Essays

In February 2019, the Centers for Medicaid and Medicare issued proposed changes to the CLIA regulations. This is potentially the largest shake-up in US regulations in decades. PLEASE NOTE: The 2019 proposed goals were replaced by CLIA 2024 goals!

New CLIA Proposed Rules for Acceptance Limits for Proficiency Testing

 Part I. Background and Summary/Comparison

 James O. Westgard, PhD and Sten Westgard, MS
March 2019

IMPORTANT NOTE: The goals proposed by CLIA 2019 were replaced/supplanted by the goals officially entered into the Federal Register in 2022, goals which were to take official effect in 2024. Thus, CLIA 2024 goals are now the goals to pay attention to, not CLIA 2019.

https://www.westgard.com/2024-clia-requirements.htm

What follows below is of historial note, useful for context....

On February 4, 2019, new CLIA “proposed rules” were published in the Federal Register to expand the list of regulated analytes and define new criteria for acceptable performance for proficiency testing (PT) [1].  These changes were apparently initiated in 2008 by the Clinical Laboratory Improvement Advisory Committee’s (CLIAC) recommendation that the acceptance limits (ALs) for PT from the original 1992 publication of CLIA rules should be updated to reflect improvements in technology and changes in the utilization of laboratory tests.  Recall that the original ALs were developed in the late 1980s based on the observed variation in proficiency testing events.  Many ALs were estimated from the results of PT surveys and often stated in the form of the Target Value (mean of the group) ± 3 SD’s of the PT peer group.  In addition, the original CLIA list of regulated analytes was based on medical practice at that time and did not include tests such as HbA1c or PSA that are routine today. Thus, these changes in ALs and expansion of the list of regulated analytes are long overdue.

Some definitions of terms

• Criteria for acceptable performance is the general language used in CLIA rules to describe Target Values (TV) and Acceptance Limits (AL) for scoring proficiency testing results.
• Target Value generally refers to the mean result for a peer group of over 10 participants, but may also be established by definitive and reference methods.
• Peer Group is a group of laboratories that have similar analytic methods (instruments, measurement procedures, reagent systems), but does not require the same reagent lot.
• Acceptance Limit is defined as the symmetrical tolerance (plus and minus) around the Target Value. The preferred format is ± a percentage of TV, but sometimes ± concentration limit is also includes, particularly for performance at low concentrations.

According to the proposed rules, acceptance limits 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… to establish performance specifications.  The implication is that the quality required for acceptable performance in PT programs is not necessarily the same as the quality required in laboratories for clinical services.  That suggests that the AL’s for PT performance may not be as demanding as needed for clinical services, even though the format of an AL as an Allowable Total Error (ATE) is commonly used for defining the quality requirements for laboratory tests.  In justifying the format for proposed ALs, the document states fixed ALs, either in percentages or concentration units, are preferred to SDs for PT…[because] they can be tied directly to objective goals for performance, such as goals for analytical accuracy and technical expectations.  Limits in the form of ATE are therefore preferred for PT rather than separate specifications for precision and bias.

Regulated analytes

            The proposed list of regulated analytes includes a few test deletions and many additions based on the following stated criteria:

• Current availability of PT materials and the number of PT programs already offering PT;
• Volume of patient testing being performed nationwide;
• Impact on patient and/or public health; and
• Cost and feasibility of implementation.

Analytes deleted include LDH isoenzymes, ethosuximide, quinidine, primidone, and procainamide. Proposed additions in general immunology include Anti-HBs, Anti-HCV, C-reactive protein (high sensitivity). Analytes added in routine chemistry include B-NP, Pro-BNP, Cancer antigen (CA) 125, Carbon dioxide, Carcinoembryonic antigen, Cholesterol low density lipoprotein, Ferritin, Gamma glutamyl transferase, Hemoglobin A1c, Phosphorus, Prostate specific antigen total, Total iron binding capacity, Troponin I, and Troponin T.  In endocrinology, additions include Estradiol, serum Folate, Follicle stimulating hormone, Luteinizing hormone, Progesterone, Prolactin, Parathyroid hormone, Testosterone, Vitamin B12. In toxicology, additions include serum Acetaminophen, Salicylate, and Vancomycin.

            The proposed rules also include extensive discussion of changes for microbiology, first by defining categories of testing (stains, susceptibility and resistance testing, antigen and/or toxin detection, and microbial identification and detection) and second by identifying major groups of microorganisms to be tested for bacteriology, mycobacteriology, mycology, parasitology, and virology. 

Acceptance Limits

Our intent here is to focus on the new recommendations for ALs for specific analytes or tests and to compare the new ALs with the old ALs from the original publication in 1992 [2] and the "Final" Rule published in 2003 [3].   In describing how the new ALs were determined, it appears that biologic variation has been considered together with “state of the art” performance.  The ATE quality goals from the Ricos database were considered to estimate ALs, followed by simulation of those ALs to determine the expected failure rate based on current PT survey performance and to assure “miss rates” of only 1 to 2 percent.  The exact process is difficult to assess from the description in the proposed rules, but this is our best guess based on the discussion on pages 1543 through 1545 in in the Federal Register [1].  It would have been helpful to have a few detailed examples to understand the exact process for setting ALs.  It would have also been useful to have more specific and up-to-date references, rather than to chapters in the 2006 and 2012 editions of the Tietz textbook (4^th and 5^th editions, vs current 2018 6^th edition).  The references tend to reflect the ideas and recommendations at the start of this project a decade ago and ignore more recent discussions, particularly related to the 2014 Milan conference on analytical specifications.   

Summary and comparison tables

The tables are presented separately in our CLIA section, as well as below.

The five tables summarize the new proposed ALs for routine chemistry, immunology, endocrinology, toxicology and hematology and compare those values to the earlier ALs in the original CLIA rules. 

For immunology, ALs have been changed from TV ± 3SD to TV ± % for 8 tests and from TV ± 2 dilutions to TV ± 3SD for 4 others.  For 6 other tests, the ALs remain the same.

For chemistry, the list has 12 additions and now includes 37 regulated analytes.  The ALs for 15 tests have been tightened, 6 remain the same, and 2 are changed from TV ± 3SD to TV ±%.  For glucose, the discussion on page 1544 describes use of both percent and concentration limts (e.g., original 10% and 6 mg/dL, whichever is greater), but the new recommendation in the table in section §493.931 shows only TV ± 8% (greater) without stating any concentration. 

For chemistry, some of the recommendations for ALs already seem out-of-date.  For example, for HbA1c, an AL of 10% is recommended in the new proposed rule.  Compare that to the criteria that have been employed in the CAP PT surveys, where the AL was 15% in 2007, tightened to 12% in 2008, 10% in 2009, 8% in 2010, 7% in 2011, 6.0% in 2014, and is currently 5.0%.  Adding HbA1c to the list of regulated analytes is certainly important, but it is also important to set the AL to be consistent with global standards and practices, which are better aligned with the CAP recommendations, the NGSP requirements for certification of manufacturers’ methods, and the IFCC global standards. 

For endocrinology, there are 10 additions to the regulated list that now totals 18.  All tests have ALs in the form of TV ±% and 6 also include six tests have ALs that have changed from TV ± 3SD to TV ±%.  Likewise, all toxicology ALs are in the form TV ±% and have generally been tightened from 25% to 20% and 20% to 15%. 

The hematology list of regulated tests remains the same and the ALs for 4 of the 10 tests have been tightened. 

References

1. CMS, CDC, HSS. Clinical Laboratory Improvement Amendments of 1988 (CLIA) Proficiency Testing Regulations Related to Analytes and Acceptable Performance. Fed Reg 2019; 84:1536-1567.
2. US Department of Health and Social Services. Medicare, Medicaid, and CLIA Programs: Regulations implementing the Clinical Laboratory Improvement Amendments of 1988 (CLIA). Final Rule. Fed Regist `992 (Feb 28);57:7002-7186.
3. US Centers for Medicare & Medicaid Services (CMS). Medicare, Medicaid, and CLIA Programs: Laboratory Requirements Relating to Quality Systems and Certain Personnel Qualifications. Final Rule.  Fed Regist Jan 24 2003;16:365-3714.

Chemistry

Immunology

Endocrinology

Toxicology

Hematology