believe that the costs passed down to laboratories to purchase the PT would be overly burdensome. In other cases, the decisions were based on the difficulty of finding any suitable PT materials. Some potential analytes were eliminated because they were too unstable for product development or shipping or because the testing methodology was not sufficiently standardized to support PT, such as vitamin D testing. After assessing the cost and feasibility of implementing PT on a case-by-case basis, we made the final cut, narrowing the analytes down to 29 potential analytes for the proposed list of analytes in subpart I.

3. Specific Analytes Proposed for Addition to Subpart I

Based upon the sequential process described previously in this final rule, information received from the PT programs, and consultation between CDC and CMS, we narrowed the list down to 29 analytes that we are proposing to add to subpart I of the CLIA regulations (Table 1).

4. Analytes Proposed for Removal From Subpart I

Recognizing that changes in the practice of clinical medicine have resulted in less frequent use of certain analytes, we used the same process to review the existing list of analytes in subpart I to determine which should be retained. In addition to requesting CLIAC's recommendations, we generally used the same criteria for retention of an analyte in subpart I as those used for determining which PT analytes to propose adding; however, as such PT testing was already available on the market, we did not consider the availability of PT material or the feasibility of implementation; therefore, we believe that PT programs already have the mechanism(s) in place to manufacture and ship PT for these analytes.

5. Process for Ranking and Assessing Existing Analytes and Proposals for Removal From Subpart I

a. Estimating Nationwide Testing Volume

We generally used the same rationale to select currently required analytes to propose for deletion. Specifically, we used the same threshold of 500,000 tests performed annually as an initial criterion for considering PT analytes. Those estimated to be lower than this threshold were considered for deletion from required PT. In particular, we focused on PT for several therapeutic drugs (ethosuximide, quinidine, primidone, and procainamide and its metabolite, N-acetyl procainamide). New drugs that are more effective or safer have entered the market since 1992 and may have replaced the use of therapeutic drugs that were included in the 1992 regulations. If so, we would expect to see a continued decline in the volume of testing for the use of such drugs. In addition to identifying decreases in testing for these drugs, we looked for probable causes of those decreases. These decreases in testing could result from new and emerging tests, including methodologies, replacing older tests, new technology, and changes to the way that the medical community orders laboratory testing. For example, the decrease in testing for LDH isoenzymes could be explained by the increased reliance on better alternative cardiac markers, especially troponin. For some anticonvulsant drugs, there may have been changes in medical practice, including alternative drugs and other treatments, possibly decreasing the need to measure them. We identified 13 currently required analytes with national test volumes less than our 500,000 annual test volume threshold.

b. Estimated Impact on Patient and Public Health

For any analyte still under consideration for removal, we performed literature reviews to determine if testing for alternative analytes or other diagnostic strategies had begun to supplant testing for the considered analyte. We took into account testing trends over the past 10 years and we attempted to project Start Printed Page 41202 expected testing trends. We then assessed the critical importance of candidates for deletion from subpart I based upon the number of guidelines available in the AHRQ NGC and the same sources used for considering inclusion in subpart I, bearing in mind that for all analytes and tests that are not listed in subpart I, laboratories must demonstrate accuracy twice per year as specified at § 493.1236(c)(1). We also considered the potential impact of deleting these analytes on clinical medicine and public health. Based on our literature review and consultation with CDC health experts, we decided not to propose the elimination of eight analytes based upon their critical importance for patient testing: carbamazepine, alpha-1-antitrypsin, phenobarbital, hepatitis B e antigen (HBeAg), antibody identification, theophylline, gentamicin and tobramycin. These are used for making important health decisions, for example, diagnosing hepatitis B (HBeAg), performing crossmatching for blood transfusions (antibody identification), or assessing compliance with medication for critically ill asthmatic patients (theophylline).

6. Analytes Proposed for Deletion From Subpart I

Based upon the sequential process described previously in this final rule, we proposed that the following analytes be deleted from subpart I: at § 493.931 LDH isoenzymes and at § 493.937 ethosuximide, quinidine, primidone, and procainamide (and its metabolite, N-acetyl procainamide).

7. Determining Criteria for Acceptable Performance

“Criteria for Acceptable Performance”, as that term is used in §§ 493.923, 493.927, 493.931, 493.933, 493.937, 493.941, and 493.959, is defined by the target value and acceptance limits. Criteria for acceptable performance is meant for PT scoring only and not intended to be used to set acceptability criteria for a laboratory's verification or establishment of performance specifications.

8. Setting Target Values

Under § 493.2, “target value” for quantitative tests is currently generally defined as either the mean of all participant responses after removal of outliers (those responses greater than 3 standard deviations from the original mean) or the mean established by definitive or reference methods acceptable for use in the National Reference System for the Clinical Laboratory (NRSCL) by the National Committee for the Clinical Laboratory Standards (NCCLS). However, in instances where definitive or reference methods are not available or a specific method's results demonstrate bias that is not observed with actual patient specimens, as determined by a defensible scientific protocol, a comparative method or a method group (“peer” group) may be used. If the method group is less than 10 participants, “target value” means the overall mean after outlier removal (as defined above) unless acceptable scientific reasons indicate that such an evaluation is inappropriate.

Based on input from PT programs, we recognize, that peer grouping is generally the way that target values are set for most analytes. Therefore, in the proposed rule, we proposed to continue allowing PT programs to use peer grouping to set the target values. In addition, we proposed removing the reference to the NRSCL and NCCLS, while retaining the other options for setting target values.

9. Changing Acceptance Limits

Because there have been improvements in technology resulting in better sensitivity, specificity, and precision, routinely using peer grouping to set target values means that the AL that were originally specified in each specialty and subspecialty of the CLIA `88 regulations in subpart I effectively allow for more tolerant acceptance criteria for most analytes than would occur if targets were set by a reference method or overall mean. Based on feedback from several HHS-approved PT programs, we believe it would be appropriate to update the ALs to reflect advancements in technology and analytical accuracy since the PT regulations were implemented in 1992. While narrowing limits may increase miss rates per challenge, we do not expect a high unsuccessful rate based on the data simulations provided by the PT programs. We expect the rates of unsatisfactory events would be low based on the simulation data and that the rates of unsuccessful events (two consecutive or two out of three testing events being unsatisfactory) would be even lower; therefore, we believed it was reasonable to propose tighter limits given current analytic accuracy. We used all data available to us to minimize the negative consequences of the proposed changes (for example, too many unsuccessful performances) to acceptance limits, including simulations provided by PT programs.

10. Changes to Percentage Acceptance Limits (ALs)

a. Basis for Using Fixed Percentage PT ALs

Currently, the CLIA regulations at §§ 493.927(c)(2), 493.931(c)(2), 493.933(c)(2), 493.937(c)(2), and 493.941(c)(2) prescribe a variety of ALs, including: a multiple of the SD of results from the mean of other participants in the peer group; fixed limit as a percentage of the assigned value; fixed limit in concentration units; and a mixture of percentage and concentration units, depending on the concentration of the analyte. For all new and currently required non-microbiology analytes, we proposed to use fixed ALs, preferably as percentage limits rather than concentration units.

There are 53 analytes (existing or proposed) for which we proposed a percentage-based AL, for which biological variability data were published. There were no biological variability data for several analytes (for example, therapeutic drugs). Where there were such data, we used AL to get as close to, or below, an accuracy goal for the test that was based on biological variability data. Then we simulated several percentage-based ALs to see if their results would have passed or failed at each simulation. We wanted to get miss rates (that is, percent of laboratories that did not meet the criteria for acceptable performance per PT challenge) of somewhere in the 1 to 2 percent range as was observed in the data provided by the PT programs for current ALs. Of the 53 analytes, 34 of the proposed ALs were tighter than or equal to biological variability limits. For 19 analytes, the limits we are proposing are looser (greater) than the limits required to meet accuracy based upon biological variability. For these 19 analytes, using ALs based on biological variability would be untenable because the current analytical accuracy for such testing would not be expected to meet such limits. White blood cell differential is the only remaining analyte that would have ALs in SD. In this case there were no biological variability data available.

In general, fixed ALs, either in percentages or concentration units, are preferred to SDs for PT for several important reasons: they can be tied directly to objective goals for performance, such as goals for analytical accuracy and technical expectations; they are constant in all PT events and do not vary because of statistical randomness, masked outliers, or small sample size; they assure the same evaluation criteria are used by all PT programs and discourage opportunities Start Printed Page 41203 for participants to “shop” for PT programs with less stringent criteria for which it is easier to achieve acceptable performance; they do not unfairly result in tighter effective ALs for peer groups that use analyzers that have tighter analytical precision; they can combine a fixed percentage and a fixed absolute concentration to allow for more robust evaluation while also fairly evaluating low analyte concentrations; and they are commonly used worldwide in other PT and external quality assessment programs.

Our analysis of existing PT and external quality assessment programs showed that ALs using two or three SDs have been used in PT in a wide variety of settings for several reasons, such as: limited experience with PT or matrix effects for a particular analyte; lack of consensus on criteria for acceptable performance; inertia with no compelling pressure for change; and analytical performance so poor that multiples of the overall SD are considered to be the only fair approach. We believe all of these reasons to some extent contributed to initial reliance on SD limits for certain analytes when CLIA `88 was implemented. We also note that while regulations promulgated under CLIA `67 used ALs of three SD for several analytes, regulations finalized under CLIA`88 replaced these with fixed limits and PT programs successfully made the transition. Therefore, we believe it is likely that the proposed changes from SD-based ALs to fixed ALs will not be problematic.

Therefore, as discussed in section II.B. of the proposed rule, we proposed to amend certain analytes in §§ 493.927, 493.931, 493.933, 493.937, and 493.941 to include fixed ALs with or without percentages. Three analytes have only concentration-based ALs (that is, no percentage-based ALs): pH, potassium, and sodium.

b. Adding Fixed Concentration Units to Fixed Percentage Units

A percentage-based criterion can be unnecessarily stringent at low concentrations—either because of technical feasibility or because medical needs at the low concentration do not require such tight precision. Thus, when percentage-based fixed criteria are used for ALs, it may be necessary to place a minimum on the percentage as currently occurs with the criterion for acceptable performance for glucose (§ 493.931) for which the AL switches from 10 percent to 6 mg/dL below a concentration of 60 mg/dL. The combined ALs direct PT programs to score with whichever of the specifications is more tolerant; at lower limits of the analytical range this will be the fixed concentration limit. Therefore, to allow for fairer and more realistic ALs, we proposed to use combinations of percentage and concentration limits as appropriate. These combination limits are similar to limits that already exist in CLIA `88 regulations for glucose and other analytes.

Therefore, we proposed to amend certain analytes in §§ 493.927, 493.931, 493.933, 493.937, 493.941, and 493.959 to include percentage-based ALs with or without additional fixed ALs.

c. Establishing ALs Based on Analytical Accuracy Goals for Proposed New and Several Current Analytes

For the newly proposed analytes and several current analytes for which current ALs are in units other than percentages such as three SDs or concentration units, we proposed to change the ALs to percentages. Over the years, there have been many proposed criteria for establishing goals for analytical performance. The various possible approaches were reviewed and a hierarchy was established based on a 1999 consensus conference. These strategies were reconsidered at the 2014 European Federation of Clinical Chemistry and Laboratory Medicine Strategic Conference in Milan. Participants in both conferences acknowledged that the ability of a test method to meet clinical needs is the highest priority, and the most defensible approach would be clinical trials in which patient outcomes could be compared using different analytical accuracy goals. This approach was not feasible for many reasons. Although clinical outcomes studies would be the most rigorous basis for establishing analytical performance goals, these are seldom possible, leaving the natural dispersion of levels for each analyte (biological variability) as the next best scientifically defensible approach for establishing analytical accuracy goals. The less the biological variability, the more stringent the analytical accuracy needs to be. This approach makes sense for two of the most important reasons to conduct patient testing: diagnosis of disease, that is, differentiating an abnormal result from a normal one, and monitoring a patient's progress during treatment. In the former case, we believe that the “within-group” biological variability is the important limiting factor defining an appropriate error goal for a test method. Furthermore, we believe the most important factor for monitoring progress is the “within individual” variability. It was not possible for us to differentiate how analytes are being used or will be used clinically, with respect to diagnosis versus monitoring. Therefore, we accounted for both needs and used an approach that accounted for both kinds of biological variability to estimate analytical accuracy goals as the basis for our proposals for acceptance limits in percentages. The advantage of using analytical accuracy goals that are expressed in terms of percentages is that they can be directly related to ALs in a mathematical way expressed as percentages.

We have assumed that a laboratory that can meet the clinical needs for test accuracy based upon biological variability should perform successfully on PT most or all of the time. Therefore, whenever possible, we have used publicly available estimates of allowed total error based upon estimates of biological variability to approximate the proposed AL. CDC has shown in a recent poster ^[10] that it is possible to design ALs based upon such accuracy goals, and it is possible to simulate the ability of a PT program to identify laboratories that cannot meet such goals, while minimizing the likelihood of misidentifying laboratories that are meeting analytical accuracy goals based upon biological variability.

Therefore, we proposed to amend ALs for certain current analytes as well as establish ALs for analytes proposed for addition in §§ 493.927, 493.931, 493.933, 493.937, 493.941 and 493.959 based on analytical accuracy goals.

d. Tightening Existing Percentage ALs as Needed

There have been significant improvements in laboratories' performance in PT for the great majority of analytes and PT unsatisfactory rates have dropped for all types of laboratories. The improvements are such that, for many analytes, laboratories that began to use PT to comply with CLIA `88 now perform as well as the hospital and independent laboratories that were previously required to perform PT under CLIA `67. Howerton, et al., showed that for almost all analytes examined, PT performance improved somewhat after CLIA `88 was implemented, but the improvements were greater for laboratories that were not previously required to perform PT. The rates of unsatisfactory PT are now roughly the same for analytes listed in subpart I, regardless of the laboratory type. This is consistent with CLIA's intent to ensure accurate clinical testing regardless of the setting where testing is performed. There are several factors Start Printed Page 41204 contributing to the improvements in PT performance, including improved analytical methods being used in all settings, technological advances resulting in improved precision, sensitivity and specificity, and increased familiarity with handling preparation, and reporting of PT samples. Therefore, for the reasons above as well as supporting simulation data date from the PT programs, we proposed to make criteria for acceptable performance for existing analytes listed in subpart I (§§ 493.927, 493.931, 493.933, 493.937, 493.941 and 493.959) tighter, so they are in closer agreement with analytical accuracy goals which are based upon biological variability and simulation data.

e. Simulating the Impact of New ALs on Unacceptable Scores for Challenges and Unsatisfactory Rates for Events

We evaluated a very specific PT data set to help set appropriate limits. The total simulations reproduced PT that covered 2 years, representing 30 challenges (three events per year; five challenges per event; 2 years) of each proposed new analyte and for the analytes for which we propose to modify ALs. We reviewed the aggregated percentage of unacceptable scores for each PT challenge using retrospective data. We then reviewed the simulation data which applied two or three new ALs for each of 84 analytes (consisting of 27 new analytes and 57 existing analytes). Based on the simulation data, we were able to make informed decisions to help us create or adjust the ALs.

Based upon our analysis of the simulation results, we further refined the proposed ALs and added potential absolute concentrations in lieu of percentage ALs, as was described previously in this final rule. We then requested narrowly tailored data from PT programs as described previously in this final rule using retrospective PT data and peer group data for scoring, as they ordinarily would do. We focused on unsatisfactory scores with the data so that we could calculate the unsatisfactory rate per analyte among all participating laboratories that might occur with each proposed AL. The final simulations were conducted by several of the PT programs and this set of data was used to determine the proposed ALs.

We compared the unacceptable scores for each challenge and each proposed AL to determine at which concentrations it would be necessary to switch to a fixed concentration AL. Using this approach, we were able to identify an AL for each analyte and, in some cases, an additional concentration-based AL. This approach enabled us to identify an AL that would be sensitive enough to identify poor-performing laboratories, yet not so sensitive that it will incorrectly identify laboratories that likely meet requirements for accuracy.

f. Limitation in Our Ability To Predict the Number of New Unsatisfactory and Unsuccessful Scores

It is not possible for us to predict the precise effect of the proposed changes on the number of unsatisfactory and unsuccessful scores. The occurrence of an unsatisfactory score for a PT event depends upon at least two of five challenges being graded as unacceptable or outside the criteria for acceptable for performance. PT programs select different combinations of samples for each event and it is impossible to predict how their selection could be modeled statistically. Finally, the distribution of unsatisfactory and unsuccessful PT scores is not randomly distributed across all participants.

++ Sections 493.923(a), 493.927(a), 493.931(a), 493.933(a), 493.937(a), 493.941(a), and 493.959(b): We proposed to amend these provisions to remove the option that PT samples, “at HHS' option, may be provided to HHS or its designee for on-site testing”.

++ Section 493.927: We proposed to amend the criteria for acceptable PT performance to permit scoring of quantitative test results for the following immunology analytes: antinuclear antibody; antistreptolysin O; rheumatoid factor; and rubella. For these analytes, we have determined that there are one or more test systems that currently report results in quantitative units; therefore, we added ALs based on percentages or target values in addition to retaining the qualitative target values. We proposed to make this allowance in CLIA for reporting PT which reflects current practice.

++ Section 493.931(b): We proposed making a technical change to the description for creatine kinase isoenzymes to be CK-MB isoenzymes, which may be measured either by electrophoresis or by direct mass determination.

++ Section 493.933: We proposed adding the following analytes: estradiol, folate (serum), follicle stimulating hormone, luteinizing hormone, progesterone, prolactin, parathyroid hormone, testosterone, and vitamin B12.

++ Section 493.937(a): We proposed revising this provision by including the requirement that annual PT programs must provide samples that cover the full range of values that could occur in patient specimens. We proposed this amendment so that PT programs must provide samples across a toxicology sample's entire reportable range rather than just provide samples within a sample's therapeutic range.

++ Section 493.941: We differentiated the criteria for units of reporting of the analyte prothrombin time. We proposed to amend the criteria for acceptable performance to reflect both in seconds and/or INR (international normalized ratio) and to add the requirement that laboratories must report prothrombin time for PT the same way they report it for patient results. We also proposed to add criteria for acceptable performance for directly measured INR for prothrombin time. Additionally, we proposed to require laboratories performing both cell counts and differentials to conduct PT for both (that is, the “or” would be changed to an “and”). Finally, we proposed changing the criteria for acceptable performance for “cell identification” from 90 percent to 80 percent. We proposed this change as the requirement of five samples per event does not allow for a score of 90 percent (that is, five samples would allow for scores of zero percent, 20 percent, 40 percent, 60 percent, 80 percent, or 100 percent). PT for cell identification is currently required in § 493.941. Further, § 493.851(a) states that “failure to attain a score of at least 80 percent of acceptable responses for each analyte in each testing event is unsatisfactory performance for the testing event.” If the requirement for acceptable performance remains at 90 percent, a laboratory can only have satisfactory performance if they receive 100 percent; however, § 493.851(a) allows satisfactory performance for both 80 percent and 100 percent.

++ Section 493.959: We proposed changing the criteria for acceptable performance for unexpected antibody detection from 80 percent accuracy to 100 percent accuracy. We proposed this change because it is critical for laboratories to identify any unexpected antibody when crossmatching blood in order to protect public health and not impact patient care.

++ Sections 493.923(b)(1), 493.927(c)(1), 493.931(c)(1), 493.933(c)(1), 493.937(c)(1), 493.941(c)(1), and 493.959(d)(1): We proposed amending these provisions to clarify that to achieve consensus, PT programs must attempt to grade using both participant and referee laboratories before determining that the sample is ungradable. We believe that this change will enhance consistency among the PT programs when grading samples. The current regulations noted previously Start Printed Page 41205 allow for scoring either with participants or with referees before calling a sample ungradable.

C. Additional Proposed Changes

We proposed to amend § 493.2 by modifying the definition of an existing term and defining new terms as follows:

• Target value: We proposed removing the reference to NRSCL and NCCLS and retaining the other options for setting target values in this final rule.

• Acceptance Limit: We proposed defining this term to mean the symmetrical tolerance (plus and minus) around the target value.

• Unacceptable score: We proposed defining this term to mean PT results that are outside the criteria for acceptable performance for a single challenge or sample.

• Peer group: We proposed defining this term as a group of laboratories whose testing process utilizes similar instruments, methodologies, and/or reagent systems and is not to be assigned using the reagent lot number. PT programs should assign peer groups based on their own policies and procedures and not based on direction from any manufacturer.

We proposed the following revisions to the regulation text at subpart A:

• Sections 493.20 and 493.25: We proposed to amend the regulations to reflect that if moderate and high complexity laboratories also perform waived tests, compliance with § 493.801(a) and (b)(7) are not applicable. However, we proposed to continue to require compliance with § 493.801(b)(1) through (6) to align the regulations with the CLIA statute (42 U.S.C. 263a (i)(4)), which does not exclude waived tests from the ban on improper PT referral.

We proposed to revise the regulation text at subpart H:

• Section 493.861: We proposed amending the satisfactory performance criteria for failure to attain an overall testing event score for unexpected antibody detection from “at least 80 percent” to “100 percent.” We proposed this change because it is critical for laboratories to identify any unexpected antibody when crossmatching blood to protect the public health and not impact patient care.

We proposed to revise the regulation text at subpart I:

• Section 493.901(a): We proposed to require that each HHS-approved PT program must have a minimum of 10 laboratory participants before offering any PT analyte. We recognize that PT programs do not grade results when there are fewer than 10 laboratory participants. This would require the laboratory to perform additional steps to verify the accuracy of their results. If at any time a PT program does not meet the minimum requirement of 10 participating laboratories during the reapproval process for an analyte or module, HHS may withdraw approval for that analyte, specialty, or subspecialty. This change reduces some burden on laboratories that have incurred the expense of enrolling in a PT program but do not receive a score or receive an artificial score requiring the laboratory to take additional steps to verify the accuracy of the analyte as required by § 493.1236(b)(2).
• Section 493.901(c)(6): We proposed to add the requirement that PT programs limit the participants' online submission of PT data to one submission or that a method be provided to track changes made to electronically reported results. Many PT programs currently allow laboratories an option to report PT results electronically, while some other PT programs only allow laboratories to report PT results electronically with no other option such as facsimile or mailed PT submission forms. However, at this time, the PT programs that do participate in the online reporting have no mechanism to review an audit trail for the submitted result. In some cases of PT referral, it has been discovered that laboratories have sent PT samples to another CLIA-certified laboratory for testing, received results from the other laboratory, and then changed their online reported results to the PT program since those results can be modified up until the PT event close date. In an effort to assist in PT referral investigations and determinations, an audit trail that includes all instances of reported results would aid in determining if a laboratory compared PT results obtained from another laboratory and changed their previously submitted results.
• Section 493.901(c)(8): We proposed to add to the requirement previously found at § 493.901 that contractors performing administrative responsibilities as described in §§ 493.901 and 493.903 must be a private nonprofit organization or a Federal or State agency or nonprofit entity acting as a designated agent for the Federal or State agency. Several PT programs have divided their administrative and technical responsibilities into separate entities or have had the administrative responsibilities performed by a contractor. We were made aware that administrative responsibilities were being performed by a for-profit entity. Because the CLIA statute (42 U.S.C. 263a(f)(3)(C)) requires PT programs to be administered by a private nonprofit organization or a State, we are proposing to amend § 493.901 to state that all functions and activities related to administering the PT program must be performed by a private nonprofit organization or State.
• Section 493.901(e): We proposed the requirement that HHS may perform on-site visits for all initial PT program applications for HHS approval and periodically for previously HHS-approved PT programs either during the reapproval process or as necessary to review and verify the policies and procedures represented in its application and other information, including, but not limited to, review and examination of documents and interviews of staff.
• Section 493.901(f): We proposed an additional requirement to the regulation that specifies we may require a PT program to reapply for approval using the process for initial applications if widespread or systemic problems are encountered during the reapproval process. The initial application for the approval as an HHS PT program requires more documentation in the application process than that which is required of PT programs seeking HHS reapproval.
• Section 493.903(a)(3): It has come to our attention that PT programs may have on occasion modified a laboratory's PT result submission by adding information such as the testing methodology which was inadvertently omitted by the laboratory. Therefore, we proposed adding the requirement that PT programs must not change or add any information on the PT result submission for any reason, including, but not limited to, the testing methodology, results, data, or units.
• Section 493.905: We proposed adding that HHS may withdraw the approval of a PT program at any point in the calendar year if the PT program provides false or misleading information that is necessary to meet a requirement for program approval or if the PT program has failed to correct issues identified by HHS related to PT program requirements. We also proposed adding a requirement that the PT program may request reconsideration should we determine that false or misleading information was provided if the PT program has failed to correct issues identified by HHS related to PT program requirements.

III. Analysis of and Responses to Public Comments

We received 107 public comments in response to the February 4, 2019, Start Printed Page 41206 proposed rule. The commenters represented individuals, PT programs, accreditation organizations, laboratory professional organizations, and businesses, including in vitro diagnostics manufacturers. Commenters were generally supportive of the proposed changes, and some noted that these changes would increase flexibility and be a positive change for both laboratories and PT programs, especially in the specialty of microbiology. A few commenters recommended clarification of proposed changes or suggested specific changes, including alternative language, to the proposed requirements. After analyzing the comments received, we have modified or deleted several provisions in this final rule. A few commenters raised issues that are beyond the scope of our proposals. We are not summarizing or responding to those comments in this final rule. However, we reviewed the comments to consider whether to take other actions, such as revising or clarifying the CLIA program operating instructions or procedures, based on the information or recommendations in those comments. Our responses to specific comments are as follows:

A. Delayed Effective Date and Ongoing Process for Updating PT Regulations (§§ 493.2 and 493.801 Through 493.959)

Comment: Several commenters requested that there be a delayed effective date or phase in approach for implementation of the updated PT requirements to give all affected constituents time to accommodate the changes. Two commenters suggested that CMS develop an ongoing process to make changes to the PT regulations to ensure timely implementation of the updates.

Response: We recognize that time will be needed for laboratories, PT programs, accreditation organizations, exempt States, and surveyors to adopt the updated PT requirements related to subparts H and I. As such we are delaying the effective date of the revisions to §§ 493.2 and 493.801 through 493.959 until 2 years after the publication of this final rule in the Federal Register . The delayed effective date reflects the timeframe that we believe PT programs will need to produce the PT samples to meet the revised regulations and incorporate any updates to PT reporting requirements. In addition, laboratories will need to implement the new PT requirements after the samples are available from the PT programs. We encourage laboratories to enroll in the new and revised analytes prior to the delayed effective date. We also appreciate the commenters' suggestions for a process to address needed PT changes more quickly on an ongoing basis. We will consider possible ways to streamline the process going forward in light of the required timeframe for rulemaking. We note that the regulations related to laboratories performing tests of moderate complexity and high complexity testing that also perform waived testing and proficiency testing enrollment, §§ 493.20 and 493.25, respectively, will be effective 30 days after the publication date of this final rule.

B. Definitions (§ 493.2)

Comment: A commenter stated that the term “unacceptable score,” as defined at § 493.2, was confusing and should be replaced with “unacceptable result.” Other commenters pointed out that the organization of sub-bullets under the definition of “target value” was incorrect as the content in (iv) does not belong under (1), but should be included as (2) under the definition.

Response: We agree with commenters that the term “unacceptable score” could be confusing because it could be interpreted to mean a total analyte event score rather than the intended meaning of referring to a single challenge or sample result. Since this term is not included in the CLIA regulations except for the proposed amendments to § 493.2, we are not finalizing this term in § 493.2 in this final rule. With respect to the proposed definition of “target value”, we agree with the commenter about the paragraphs included under that definition and are making the recommended change in this final rule.

Comment: While several commenters supported the inclusion of a definition for “peer group” in the proposed rule, other commenters expressed concerns about our proposal. Three commenters approved of our proposal to disallow peer-grouping to the reagent lot level, while two commenters did not agree with the proposal. One commenter noted that matrix effects, known to cause PT materials to behave differently from unmodified patient samples, are the reason underlying the need to use peer grouping to set target values and grade PT results. This commenter was concerned that the final rule would not account for the existence of matrix effects by not allowing peer grouping. One commenter suggested we consider conducting a scientific study to assess the contribution of calibration errors versus matrix effects in causing differences in PT results.

Response: In response to the comments about peer-grouping to the reagent lot level, PT is one of the important ways to detect problems in FDA-cleared/approved test methods. Differences between reagent lots used during testing may occur due to the manufacturing process. Allowing peer grouping to the lot level may inhibit the detection of these problems. We are not prohibiting PT programs from interacting with manufacturers to discover problems with reagent lots. However, the PT program has the responsibility for interpreting correct PT results. If a PT program determines that a specific reagent lot failure occurred, it should inform the affected laboratories and manufacturer. Concerning the comment about matrix effects, currently CLIA requires PT programs to demonstrate through a scientific protocol that bias, such as matrix effects, existed in PT materials before allowing peer-grouping to grade results. We are aware that PT programs have typically not used a scientific approach to determine if a peer group should be used as the process of demonstrating matrix effects is expensive and time-consuming. This rule finalizes the proposed definitions for both “peer group” and “target value” and will continue to allow peer-grouping for evaluation of PT results, without requiring prior demonstration of matrix effects. We do not expect there will be a change in how peer groups are identified by PT programs. Therefore, there will be no change in how target values are determined based upon the mean of peer group results. In response to the proposed study of commutability to demonstrate differences in PT results based on calibration errors, the comment is outside the scope of this final rule.

Comment: Two commenters suggested that CLIA should not require removal of outliers using a three standard deviation (3 SD) criterion when grading PT, as required under the proposed definition of target value in § 493.2. One commenter noted that the requirement to remove outliers was done to get a better estimate of the SD, which would only apply to one analyte after the final rule is effective. The other commenter stated that outlier removal using a 3 SD limit is not recommended according to ISO 13528:2015. Both commenters noted the need for robust methods to remove outliers, which can be especially problematic when the PT peer group is very small, such as a group that includes only 5 to 20 results.

Response: It is important that outliers be removed to set target values. Because a spurious PT result, including one due to a transcription error, could affect the peer group mean, especially when the peer group has relatively few laboratory Start Printed Page 41207 participants, PT programs should continue to discard aberrant results when calculating the peer group target. At this time, we do not have sufficient information to provide additional or alternative options for outlier removal. However, we recognize the need for PT programs to have valid modern approaches for outlier removal. Therefore, we are retaining the requirement to remove outliers as described in the definition for target value, using a 3 SD criterion. Regarding the comment referencing ISO requirements, we note that ISO standards do not apply to CLIA.

Summary of Final Actions

• We did not receive any comments on the proposed definition of “acceptance limit” and are finalizing the definition with a clarifying technical edit.
• Based on the public comments received, we are finalizing the proposed definition of “peer group” with a clarifying technical edit.
• We are revising and finalizing the proposed definition for “target value.” We have corrected the organization of the paragraphs and have moved the content of subparagraph (iv) to paragraph (2).
• We are not finalizing the proposed definition of “unacceptable score.”

C. Enrollment and Testing of Samples (§§ 493.20(c) and 493.25(d))

Comment: A number of commenters expressed concerns or requested clarification about the proposal to amend §§ 493.20(c) and 493.25(d) to reflect that if laboratories certified to perform moderate and high complexity testing, respectively, also perform waived tests, compliance with § 493.801(a), which requires enrollment in PT, and (b)(7), requiring PT for the primary method of patient testing, are not applicable for the waived tests. However, as proposed, if laboratories voluntarily enrolled in PT for their waived testing, § 493.801(b)(1) through (6) would apply in cases of improper PT referral for those tests. Commenters expressed that laboratories may be discouraged from voluntarily enrolling in PT for waived tests if the possibility of sanctions for referred PT existed. Two commenters recommended that PT should be required for all testing, including waived testing. One commenter requested clarification of whether laboratories would need to verify the accuracy of waived tests twice per year.

Response: Subsection (d)(2)(C) of the CLIA statute states that subsections (f) and (g) shall not apply to a laboratory issued a Certificate of Waiver. Subsection (f) is related to issuing standards that, at a minimum, allow a laboratory to consistently perform testing to ensure accurate and reliable test results, including the requirement for all laboratories that perform nonwaived testing to enroll in an approved PT program and to verify the accuracy of tests twice per year. Subsection (g) speaks to inspecting laboratories for compliance with subsection (f) and are generally done on a biennial basis. However, sanctions related to PT referral are in subsection (i), which is not limited to nonwaived laboratories but rather allows sanctions to be taken against “any laboratory”, including a Certificate of Waiver laboratory, that intentionally refers PT samples to another laboratory. Some Certificate of Waiver laboratories and other laboratories that perform waived testing have voluntarily chosen to enroll in PT for waived testing over the history of the CLIA program to ensure the quality of their testing. We have no reason to believe these laboratories will be discouraged from continuing their enrollment in PT. As a result, we are finalizing the new requirements at §§ 493.20(c) and 493.25(d) to ensure that the CLIA regulations align with the statute.

Summary of Final Actions

• We are finalizing the proposed revisions at §§ 493.20(c) and 493.25(d).
• We are finalizing the proposed revisions at §§ 493.801 and 493.861. Section 493.801 will require laboratories to report PT results for microbiology organism identification to the highest level that they report results on patient specimens. Section 493.861 will amend the satisfactory performance criteria for failure to attain an overall testing event score for unexpected antibody detection from “at least 80 percent” to “100 percent.” We received no comments on the proposed revisions at §§ 493.801 and 493.861.

D. PT Program Approval and Administration (§§ 493.901, 493.903, 493.905)

Comment: Two commenters urged CMS not to change the current codes used for specific analytes when PT programs report PT results to CMS and to create new codes for the analytes being added.

Response: We understand the commenters to be referring to certain analyte-specific codes that are used as an internal data system designation for PT programs to report PT analyte results to us. Although these codes are not explicitly referenced in the regulations, we agree with the commenters and note that the current analyte-specific codes for PT will remain the same. New analyte-specific codes will be generated for the newly required PT analytes.

Comment: Many commenters remarked on the requirement proposed at § 493.901(a) having at least 10 laboratory participants for an analyte before a program is approved to offer that analyte. Commenters stated that this requirement could inhibit development of new PT, and be detrimental to both laboratories and PT programs, especially smaller programs, which could find it harder to compete. Some commenters pointed out that PT programs offering newly required analytes would naturally have relatively fewer participating laboratories. One commenter requested clarification on whether this requirement would apply only to newly required analytes or to all PT analytes. Some commenters pointed out that PT programs may not initially know how many laboratories would enroll, and the programs would need time to develop their market. One commenter stated that this requirement would be a burden and result in more ungraded events.

Response: The requirement for at least 10 laboratory participants would only apply for PT analytes required in subpart I, and therefore, should not impact the development of PT for new or emerging analytes to the extent that they are not listed in subpart I. We realize that PT programs seeking HHS approval for the first time may not know how many laboratories would enroll in their program, and we did not intend to require at least 10 laboratory participants when PT programs apply for initial approval. We intend to review the number of laboratory participants for each program and each HHS-approved analyte during the annual reapproval process. If a PT program has fewer than 10 participants, we may not reapprove the PT program for a specific analyte. As a result of the comments, in this final rule, we are clarifying the requirement at § 493.901(a) to state “for each specialty, subspecialty, and analyte or test for which the proficiency testing program is seeking reapproval” to better reflect the PT approval process.

Comment: A number of commenters representing several PT programs and accreditation organizations commented on the requirement proposed at § 493.901(c)(6) that for those results submitted electronically, a mechanism to track changes to any result reported to the proficiency testing program and the reason for the change. There was general opposition due to perceived burden and expense, both to PT Start Printed Page 41208 programs and laboratories, and possibilities for errors. Some commenters stated that they are currently unable to know when every PT result is entered or changed if done electronically based on the technology used for laboratories to submit results. There were also questions about the circumstances under which PT programs would be required to provide audit trails. One commenter agreed with this proposed change but recommended that we provide more guidance to laboratories on how to meet this requirement.

Response: We appreciate the information provided by the commenters expressing the challenges with meeting this requirement. We do require laboratories to maintain documentation of their submissions to PT programs (see § 493.801(b)(5)). However, based on the comments received, we are not finalizing the requirement proposed at § 493.901(c)(6).

Comment: Several commenters expressed concerns about the requirement proposed at § 493.901(c)(9) that a contractor performing administrative responsibilities as described in §§ 493.901 and 493.903 must be a private nonprofit organization or a Federal or State agency, or an entity acting as a designated agent for the Federal or State agency. A commenter noted that many essential PT program functions are currently performed by for-profit entities or subcontractors. There was a general consensus among commenters that many important administrative functions could not be performed without contractual arrangements with for-profit entities, such as transportation services.

Response: We recognize that some functions required as part of the PT process, such as transportation services, are provided by for-profit entities. Other business functions may also be provided by for-profit contractors, such as obtaining and manufacturing the PT specimens/products, initial testing to establish approximate target values as prescribed by the PT program, aliquoting and labeling samples, testing to assure homogeneity and stability of samples, long-term storage of samples for use in future PT events, and storage of aliquoted PT samples for additional testing as may be requested by the clients, or required by us. Also, “for-profit” entities can be used or contracted for distributing/mailing out the PT kits to the laboratories. This proposed requirement was not intended to address those aspects of PT program operations, but rather the technical and scientific responsibilities as described in §§ 493.901 and 493.903. These technical and scientific responsibilities include, but are not limited to, processes for selecting appropriate target values to be included in challenges as part of the annual PT program or grading PT results, determining target values, reporting scores to CMS, and determining organisms included in microbiology PT samples. In an effort to clarify the intent of the proposed requirement, we are changing “administrative responsibilities” to “technical and scientific responsibilities” in the provision being finalized at § 493.901(c)(8), previously proposed at § 493.901(c)(9).

Comment: While commenters agreed with the requirement proposed at §§ 493.901(e) to allow HHS to require on-site visits as part of the initial approval of PT programs, they indicated the need for sufficient advance notice of an on-site visit. Also, there were two suggestions to use an independent third party if on-site visits were to be conducted.

Response: We would coordinate the timing of the visit with the PT program and generally provide advance notice of the on-site visit. On-site visits will be conducted by CMS, and not by a third party. As a result, we are finalizing the new requirement at § 493.901(e) as proposed.

Comment: We received comments concerning the requirement proposed at § 493.901(f) that HHS may require a PT program to reapply for approval using the process for initial applications if significant problems are encountered during the reapproval process. While no commenters disagreed with the proposed requirement, one commenter requested that we use this option sparingly, and another commenter requested clarification on when this option would be used.

Response: We intend to use this option cautiously and only when issues arise that we consider be significant, for example, complaints of quality issues related to the PT program. As a result, we are finalizing the new requirement at § 493.901(f).

Comment: Commenters suggested clarification was needed regarding the requirement proposed at § 493.903(a)(3) that PT programs must not change or add any information on the PT result submission. They requested clarification on what data could not be changed, noting that some changes, such as adding or changing a method code, would not necessarily affect test results submitted but would be important for appropriate peer grouping. Commenters expressed concern that PT programs would not be able to add a methodology if inadvertently left off by the laboratory, thus affecting appropriate peer grouping. Commenters questioned if exceptions might be made if errors were made by the PT program and not the laboratory.

Response: As explained in the proposed rule (84 FR 1536, 1547), it is not appropriate for a PT program to change or add information on the PT result submission from a laboratory, including, but not limited to, the testing methodology, results, data, or units. If a laboratory inadvertently enters the wrong methodology or omits a methodology, the PT program should not assume to know the correct methodology and make that change or addition. We would consider it acceptable for the PT program to enter the methodology in cases where the PT program form does not include the methodology used by the laboratory for testing and the laboratory has manually written the methodology on the result submission form. This would also apply to units of measure. Under no circumstances should a PT program change a laboratory's submitted result. It is the laboratory's responsibility to provide correct and complete information and to investigate and correct errors that lead to PT failures. As a result, we are finalizing the requirement at § 493.903(a)(3) as proposed.

Comment: Commenters expressed concerns regarding the potential impact on laboratories and PT programs of the requirement proposed at § 493.905(a) allowing HHS to withdraw the approval of a PT program at any point in the calendar year if the PT program provides false or misleading information required for program approval or if the PT program fails to correct issues identified by HHS related to PT program requirements.

Response: We may withdraw approval of the PT program if HHS determines the PT program fails to meet any of the required criteria for approval. After we withdraw approval of a PT program, approval of the PT program would remain in effect for 60 days from the date of written notice to the PT program of this action. A PT program will be required to notify all of its participating laboratories of our withdrawal of approval within 30 days from the date of written notice to the PT program. We believe the 30-day notification by the PT program in this situation, and the additional 30 days before approval is withdrawn, gives laboratories sufficient time to enroll in an alternative PT program. PT programs may request reconsideration from us in accordance with subpart D of part 488 regarding the Start Printed Page 41209 withdrawal of approval if the false or misleading information or issues identified by us have been addressed within 60 days. We believe that the 60-day timeframe gives the PT programs sufficient time to mitigate any issues related to withdrawal of approval.

Summary of Final Actions

• We are finalizing the proposed changes to §§ 493.901(a), (c)(8), (e), (f), 493.903(a)(3), and 493.905.
• Based on comments received, we are not finalizing the proposed addition at § 493.901(c)(6).

E. Proposed Changes to Microbiology PT (§§ 493.911 Through 493.919)

Comment: Commenters suggested clarification is needed regarding methods or platforms for which PT is proposed to be required, specifically for laboratories that use molecular, nucleic acid amplification, mass spectrometry testing or next generation sequencing for microorganism identification and susceptibility testing in all microbiology subspecialties. A commenter also questioned whether PT is required only for FDA-cleared test systems. The commenters stated this clarification would help prevent confusion among laboratories.

Response: PT is not required by method or specific technology for microbiology subspecialties (§§ 493.911 through 493.919), including whether a test system is FDA-cleared, or analytes in non-microbiology specialties or subspecialties (§§ 493.921 through 493.959). Regardless of the method, a laboratory uses for microorganism identification and susceptibility testing, PT is required for these categories of microbiology testing. When CLIAC deliberated on appropriate PT for microbiology, they suggested the inclusion of broad categories of testing performed in microbiology, rather than the types of services offered by laboratories, described in §§ 493.911 through 493.919, to allow flexibility for the inclusion of new technologies. Each laboratory needs to identify the method or test system used when submitting PT results for programs to properly grade the PT. If a laboratory performs microbiology testing for which PT is not available or required, they need to verify the accuracy of those procedures at least twice per year, as described at § 493.1236(c)(1). If available, voluntary PT may be a way the laboratory chooses to meet this requirement.

Comment: Commenters supported the removal of the types of services offered by laboratories in each microbiology subspecialty and replacement of the types of services with general categories of testing for which PT is required. However, they had questions about the proposed option in bacteriology for detection of growth or no growth in culture media. They questioned whether this option was included or relevant for all microbiology subspecialties and all specimen types and whether it should be removed as an option under the category for identification of bacteria since bacteria are not identified when only growth is detected. A commenter also noted that this category may not be appropriate for cultures from normally sterile sites or those that are expected to contain normal flora. Another commenter requested for clarification of how this category would apply to urine colony counts. A commenter suggested changing the language in bacteriology to “presence or absence of bacteria without identification,” with similar changes in other subspecialties. Another commenter suggested changing the language in bacteriology to “growth or no growth in culture media or identification of bacteria to the highest level that the laboratory reports results on patient specimens.” Other language changes suggested by commenters included revising this category to “growth or no growth of acid-fast bacilli” in mycobacteriology and “growth of yeast, growth of mold, or specimen negative for fungi” in mycology.

Response: We recognize the need for clarification of this option based on the comments received. The option was proposed in bacteriology at § 493.911(a)(1)(iv)(A); mycobacteriology at § 493.913(a)(1)(ii)(A); and mycology at § 493.915(a)(1)(ii)(A) under the proposed categories for microorganism detection and identification. Similar language proposed for parasitology at § 493.917(a)(1)(ii)(A) specified detection of the presence or absence of parasites. This option was not proposed for virology. Specimen types are not included in any of the PT categories in microbiology and a challenge for growth or no growth, or presence or absence, was not proposed and may not be appropriate for all specimen types or sites, or appropriate as a response for all laboratories. It is one of two options included under the category of detection and identification of bacteria, mycobacteria, fungi and aerobic actinomycetes, and parasites, in the respective microbiology subspecialties. It was proposed as an option for laboratories that perform limited microbiology testing to detect the presence of microorganisms and then refer growth from culture or specimens containing the microorganisms detected to another laboratory for identification. In response to the question about applicability of this option for laboratories that perform urine colony counts, PT is not required for colony counts. If the laboratory performs identification of the bacterial growth, PT is required for the identification. If the laboratory performs the colony count only and refers the isolate for identification, an appropriate result for the PT challenge would be to report detection or growth of bacteria. In response to the suggestions for revisions to the language for this option in each of the subspecialties, after considering the suggestions from commenters, for clarification in this final rule we have changed the language at § 493.911(a)(1)(iv)(A) to “detection of the presence or absence of bacteria without identification.” We changed the language at § 493.913(a)(1)(ii)(A) to “detection of the presence or absence of mycobacteria without identification, “and at § 493.915(a)(1)(ii)(A) to “detection of the presence or absence of fungi and aerobic actinomycetes without identification.” In parasitology, we added “without identification” to the end of the phrase currently at §§ 493.917(a)(1)(ii)(A) and 493.917(b)(1) to be consistent with the other microbiology subspecialties. In these subspecialties, we also revised the performance criteria at §§ 493.911(b)(1), 493.911(b)(7)(i), 913(b)(1), 493.913(b)(5)(i), 493.915(b)(5)(i), and 493.917(b)(5)(i) to correspond to these changes. For example, in bacteriology this change now specifies that the performance criterion is the correct detection of the presence or absence of bacteria without identification. This may be achieved when performing a culture and looking for bacterial growth or when using another test method that detects the presence of bacteria without any type of identification being performed.

Comment: Two commenters recommended clarification of the proposed categories of direct antigen and toxin detection, with specific questions about the applicability of this category and which antigens or toxins are required in the subspecialties of bacteriology (§ 493.911), mycobacteriology (§ 493.913), and mycology (§ 493.915). One commenter questioned whether the intent of the proposal was to require PT for only Clostridium difficile toxin or also for other toxins in bacteriology. The same commenter requested clarification on which direct antigen tests are proposed to be required in mycology. Another commenter questioned whether antigen Start Printed Page 41210 detection was intended to be required for mycobacteriology, as it was not proposed and no programs currently offer this PT.

Response: The requirement for PT for laboratories that perform direct antigen testing has been part of the CLIA regulations in the subspecialties of bacteriology and virology since PT was first required in 1994 and it was included as one of the required categories of microbiology PT in the proposed rule. As with other microbiology PT, the microorganisms for which it is required are not specified in the regulations. Rather, the regulations require that PT programs determine the reportable bacteria or viruses to be detected using direct antigen techniques. In this rule, required PT for direct antigen detection is included in bacteriology at § 493.911(a)(1)(ii); mycology at § 493.915(a)(1)(i); parasitology at § 493.917(a)(1)(i); and virology at § 493.919(a)(1)(i). Required PT for toxin detection is included in bacteriology at § 493.911(a)(1)(iii). As in the previous rule, the microorganisms for which direct antigen or toxin detection are required are not specified in the regulations. Rather, in all subspecialties for which this category is required, the regulations state the PT program determines the organisms to be reported by direct antigen or toxin detection. PT for direct antigen or toxin detection may be part of a combination module or offered as an individual five-challenge module in each subspecialty. If a laboratory performs direct antigen or toxin testing for which PT is not available, they are required to verify the accuracy of those procedures at least twice per year, as described at § 493.1236(c)(1).

Comment: A few commenters addressed the proposed requirements for microbiology stains, with agreement that Gram stain PT should require bacterial morphology as well as gram-reaction. Commenters requested for clarification regarding the level of detail required for bacterial morphology as part of PT and whether Gram stain PT would be required when a Gram stain is performed as part of organism identification. Commenters also questioned the proposed inclusion of Gram stains and acid-fast stains in bacteriology and mycobacteriology, but lack of requirements for stain challenges in other microbiology subspecialties.

Response: In this rule, we are finalizing the proposed requirement at § 493.911(b)(1) that includes bacterial morphology when performing Gram stain PT. This may apply to either a Gram stain required as an individual challenge or as part of bacterial identification. PT program instructions specify which tests are to be performed on each sample, thus identifying which samples require Gram stains. Morphology should include the basic shape and arrangement of bacteria. However, as stated at § 493.911(b)(1), the PT program determines the reportable staining and morphological characteristics to be interpreted by Gram stains. In response to the commenters who questioned whether PT was proposed for stains in mycology and virology, at this time, PT programs do not offer challenges for stains in these subspecialties. Thus, they were not proposed. In parasitology, although specific stains were not proposed as a required PT category, the sample types required at § 493.917(a)(2) include PVA (polyvinyl alcohol) fixed specimens and blood smears, both of which are used in parasite identification. Because a variety of stains are used by laboratories to facilitate identification of intestinal, blood, and tissue parasites, and in some cases, parasites can be identified directly in wet mounts without using a stain, no stains were included for this microbiology subspecialty. Each laboratory participating in PT for parasite identification should follow the staining procedures they use for patient specimens.

Comment: Commenters supported the removal of specific lists of microorganisms from the microbiology subspecialty requirements and replacement with general groups of organisms to be included over time. In addition, commenters requested clarification of the required groups in bacteriology, mycology, and virology. In bacteriology, one commenter suggested expansion of the groups to include Gram-negative cocci or coccobacilli, and another requested clarification of whether the groups of cocci include coccobacilli or diplococci. A third commenter suggested bacterial strains included in PT should be those routinely encountered in specimens. In mycology, two commenters expressed concern about inclusion of dimorphic fungi as a required category, noting that the majority require handling in a biosafety level 3 laboratory and are unable to be shipped. Comments pertaining to groups of organisms for virology recommended viral groups that must be included, and one organization questioned whether a PT program needed to offer all viruses and all specimen sources to be approved for virology PT. Specifically, the commenter questioned whether a program could offer PT challenges for susceptibility or resistance testing based on a single specimen source, such as urine. Another commenter requested for clarification regarding appropriate specimen sources to be included in virology modules and questioned whether combinations of viruses needed to be incorporated in a single PT sample.

Response: The PT requirements for the microbiology subspecialties specify that the organisms included are those that are commonly occurring in patient specimens or are important emerging pathogens. The groups identified for each of the five subspecialties are general groups to be included over time and annually, if appropriate for the sample sources. They are not intended to be the only groups that could potentially be included. In bacteriology, Gram-positive or Gram-negative coccobacilli or diplococci could be included as challenges in addition to, or as more specific subgroups of the individual morphologies listed for bacteriology at § 493.911(a)(3). No changes are being made in this final rule to the bacteriology groups that were proposed. As stated by the commenters for mycology, dimorphic fungi were proposed at § 493.915(a)(3)(ii)(C) as a group of organisms to be included in mycology over time and more specifically, required on an annual basis. We recognize the commenters concerns with the proposed inclusion of this group of fungi, some of which must be manipulated at a biosafety level 3. In response to these concerns, we have removed the dimorphic fungi from the groups of annually required organisms in mycology. However, over time, we encourage PT programs to include a variety of organisms in each subspecialty, as appropriate, to test a laboratory's ability to detect and identify the spectrum of organisms that might be found in patient specimens. In mycology, this may occasionally include dimorphic fungi, such as Sporothrix schenckii, that can be handled under biosafety level 2 conditions. In response to the questions about the PT requirements for virology at § 493.919(a)(3), the proposed rule did not specify that all viruses or specimen sources needed to be included for a PT program to be approved. However, it was proposed that if appropriate for sample sources offered, the types of viruses included annually must be representative of the groups of medically important viruses listed. Generally, with this rule, PT programs must continue to offer the same types of virology challenges and modules that have been offered in the past. Lastly, PT Start Printed Page 41211 samples containing combinations of viruses were not proposed and are not required in this final rule.

Comment: Several commenters indicated that the proposed requirement in all microbiology subspecialties for laboratories to detect and identify organisms to highest level performed on patient specimens was unclear. One commenter recommended changing the description of the category for identification of bacteria to “the highest level that the laboratory reports results on patient specimens.” Two commenters suggested identification needed to be clarified as to whether the intent was presumptive or definitive identification and others questioned how this requirement should be applied with respect to identification at the genus or species level. The commenters stated more specific and better-defined criteria are needed, as well as the incorporation of language to allow for abbreviated reporting frequently used in reporting mixed cultures. They also questioned whether this information would need to be transmitted from PT programs to CMS and State agencies and one noted it would take time to implement this requirement. Another commenter stated it is the responsibility of inspectors to review patient reporting practices and not that of PT programs.

Response: We agree that the language proposed in all subspecialties for identification of microorganisms to the highest level that it performs procedures on patient specimens may be unclear, and we agree that the revised description provided by the commenter earlier more clearly specifies that this requirement refers to how a laboratory reports results on patient specimens. As a result, we have incorporated the change suggested by the commenter and made conforming changes in this rule for all subspecialties at §§ 493.911(b)(2), 493.913(b)(2), 493.915(b)(2), 493.917(b)(2), and 493.919(b)(2). We expect that this will clarify that if a laboratory reports patient results to the genus level, that is the expectation for PT. Similarly, if a laboratory reports patient results to the species level, that would be the expectation for reporting patient results. In response to the question about incorporation of language to allow for reporting abbreviated results, if this is the practice for reporting results to the highest level on patient specimens, it may be an acceptable PT practice as well. In all subspecialties, PT programs determine the organisms that must be reported as part of their identification. We believe the delayed implementation of specific portions of this final rule will allow PT programs to incorporate updates needed for reporting results to CMS. We agree with the commenters who stated that it is the responsibility of laboratory inspectors to review patient reporting practices and not the responsibility of PT programs and this was part of a CLIAC recommendation made prior to the development of the proposed PT rule. It was not our intent that PT programs take on this responsibility and it was not included in the proposed rule.

Comment: Multiple commenters supported the proposed changes to decrease the required percentage of mixed culture challenges from at least 50 percent to at least 25 percent in bacteriology, mycobacteriology, and mycology. The change, if finalized, would specify that at least 25 percent of the PT samples must contain mixtures of the principal organisms and appropriate normal flora.

Response: We agree with the commenters and appreciate their support of these proposed changes. This is in alignment with a CLIAC recommendation stating such and was proposed at §§ 493.911(b)(1), 493.913(b)(1), 493.915(b)(1). We are finalizing these changes in this rule.

Comment: Some commenters recommended changes to the microbiology subspecialties for which susceptibility or resistance testing PT was proposed to be required. A commenter noted that it would be difficult to comply with the requirement for susceptibility or resistance testing in mycology since samples are limited, there are few FDA-cleared methods or breakpoints for fungi, and there is extensive variability in the testing. Another commenter recommended that susceptibility or resistance testing may not be added to required PT in mycology and may be removed in mycobacteriology since few laboratories perform this testing. A third commenter stated the value of requiring PT for M. tuberculosis susceptibility testing is limited since programs often send out the same strain that is susceptible to all drugs tested. With respect to virology, a commenter disagreed with requiring susceptibility or resistance testing in this subspecialty and proposed requiring PT for viral loads. Another commenter indicated that since only one PT program currently offers antiviral susceptibility testing, that does not meet the specified criterion of requiring that three programs offer PT for an analyte or test, and it may not be required in virology. Finally, a commenter questioned whether a PT program should be required to offer susceptibility or resistance testing PT in virology if they offered other virology PT.

Response: We agree with the commenters' reasons for suggesting that PT not be required for susceptibility or resistance testing in mycology and virology at this time. Therefore, we are removing the proposed requirements for inclusion of this category of required PT § 493.915(a)(1)(iii) for mycology and at § 493.919(a)(1)(iii) for virology in this final rule. If this testing becomes less variable and PT availability increases in these subspecialties in the future, we may propose to include it in rulemaking at that time. In the meantime, if a laboratory performs susceptibility or resistance testing on patient specimens in mycology or virology, they are required to verify the accuracy of those procedures at least twice per year, as described at § 493.1236(c)(1). Voluntary PT may be a way the laboratory chooses to meet this requirement. With respect to the requirement for susceptibility or resistance testing in mycobacteriology, we are aware that small numbers of laboratories perform this testing and subscribe to PT and that only one program currently offers susceptibility testing PT in mycobacteriology. We also recognize that PT programs are less likely to send out resistant strains of mycobacteria, especially M. tuberculosis, due to biosafety concerns when shipping or working with these organisms. For these reasons, in addition to the fact that mycobacteriology is unique in that only two PT events per year are required, we are removing the requirement at § 493.913(a)(1)(iii) for susceptibility or resistance testing in mycobacteriology in this final rule. As stated previously, if a laboratory performs susceptibility or resistance testing on patient specimens in mycobacteriology, they are required to verify the accuracy of those procedures at least twice per year, the same frequency as required PT in this subspecialty. Laboratories may choose to subscribe to voluntary PT as a way to meet the requirement or they may use another mechanism to meet the requirement that does not include shipping strains of organisms that require special precautions.

Comment: Commenters questioned or requested clarification of the proposed requirements specified for antimicrobial susceptibility or resistance testing, including clarification of the definition or intent of resistance testing, questioning whether it meant testing for resistance mechanisms or markers for specific organisms. One commenter stated clarification was needed as to whether susceptibility testing is optional if a laboratory performs Start Printed Page 41212 identification. Another commenter suggested the language for this category of PT in bacteriology be clarified to state “antimicrobial susceptibility or resistance testing of select bacteria.”

Response: The category of antimicrobial susceptibility or resistance testing was included in the proposed rule in for the subspecialties of bacteriology at § 493.911(a)(1)(v); mycobacteriology at § 493.913(a)(1)(iii); mycology at § 493.915(a)(1)(iii); and virology at § 493.919(a)(1)(iii). Resistance testing was included in this proposed category as it was previously recommended by CLIAC to be required along with susceptibility testing. As discussed in the previous comment, the proposed requirement for susceptibility or resistance testing in mycobacteriology, mycology, and virology has been removed from this final rule. With respect to the proposed requirement for this category in bacteriology, we agree with the commenters that the interpretation of “resistance testing” may not be clear, and that in some cases, bacterial resistance may be determined as part of an organism identification. For these reasons, we have removed resistance testing from the required category proposed in bacteriology and in this final rule we are requiring antimicrobial susceptibility testing of select bacteria, as suggested by the commenter, at § 493.911(a)(1)(v), since antimicrobial susceptibility testing is not performed on every bacterium that is isolated in a culture and PT programs specify which challenges require that susceptibility testing be performed. This also addresses the comment suggesting a change in the description of this bacteriology category for clarification. If laboratories perform resistance testing separate from bacterial identification, they are required to verify the accuracy of those procedures at least twice per year, as previously stated, and may enroll in voluntary PT to do so. In response to the recommended clarification of whether susceptibility testing is optional when a laboratory performs identification, laboratories must follow PT program instructions when determining which tests to perform on a microbiology sample. The programs must clearly identify which samples require that susceptibility testing be performed on bacteria that are identified and those results reported for PT purposes.

Comment: Several commenters agreed with the proposed increase in the number of required susceptibility or resistance testing challenges from one to two per event in all microbiology subspecialties except parasitology, where PT for susceptibility testing is not required. They indicated that increasing the number of challenges and requiring one Gram-positive and one Gram-negative challenge per event in bacteriology would help identify issues with patient testing. Other commenters disagreed with this proposed change, expressing concerns that this requirement would provide too much information to laboratories about PT sample content and make the PT results more predictable. One commenter stated that including two susceptibility challenges per event lacked value and relevance. Others suggested that requiring a mixture of challenges throughout the year was preferred over the requirement to include one Gram-positive and one Gram-negative challenge per event.

Response: We agree with the commenters who supported the proposed change to increase the number of required susceptibility or resistance challenges to two per event and are finalizing that change in this rule at § 493.911(a)(4). This change was recommended by CLIAC, and we believe it will provide a better assessment of laboratory testing performance over time. We also agree with the commenters who suggested that we should not specify a predictable pattern of susceptibility testing challenges in bacteriology, requiring that each event must include one Gram-positive and one Gram-negative challenge. As a result, in this rule, we are revising the requirement to indicate that each year, a minimum of two samples per testing event of susceptibility testing challenges must include a mixture of Gram-positive and Gram-negative challenges.

Comment: A PT program commented on the proposed requirements to change scoring for the microbiology subspecialties by including separate category scores in addition to the overall subspecialty scores. The program inquired about the intent of this proposed change and suggested that it would increase the complexity of determining scores and it may be especially challenging to score laboratories that perform a mixture of detection and identification procedures. The commenter also noted the proposed scoring method would give PT programs discretion in the interpretation of the requirement which could result in laboratories choosing the program that uses the most advantageous method. The commenter advocated for simplifying the subspecialty scoring process rather than increasing complexity for efficiency and increasing the value to laboratories.

Response: The four categories of testing proposed for microbiology PT were recommended by CLIAC to replace the types of laboratory services that are part of the current regulations. The types of services guided the scoring of microbiology subspecialties since there are no specific analytes in this laboratory specialty. However, since only a single score is given for each subspecialty, many times representing a combination of results for different types of testing, it is not possible for laboratory surveyors to readily determine if a laboratory is having problems with one area of their microbiology testing. No changes were made to the scoring process for microbiology in the proposed rule other than aligning the requirements for evaluation of a laboratory's performance at §§ 493.911(b) through 493.919(b) to be consistent with the categories of testing and facilitate the identification of problems in any one of the categories.

Summary of Final Actions

• We are finalizing the proposed revisions at §§ 493.911 through 493.919 by removing the types of services listed for each microbiology subspecialty and inserting a more general list of organisms.
• We are finalizing the proposed revisions at §§ 493. 911(a), 493.913(a), and 493.915(a) that are related to growth or no growth and mixed culture requirements (50 percent to 25 percent).
• We are finalizing the proposed performance criteria revisions at §§ 493.911(b), 493.913(b), 493.915(b), 493.917(b), and 493.919(b).
• We are finalizing the proposed addition of “without identification” to the end of the phrase currently in the subspecialty of parasitology at § 493.917(a)(1)(ii)(A) to be consistent with the other subspecialties.
• We are finalizing the proposed revised requirement at §§ 493.911(b)(2), 493.913(b)(2), 493.915(b)(2), 493.917(b)(2), and 493.919(b)(2) to clarify and emphasize that laboratories should detect and identify organisms to the highest level that they report results on patient specimens.
• We will amend §§ 493.911(b)(1), 493.913(b)(1), 493.915(b)(1), 493.917(b)(1), 493.919(b)(1) to clarify that for the purpose of achieving consensus, PT programs must attempt to grade using both participant and referee laboratories before determining that the sample is ungradable.

• We are finalizing the proposed revisions to § 493.911(a) through (b) related to Gram stains, direct antigen detection, bacterial toxin detection, and performance and scoring related to Start Printed Page 41213 direct antigen and bacterial toxin detection for the subspecialty of bacteriology.

• We are finalizing the proposed addition to § 493.915(a) related to requiring direct antigen testing for the subspecialty of mycology.
• We are finalizing the proposed addition to § 493.917(a) related to requiring direct antigen testing for the subspecialty of parasitology.
• We are finalizing the proposed revision to § 493.919(a) related to requiring direct antigen testing for the subspecialty of virology.
• We are removing the reference to resistance testing in the subspecialty of bacteriology and have removed references to “resistance testing” in the requirement for antimicrobial susceptibility testing of select bacteria at § 493.911.
• We are not finalizing the proposed requirements for PT of antimicrobial susceptibility and resistance testing in the subspecialties of mycobacteriology, mycology, and virology and have removed the requirement at §§ 493.913, 493.915, and 493.919.

F. Proposed Changes to PT for Non-Microbiology Specialties and Subspecialties (§§ 493.921 Through 493.959)

1. Required Analytes

Comment: Several commenters agreed that the list of required analytes should be updated. Some commenters stated that the process for analyte inclusion and removal was thorough, understandable, and transparent. One commenter stated the inclusion threshold for new analytes that only included three PT programs, rather than four, could result in an unfair market advantage, raise PT costs for laboratories, or result in logistical difficulties in obtaining PT.

Response: In response to the comments, we reviewed our analyses and determined that there were no proposed analytes that would not have made the requirement for being offered by at least four PT programs, as was suggested by the commenter. We believe that the fact that there are already at least three programs available to choose from for each new analyte or test gives laboratories several options and should not result in increased costs or logistical difficulties in obtaining PT. All PT programs received notification of the proposed analytes or tests at the same time when the proposed rule was published. Whether a PT program elects to offer a particular analyte is a business decision of the PT program, and outside of our purview.

Comment: A small number of commenters mentioned concerns about the possibility that either inclusion of the PT analytes or the ALs we proposed would have a negative impact on access to testing. A few commenters suggested that for the ALs proposed for some analytes, some existing test systems would not meet the new requirements. For example, one manufacturer stated that the proposed ALs for creatine kinase isoenzymes may be challenging for some testing platforms to meet. A similar comment was made for proposed ALs for troponin I and hematocrit.

Response: During the phase in period, manufacturers will have time to improve test accuracy, and laboratories will have time to switch to higher accuracy test methods if those they use do not provide results that are able to meet the criteria for acceptable performance specified in the regulations. Clinicians and patients should be able to expect accurate testing, and assuring overall accuracy is the goal of performing PT. Therefore, these changes should drive the health care system toward more accurate methods. We have no reason to believe that access to testing will be impacted.

Comment: Several commenters supported the list of analytes that were proposed for addition and deletion, and commenters supported the process we used for determining the list of analytes for which PT is to be required. No commenters questioned any of the proposed new analytes. However, one commenter stated that a current analyte, T3 uptake, should be deleted because it lacked clinical utility. An accreditation organization and an individual commented that determination of creatine kinase (CK) MB fraction by electrophoresis should be discouraged, and therefore, it should be excluded from the required PT for creatine kinase isoenzymes. Rather, the commenters noted that PT should only be required for laboratories that use immunochemical methods when testing for this analyte. Some commenters recommended inclusion of analytes that we had considered but decided not to include. One commenter suggested that we require PT for several immunosuppressant drugs for which PT is not currently required.

Response: We had initially considered all the analytes that commenters recommended for either inclusion or deletion, but the suggested analytes did not meet one or more of our inclusion or deletion criteria. Both the inclusion and deletion processes, which were described in the proposed rule, were based upon per-analyte estimates of the availability and the number of programs already offering PT, the nationwide volume of patient testing, the impact on patient or public health of offering PT, and the cost and feasibility of PT implementation. We did not propose deletion of T3 uptake because test volumes were above the threshold for consideration. With respect to the suggestion to discourage laboratories from using electrophoretic methods to test for CK-MB isoenzymes, the method used is not a basis for requiring or not requiring PT for any test or analyte. Each laboratory needs to identify the method or test system used when submitting PT results for programs to properly grade the PT. To the extent that test results are used for clinical decision making, the test results should be accurate. The immunosuppressant drugs that were suggested were not done in sufficient volumes to meet the threshold for consideration in the proposed rule, so they were not proposed to be required.

Comment: For a few analytes that can be detected or quantified in more than one way, some commenters requested clarification concerning which analyte would require PT. For example, a commenter questioned if PT was proposed to be required whether LDL cholesterol was calculated or measured directly. Several commenters requested clarification concerning whether drugs were to be measured in total or free forms. One commenter mentioned a need to specify the sample type that should be tested if the analyte can be tested in more than one type of body fluid.

Response: For LDL cholesterol, which can be measured both directly and as an estimation based on other measured lipids, PT is only required for directly measured (not calculated) LDL cholesterol. For all drugs, we intend that the measured form must be total drug. For the specialty of chemistry, in subpart I the sample types for which PT is required are specified for each under each subspecialty, at § 493.931(b) for general chemistry, § 493.933(b) for endocrinology, and § 493.937(b) for toxicology. If a laboratory performs patient testing on other sample types than those listed, they are required to verify the accuracy of testing with those alternative sample types at least twice per year, as described at § 493.1236(c)(1). If available, voluntary PT may be a way the laboratory chooses to meet this requirement.

Comment: A few commenters requested clarification of what should be considered high sensitivity C-reactive protein, as opposed to traditional C-reactive protein, as included in the Start Printed Page 41214 proposed rule. A related comment suggested that we should require PT for all assays for C-reactive protein.

Response: Although traditional C-reactive protein has been used as a general marker of inflammation for many years, it did not meet the threshold for inclusion as a required PT analyte. In this rule we are finalizing the proposed PT requirement for high sensitivity C-reactive protein and we appreciate the need to define which test methods would be considered “high sensitivity” testing. High sensitivity C-reactive protein concerns testing related to cardiac ischemia, either for frank cardiac events or for risk stratification, which requires more sensitive test methods to detect lower concentrations. We are deferring to laboratories to know whether their assay is a high sensitivity method used to detect cardiac pathology, or the traditional, less sensitive C-reactive protein. PT programs must label their PT offerings accordingly.

Comment: One commenter suggested that we should specify the N-terminal region of pro-B-natriuretic peptide (BNP), which was included as a required analyte in the proposed rule because this is the epitope usually detected by antibodies used in most test methods.

Response: We agree with the commenter that the N-terminal region of pro-B-natriuretic peptide (BNP) is the part of the peptide that is usually measured, but we did not want to restrict the requirement for PT. Therefore, in this rule we are finalizing the name as proposed: proBNP.

2. Scoring and Acceptance Limits

Comment: With respect to scoring and ungradable samples, one commenter requested clarification about how performance on an analyte was determined for a PT event when one of the PT samples was not able to be graded. The commenter questioned what the denominator of graded samples would be. An accreditation organization agreed with our proposal to require PT programs to attempt to reach consensus using both laboratory and referee laboratories before deciding a sample is ungradable due to lack of consensus.

Response: If a sample for a particular PT event is ungradable, for example, because consensus could not be reached, it is still considered to be part of the denominator of five PT samples for that event, and in this case, the laboratory is given credit for passing the challenge. Therefore, if one of the remaining PT samples in the event is missed, the event score is 80 percent, and the event score is “satisfactory” for the majority of required PT.

Comment: Several commenters stated that the process used for simulating the impact of scoring PT using several alternative ALs to determine the optimal limit to require was unclear.

Response: As discussed in the proposed rule, we requested PT programs to examine the impact of various ALs on their aggregated sample failure rates, using the peer grouping approaches they had previously used. A number of the PT programs provided simulated results, applying various possible percentage-based ALs to actual results from previous PT events, and were able to help us select appropriate ALs. We selected ALs using a target miss rate (per sample) in the 1 to 2 percent range. Our intent was to assure that the ALs would work across the clinically important range and not inappropriately fail results that were accurate for clinical decision making. Therefore, we examined error rates at all concentrations that PT programs used throughout the 2 years of PT data they shared with us.

Comment: We received a number of comments related to the proposal to use percentage-based ALs whenever possible. While some commenters supported the proposed changes, others suggested changes to specific proposed ALs for both current and newly proposed analytes. Generally, these comments concerned whether the proposed limits would be workable across the clinically important measurement interval for all test methods and platforms. In almost all cases, the comments recommended less stringent ALs, either across the entire analytical measurement range or specifically at low concentrations, where test methods are generally less accurate. Commenters pointed out that unless there is allowance for low concentrations, PT programs would be discouraged from using PT samples with low concentrations, to the detriment of assuring accurate testing across the analytical range. Supporting this, some commenters stated that it is not clinically important to be as accurate as the percentage-based limits would require. Commenters suggested that we use a combination of a percentage and a concentration limit for certain analytes, such that PT samples with relatively low concentrations would be more fairly assessed. In some cases, commenters recommended a concentration limit that differed from a concentration limit we had proposed. A small number of commenters were generally concerned about moving from familiar 3 SD-limits to percentage based ALs for some currently required analytes.

Response: In response to commenters' concerns about the use of percentage limits when scoring PT analytes at low concentrations, in this final rule, we are including “concentration limits” such as are already used for glucose and some other analytes for many newly required analytes and some previously required analytes. When adding concentration limits and using combined ALs, programs are directed to score with whichever of the specifications is more tolerant, allowing for fairer and more realistic ALs that will allow PT programs to cover the clinically important range of results. We re-examined previously acquired simulation data from PT programs and have added concentration limits for 13 analytes. Specifically, we created concentration thresholds for alanine aminotransferase, aspartate aminotransferase, cholesterol (high density lipoprotein), CK-MB isoenzymes, glucose, carcinoembryonic antigen, human chorionic gonadotropin, vitamin B12, acetaminophen, carbamazepine, lithium, phenobarbital, and salicylate. Concerning the switch from current 3 SD limits to percentage-based limits, we believe that the new ALs will be workable, fair, and clinically relevant. As stated in the proposed rule, ALs based on analytical variability within a peer group, such as the use of 3 SD limits, are ill-suited to know whether testing results are sufficiently accurate for clinical purposes.

Comment: Two commenters noted that CLIA ALs have been used in ways other than their intended purpose of identifying laboratories with unacceptable performance. One commenter noted that ALs have been used as goals for ideal performance, for example, setting quality control acceptable limits. Another commenter pointed out that ALs have been used for verifying analytical performance, for example, accuracy.

Response: We agree with the comments and reemphasize that ALs must not be used as the criteria to establish performance goals in clinical laboratories. Goals for accuracy and precision must be based upon clinical needs and manufacturer's FDA-approved or -cleared labeling; PT performance is not the best assessment of these. Proficiency testing is intended to identify laboratories that are not performing with acceptable analytic accuracy; it is not intended, nor suited, to provide goals for analytical accuracy or clinical performance. Start Printed Page 41215

Comment: Many commenters stated the proposed AL for hemoglobin A1c (HbA1c) was too loose and not reflective of the testing accuracy of current test methods. Many individuals and organizations commented that the AL should be 6 percent, and several recommended lowering the limit to 5 percent. Several comments requested that CLIA ALs should not “change” from the current 6 percent, despite the fact that HbA1c is currently not a CLIA-required PT analyte, and therefore, no ALs are specified in the regulations. Many commenters expressed concerns that using a threshold higher than 6 percent would in some way subvert the substantial progress made by the National Glycohemoglobin Standardization Program (NGSP), working collaboratively with test method manufacturers, to improve accuracy of HbA1c testing. Commenters suggested that manufacturers would allow the accuracy of their test methods to deteriorate if CLIA added HbA1c with an AL as loose as 10 percent. A PT program proposed that we use an AL of 10 percent for non-commutable PT materials and a limit of 6 percent for commutable (accuracy-based) PT materials. Another PT program commented in favor of a 10 percent limit, noting that non-commutable PT materials may be less accurate with certain test methods and, moreover, PT is not intended to directly reflect accuracy needed for clinical testing.

Response: We appreciate the importance of HbA1c for diagnosis and monitoring patient management, and the need for testing accuracy that is sufficient to meet clinical needs, and we support the progress that continues to be made to improve the accuracy of HbA1c testing. As mentioned in the previous comment, CLIA PT ALs are intended to identify, and hopefully remediate, laboratories that are not providing results as accurate as their peers. CLIA PT ALs should not be used as accuracy goals by manufacturers or by standardization initiatives such as the NGSP. CLIA should not impose a requirement that limits access to critically important patient testing, especially if it is based on PT results that may not reflect the accuracy of patient testing.

One PT program has demonstrated progressive improvements in accuracy of testing by laboratories enrolled in their accuracy-based PT program, which uses commutable patient samples. We are aware that, over time, the program has incrementally tightened their ALs for the accuracy-based PT. This progress has been possible without CLIA requiring PT for HbA1c, and therefore, adding a PT requirement for HbA1c should not impede further progress in the future. Accreditation organizations have the flexibility to require their laboratories to meet a more stringent requirement than CLIA. They also have the option of using the CLIA limit and using a second, more stringent, AL for educational purposes. Either approach would allow these organizations to continue to tighten the limits for HbA1c for their accredited laboratories. We acknowledge the importance of standardization programs, like the NGSP, having the latitude to continuously adjust their accuracy goals to monitor and encourage improvements in the accuracy of HbA1c testing. We do not believe that a CLIA AL that is looser than the limit in use by the accuracy-based PT program would cause manufacturers to allow testing accuracy to deteriorate, as many commenters have suggested.

The AL adopted in CLIA regulations must not be too tight for laboratories that do not participate in an accuracy-based PT program that uses commutable PT materials. In simulation studies performed before issuing the proposed rule, laboratories using non-commutable PT samples had poorer performance, especially when scoring using any AL less than 10 percent. This might have occurred because laboratories not enrolled in accuracy-based PT use different test methods or because the PT they use is non-commutable. CLIA does not specify whether laboratories are required to participate in PT based on whether it is commutable or non-commutable. The same AL apply regardless of the PT samples' commutability.

After analyzing the comments received in response to the proposed rule, we requested the PT programs that offer HbA1c to simulate results that would be obtained if they used 5 percent, 6 percent, 8 percent, and 10 percent as the AL. We requested programs to indicate miss rates and unsatisfactory rates based upon different HbA1c concentrations in their materials, and to disclose performance based upon their testing platform or peer groups used. Based upon these more recent simulated results, we found that it will be possible to use a tighter AL than 10 percent. After this analysis, we are setting the AL for HbA1c at 8 percent in this final rule. The performance improvements we saw between the first and later simulations may reflect improvements in the accuracy of testing for HbA1c.

Comment: Commenters stated that rather than using the proposed AL of 20 percent for LDL cholesterol, we should require an AL of 12 percent, which is the accuracy target used by the National Cholesterol Education Program.

Response: Because the commenters suggested an AL tighter than was proposed, we requested PT programs to simulate the impact of using that limit. Based upon reanalysis of new data shared by PT programs, we confirmed that the proposed AL of 20 percent is appropriate for scoring PT for LDL cholesterol, and we are finalizing that limit in this rule.

Comment: With respect to PT for blood lead, we proposed a change from the current AL of ±4 mcg/dL or 10 percent (greater) to ±2 mcg/dL or 10 percent (greater). One commenter supported the proposed AL, consistent with efforts to improve the ability of laboratories to detect very low concentrations of blood lead in patient specimens. Conversely, another commenter stated that the reduction of the concentration AL from 4 mcg/dL to 2 mcg/dL would result in more instances of nonconsensus, which would result in more ungraded samples and events. Another commenter expressed concerns about the impact of the proposed limits on failures for certain testing platforms.

Response: We agree with the commenter who emphasized the public health importance of the need for accuracy at low concentrations of blood lead, to detect and prevent cases of childhood lead poisoning, and are finalizing the proposed AL for blood lead at 2 mcg/dL ±10 percent (greater) in this rule. We appreciate the commenters' concerns, however, one outcome of more stringent ALs may be that laboratories switch to test methods that are more accurate across the range of testing and better able to meet clinical needs. We believe that manufacturers of analytical platforms that may fail to achieve consensus, or otherwise perform poorly, will improve their accuracy during the phase-in period. To address concerns regarding unintended consequences that may increase health disparities, we will monitor changes in PT participation for all analytes after this rule becomes effective as this is required as part of PT oversight under CLIA. This includes the methods used for testing each PT analyte required by CLIA.

Comment: A few commenters provided suggestions related to the addition of troponin I and troponin T as required analytes in routine chemistry. One commenter was concerned that adding troponins to the required list for PT may potentially limit access to point-of-care cardiac triage testing of potential cardiac events in rural settings. The Start Printed Page 41216 same commenter also suggested that the ALs for troponin I and troponin T should be expanded to ±40 percent, with no suggested changes to the associated concentration limits. A couple of commenters suggested that the same, percentage-based AL would work for both generic and high sensitivity troponins. A small number of commenters suggested that we should require PT for high sensitivity troponin assays in addition to traditional troponin assays.

Response: Troponin I and troponin T are used to make decisions about the use of lifesaving, yet not risk-free, interventions, such as cardiac catheterization and therapeutic thrombolysis. Therefore, it is important that such testing be both accessible and accurate. We believe that requiring PT for the troponins is important and must not inhibit access to testing. We reviewed our simulation data to see if the same concentration limit would work for both troponin I and T. We determined that we must use the proposed, different ALs, and, therefore, are finalizing the AL for troponin I as ±0.9 ng/mL or 30 percent (greater) and for troponin T as ±0.2 ng/mL or 30 percent (greater). At the time we proposed these changes, troponin I and T were not frequently tested as “high sensitivity” analytes, that is, at very low limits of detection. Also, there were not enough PT program offerings to meet our threshold for inclusion for high sensitivity troponins. Therefore, we are not requiring PT for “high sensitivity” troponin I or T.

Comment: A few commenters stated that some proposed percentage-based ALs were too tight, regardless of whether a concentration threshold was included. Commenters stated that the proposed percentage ALs for immunoglobulin A (±15 percent), immunoglobulin E (±15 percent), amylase (±15 percent), and leukocyte count (±5 percent) were too tight. The commenters recommended ALs be set at ±20 percent for immunoglobulin A, ±25 percent for immunoglobulin E, and ±10 percent for leukocyte count. No recommendation was provided for amylase.

Response: We re-examined simulation data that had been submitted by PT programs and revised percentage limits as appropriate. Specifically, in this rule we are finalizing the AL for immunoglobulin A to ±20 percent, amylase to ±20 percent, and leukocyte count to ±10 percent. We determined that adding a concentration limit for these analytes was not necessary or adequate to make the AL workable at a lower concentration. For immunoglobulin E, we did not determine that it was necessary to increase the AL to ±25 percent; therefore, we are finalizing the AL for immunoglobulin E in this rule at ±20 percent.

Comment: Some commenters expressed concerns related to proposing ALs based on allowable total error derived from estimates of biological variability (BV). There was a comment that the use of BV data was in flux at this time. One commenter noted that estimates of BV that we used may be incorrectly wide due to errors in the way estimates were made, specifically that they may overestimate BV because the results are based upon analytical test methods that have inherent variability. One commenter stated that BV cannot be directly related to clinical outcomes. The same commenter stated that when setting ALs both BV and state-of-the art performance should be considered.

Response: We appreciate the concerns expressed and note that the ALs we proposed were not based strictly on estimates of BV. Moreover, we are aware that the field of estimating BV data has changed in the last few years. However, any impact of suboptimal estimations of BV on the ALs we proposed was likely negligible because we always tested potential ALs using simulations. ALs that were too tight to be workable were eliminated even if they were not as stringent as our estimates of BV might have suggested were necessary. In other words, consistent with one of the comments, we used state-of-the-art performance, demonstrated through simulations, to finalize the proposed ALs. In some cases, we showed through simulations that it was possible to use ALs that are tighter than the “minimal” threshold based upon estimates of BV and in these cases we used a somewhat tighter AL, but only if the data from PT programs supported the tighter limit. As a result, changes in the estimates of BV we used would not have affected our proposed ALs.

After re-examining the literature, we reconfirmed that BV is the only tenable approach to establishing new limits. We agree that clinical outcomes may not be reflected in BV data, but the preferred outcomes studies were not available to us.

Comment: Commenters generally favored the proposal to require separate PT for cell identification and differentials rather than including an option to participate in PT for one or the other. It was pointed out that the results can be used for different purposes in patient treatment. There were questions, however, questioning whether there should be separate scores for cell identification and differentials or if they should be averaged. One commenter recommended that the three standard deviation criteria for acceptable performance for differentials should be changed to a percentage-based criterion and another suggestion was made to include ±1.0 (whichever is greater) for low target values or absolute values (that is, basophils). An additional commenter requested clarification as to whether PT would be required for both manual and automated flow through differentials for laboratories that use platforms that can report flow through differentials.

Response: We appreciate the support from commenters who recognized the need to recognize cell identification and differentials as two separate analytes and are finalizing that change in this rule. As separate analytes they may be scored individually. We are finalizing the criteria for acceptable performance for both analytes in this rule. We are not changing the criterion for differentials to percentage-based because we have no BV data on which to base that change. As such, we are also not including the ±1.0 option for low target values. In response to the question regarding PT requirements for laboratories that perform both manual and automated flow through differentials, a laboratory should perform PT in the same manner as they perform testing on patient specimens. PT is required for the primary method of testing used for patient testing.

Comment: A few commenters supported the proposal to change the consensus requirement for cell identification from 90 percent to 80 percent. One commenter requested for clearer justification for the change.

Response: This change was proposed because it is not possible to score 90 percent on a 5-challenge PT panel. We are finalizing the change in this rule.

Comment: An accreditation organization made several suggestions about how standard deviations should be calculated when they are required as ALs for white blood cell differentials. For peer group sizes of 20 or more, they recommended that we continue to require elimination of outliers before calculation of the standard deviation. The commenter stated that when the peer group size is between 5 and 19 laboratories, robust methods as described in ISO 13528, ISO Guide 35, or ASTM E-691, should be used. They recommended that, alternatively, the standard deviation could be an average standard deviation determined from previous rounds of PT, calculated according to ISO 13528. They also noted that mention of 3 SD to set ALs should Start Printed Page 41217 be removed from parts of the regulation that no longer include 3 SD limits.

Response: As mentioned by the commenter, this final rule includes only one analyte with a three standard deviation limit. We agree that this recommendation would allow more accurate estimates of 3 SD ALs for relatively small peer group sizes. We also agree that robust statistical methods must be used to calculate the standard deviations when the peer group size is between 5 and 19 laboratories. However, we are not specifying the statistical approach that needs to be used. We appreciate the commenter's suggestion to remove reference to 3 SD ALs in relevant sections of this final rule and have done so in §§ 493.931(c)(2) and 493.933(c)(2).

Comment: A few commenters recommended that the international normalized ratio (INR) should be listed as a separate analyte in the specialty of hematology, the same way blood cell counts and white blood cell differentials are separate analytes, rather than including INR as a mechanism for reporting prothrombin time results, as was proposed. The commenters agreed that laboratories should report prothrombin time results in seconds, as an INR, or both as appropriate, in the same way that they report patient results. Commenters also stated that separating the prothrombin time and INR would allow for separate ALs for each of them.

Response: It is important for laboratories to report PT results the same way that they report patient results. If patient results are reported in seconds or as INR results, laboratories should report the same way to PT programs. If the laboratory reports patient results in both seconds and as an INR, they should report both to PT programs. The AL for prothrombin time at ±15 percent is applicable for both seconds and INR. When we referenced “directly measured INR” in the preamble to the proposed rule, we were referring to those devices that internally calculate and display the INR value rather than giving a value in seconds. The 15 percent AL for INR applies regardless of how it is derived.

Comment: Two commenters remarked on the proposed change to the criteria for acceptable performance of unexpected antibody detection in immunohematology from 80 percent to 100 percent accuracy. While one commenter agreed with this proposed change, the other disagreed. The opposition was concerned with the possibility that laboratories that use less sensitive, but safe, methods could be penalized, and it could limit patient access to care.

Response: We believe that the criteria for acceptable performance for unexpected antibodies should be 100 percent rather than 80 percent. We are finalizing this change because it is critical for laboratories to detect any unexpected antibody when crossmatching blood to protect the public health and not impact patient care. It is important that antibodies are detected to lessen the possibly of a transfusion reaction due to incompatible blood products.

Comment: Concerning appropriate units for reporting PT results or some other aspect of the AL, some commenters noted that we inadvertently deleted titers for some ALs. It was pointed out that for some analytes we incorrectly suggested that the AL should be qualitative. Some commenters noted inaccuracies in the units we used for quantitative analytes.

Response: We appreciate the commenters careful examination of the proposed limits and we made appropriate adjustments that are now reflected in the final rule. In response to comments about proposed units for reporting PT results, unintentional uses of incorrect units have been corrected in this final rule.

Summary of Final Actions

• We are finalizing the proposed revision at §§ 493.923(a), 493.927(a), 493.931(a), 493.933(a), 493.937(a) and 493.941(a) to remove the option that PT samples “at HHS option, may be provided to HHS or its designee for on-site testing.”
• We are finalizing the proposed addition of 29 analytes and the deletion of five analytes. See section II of this final rule. Additional analytes can be found in section II.B.1. of this final rule, Table 1, and deleted analytes are listed in section II.B.6 of this final rule.
• We are amending §§ 493.923(b)(1), 493.927(c)(1), 493.931(c)(1), 493.933(c)(1), 493.937(c)(1), 493.941(c)(1), and 493.959(d)(1) to clarify that for the purpose of achieving consensus, PT programs must attempt to grade using both participant and referee laboratories before determining that the sample is ungradable.

• Section 493.927 (General Immunology)

++ We are correcting typographical or editorial errors in the proposed criteria for acceptable performance for alpha-1-antitrypsin, alpha-fetoprotein (tumor marker), complement C3, complement C4, antinuclear antibody, antistreptolysin O.

++ We are modifying the proposed AL for immunoglobulin A (IgA) of ±15 percent and finalizing the AL for IgA as ±20 percent based on public comments.

++ We are finalizing the proposed criteria for acceptable performance for antinuclear antibody, antistreptolysin O, rheumatoid factor, and rubella.

• Section 493.931 (Routine Chemistry)

++ We are finalizing the proposed ALs in the criteria for acceptable performance.

++ We are correcting the units for prostate specific antigen (total).

++ We are making a technical change to CK-MB isoenzymes to address measurement by electrophoresis or direct mass determination.

++ We are also modifying the proposed criteria for acceptable performance for hemoglobin A1c of ±10 percent and finalizing the AL for hemoglobin A1c to ±8 percent based on public comments.

• Section 493.933 (Endocrinology)

++ We are finalizing the proposed percentage based ALs in the criteria for acceptable performance.

• Section 493.937 (Toxicology)

++ We are finalizing the proposed concentration limits and percentage based ALs in the criteria for acceptable performance.

++ We are finalizing the proposed requirement that PT programs must provide samples that cover the full range of samples that could occur in patient specimens.

++ We are correcting the units for phenytoin and vancomycin.

• Section 493.941 (Hematology)