College of American Pathologists (CAP) Survey Data
Updated 01/00
The American Diabetes Association (ADA) recommends that laboratories use only GHB assay methods that have been NGSP certified. The ADA also recommends that all laboratories performing GHB testing participate in the College of American Pathologists (CAP) fresh sample proficiency testing survey (see ADA Recommendations section on this website for more details).
In 1998 the CAP fresh sample survey for GHB (GH2) was expanded to enable all laboratories to participate. The survey included three fresh whole blood specimens (at three GHB levels). Due to the large number of participating laboratories, 5 subsets of 3 specimens each were prepared; each subset was sent to a different group of laboratories. With the first survey of 1999 all laboratories received the same set of three fresh pooled specimens.
CAP GH2 data for the second survey of 1999 is summarized below. Results from laboratories reporting HbA1c or equivalent and those reporting total GHB are included, although results from methods reporting total GHB cannot be directly compared to NGSP Reference values. Note that the NGSP target or reference values are based on replicate analyses using three NGSP certified secondary reference methods.
1999 GH2-B (fresh pooled samples)
* = NGSP certified methods (at the time of the survey)
|
|
CAP sample |
CAPsample |
CAP sample |
||||
|
NGSP Reference Valuet |
5.2 |
8.2 |
11.6 |
||||
|
|
no. labs |
Median |
%CV |
Median |
%CV |
Median |
%CV |
|
Methods reporting HbA1c (or equivalent) |
|||||||
|
* Bayer DCA 2000 |
201 |
4.9 |
3.8 |
8.0 |
3.8 |
11.3 |
4.2 |
|
* Beckman Diatrac |
32 |
4.8 |
10.1 |
7.8 |
7.6 |
11.6 |
7.6 |
|
* Beckman Synchron LX20 |
63 |
5.2 |
8.1 |
8.4 |
6.5 |
11.8 |
6.6 |
|
* Bio-Rad Diamat |
29 |
5.1 |
3.3 |
8.3 |
1.7 |
11.8 |
1.2 |
|
* Bio-Rad Diastat |
118 |
5.5 |
5.6 |
8.1 |
4.9 |
11.9 |
4.7 |
|
* Bio-Rad Variant A1c |
198 |
5.6 |
2.8 |
8.4 |
2.2 |
11.8 |
2.6 |
|
* Bio-Rad Variant Express |
18 |
5.1 |
4.1 |
8.0 |
2.8 |
11.6 |
1.5 |
|
* Helena Glyco-Tek |
9 |
5.4 |
- |
8.2 |
- |
11.5 |
- |
|
* Primus (affinity) |
12 |
5.0 |
2.3 |
7.9 |
4.4 |
11.5 |
2.9 |
|
* Roche Cobas Integra |
68 |
5.4 |
3.6 |
8.5 |
3.2 |
12.0 |
3.8 |
|
* Roche Cobas Mira |
9 |
5.7 |
- |
8.3 |
- |
11.9 |
- |
|
* Roche/BM Tina Quant II |
130 |
5.4 |
4.3 |
7.8 |
4.8 |
11.6 |
5.1 |
|
* Tosoh A1c 2.2 Plus |
147 |
5.2 |
3.6 |
8.3 |
2.7 |
12.0 |
2.4 |
|
Abbott IMX |
657 |
5.3 |
6.9 |
7.8 |
7.1 |
10.7 |
7.2 |
|
Abbott Vision |
23 |
5.3 |
7.5 |
8.1 |
4.7 |
11.7 |
4.0 |
|
Bio-Rad HbA1c column |
6 |
4.3 |
- |
7.6 |
- |
10.7 |
- |
|
Methods reporting Total GHB |
|||||||
|
Abbott IMX |
364 |
6.1 |
8.5 |
9.9 |
8.3 |
14.0 |
8.0 |
|
Bio-Rad Variant |
19 |
5.1 |
6.1 |
9.3 |
3.8 |
14.2 |
3.0 |
|
Bio-Rad Variant Express |
8 |
4.9 |
- |
9.0 |
- |
14.2 |
- |
|
Helena Glyco-Tek |
20 |
6.5 |
6.6 |
10.7 |
3.7 |
15.6 |
2.9 |
|
Isolab Glyc-Affin GHB |
9 |
5.5 |
- |
10.1 |
- |
15.2 |
- |
|
Primus |
12 |
5.6 |
4.6 |
10.6 |
3.9 |
16.6 |
2.9 |
t = Assigned as the mean value of 6 replicate analyses over two days using three NGSP certified secondary reference methods.
In 1999, based on data from the GH2 -B survey, 80% of laboratories reported results as HbA1c or equivalent - compared to 50% in 1993. Approximately 60% of those laboratories reporting results as HbA1c (48% of the total number of laboratories) used an NGSP certified method.
For all NGSP certified methods, the method-specific medians were within +0.5% HbA1c of the NGSP target at all HbA1c levels. A few non-certified methods (e.g., Abbott IMx and Bio-Rad HbA1c column) were outside this limit.
Method-specific, between-laboratory CV’s ranged from 1.2% to 10.1%. NGSP certified methods had both the highest and lowest between-laboratory CVs but most NGSP certified methods had between-laboratory CVs less than 5%. It should be noted that a method must have a total imprecision < 5% (NCCLS EP5) in order to be NGSP certified; the NGSP evaluates precision in one laboratory using one lot of reagents and calibrators. CAP precision reflects between-laboratory reproducibility, often with more than one lot of reagents and calibrators.
New NGSP certification criteria were implemented in December 1999 (after the GH2-B survey was completed). Some methods that are currently certified may not pass the new criteria. By the second survey of 2000 all methods that are listed as certified will have passed the new NGSP Assessment of Agreement criteria.