The national glycohemoglobin standardization program: a five-year progress report.

BACKGROUND: The Diabetes Control and Complications Trial (DCCT) and the United Kingdom Prospective Diabetes Study (UKPDS) demonstrated conclusively that risks for complications in patients with diabetes are directly related to glycemic control, as measured by glycohemoglobin (GHB). In 1994, one year after the DCCT results were reported, the American Diabetes Association (ADA) set specific diabetes treatment goals. However, 1993 College of American Pathologists (CAP) Survey results indicated a lack of comparability of GHB test results among methods and laboratories that represented a major obstacle to meaningful implementation of the ADA guidelines. Thus, an AACC subcommittee was formed in 1993 to develop a standardization program that would enable laboratories to report DCCT-traceable GHB results. This program was implemented in 1996 by the National Glycohemoglobin Standardization Program (NGSP) Steering Committee. APPROACH: We review the NGSP process and summarize progress in standardization through analysis of CAP data. CONTENT: Since 1996, the number of methods and laboratories certified by the NGSP as traceable to the DCCT has steadily increased. CAP GH2-B survey results reported in December 2000 show marked improvement over 1993 data in the comparability of GHB results. In 2000, 90% of surveyed laboratories reported GHB results as hemoglobin A(1c) (HbA(1c)) or equivalent, compared with 50% in 1993. Of laboratories reporting HbA(1c) in 2000, 78% used a NGSP-certified method. For most certified methods in 2000, between-laboratory CVs were <5%. For all certified methods in 2000, the mean percent HbA(1c) was within 0.8% HbA(1c) of the NGSP target at all HbA(1c) concentrations.

Use of GHb (HbA1c) in screening for undiagnosed diabetes in the U.S. population.

OBJECTIVE: To evaluate the use of GHb as a screening test for undiagnosed diabetes (fasting plasma glucose > or =7.0 mmol/l) in a representative sample of the U.S. population. RESEARCH DESIGN AND METHODS: The Third National Health and Nutrition Examination Survey included national samples of non-Hispanic whites, non-Hispanic blacks, and Mexican Americans aged > or =20 years. Of these subjects, 7,832 participated in a morning examination session, of which 1,273 were excluded because of a previous diagnosis of diabetes, missing data, or fasting time of <8 h before examination. Venous blood was obtained to measure fasting plasma glucose and GHb in the remaining 6,559 subjects. Receiver operating characteristic curve analysis was used to examine the sensitivity and specificity of GHb for detecting diabetes at increasing GHb cutoff levels. RESULTS: GHb demonstrated high sensitivity (83.4%) and specificity (84.4%) for detecting undiagnosed diabetes at a GHb cutoff of 1 SD above the normal mean. Moderate sensitivity (63.2%) and very high specificity (97.4%) were evident at a GHb cutoff of 2 SD above the normal mean. Sensitivity at this level ranged from 58.6% in the non-Hispanic white population to 83.6% in the Mexican-American population; specificity ranged from 93.0% in the nonHispanic black population to 98.3% in the non-Hispanic white population. CONCLUSIONS: GHb is a highly specific and convenient alternative to fasting plasma glucose for diabetes screening. A GHb value of 2 SD above the normal mean could identify a high proportion of individuals with undiagnosed diabetes who are at risk for developing diabetes complications.

Prevalence of diabetes, impaired fasting glucose, and impaired glucose tolerance in U.S. adults. The Third National Health and Nutrition Examination Survey, 1988-1994.

OBJECTIVE: To evaluate the prevalence and time trends for diagnosed and undiagnosed diabetes, impaired fasting glucose, and impaired glucose tolerance in U.S. adults by age, sex, and race or ethnic group, based on data from the Third National Health and Nutrition Examination Survey, 1988-1994 (NHANES III) and prior Health and Nutrition Examination Surveys (HANESs). RESEARCH DESIGN AND METHODS: NHANES III contained a probability sample of 18,825 U.S. adults > or = 20 years of age who were interviewed to ascertain a medical history of diagnosed diabetes, a subsample of 6,587 adults for whom fasting plasma glucose values were obtained, and a subsample of 2,844 adults between 40 and 74 years of age who received an oral glucose tolerance test. The Second National Health and Nutrition Examination Survey, 1976-1980, and Hispanic HANES used similar procedures to ascertain diabetes. Prevalence was calculated using the 1997 American Diabetes Association fasting plasma glucose criteria and the 1980-1985 World Health Organization (WHO) oral glucose tolerance test criteria. RESULTS: Prevalence of diagnosed diabetes in 1988-1994 was estimated to be 5.1% for U.S. adults > or = 20 years of age (10.2 million people when extrapolated to the 1997 U.S. population). Using American Diabetes Association criteria, the prevalence of undiagnosed diabetes (fasting plasma glucose > or = 126 mg/dl) was 2.7% (5.4 million), and the prevalence of impaired fasting glucose (110 to < 126 mg/dl) was 6.9% (13.4 million). There were similar rates of diabetes for men and women, but the rates for non-Hispanic blacks and Mexican-Americans were 1.6 and 1.9 times the rate for non-Hispanic whites. Based on American Diabetes Association criteria, prevalence of diabetes (diagnosed plus undiagnosed) in the total population of people who were 40-74 years of age increased from 8.9% in the period 1976-1980 to 12.3% by 1988-1994. A similar increase was found when WHO criteria were applied (11.4 and 14.3%). CONCLUSIONS: The high rates of abnormal fasting and postchallenge glucose found in NHANES III, together with the increasing frequency of obesity and sedentary lifestyles in the population, make it likely that diabetes will continue to be a major health problem in the U.S.

Monitoring glycemia in diabetes. Short-term assessment.

The monitoring of glycemic status is considered a cornerstone of diabetes care. This article reviews current recommendations for routine glycemia monitoring, with emphasis on practical applications. A description of the newly developed National Glycohemoglobin Standardization Program also is provided.

Comparisons of studies on diabetic complications hampered by differences in GHb measurements.

OBJECTIVE: To compare glycated hemoglobin (GHb) values of the relationship between glycemic control and complications of diabetes from laboratories involved in long-term studies (Steno, Oslo, Stockholm, Diabetes Control and Complications Trial, and Linköping.) RESEARCH DESIGN AND METHODS: Blood samples were collected from 25 subjects selected to represent the clinically relevant measurement range. Fresh whole-blood samples were distributed and analyzed within 4 days of sample collection. Pretreatment of samples and analyses of GHb were performed according to the routine method of each study's central or reference laboratory. Results from each laboratory were compared with the group mean, i.e., the mean of all results for each sample. RESULTS: Regression analyses with the group mean values as independent variables and results from each laboratory as dependent variables showed that Oslo's result had a slope significantly different from the group mean. Laboratories used by the DCCT, Oslo, and Steno studies gave, on average, 0.4, 0.4, and 0.7% higher HbA1c readings than the group mean, respectively, while HbA1c results from Linköping and Stockholm were, on average, 0.6 and 1.0% lower, respectively. CONCLUSIONS: There were large differences in GHb values among laboratories participating in studies of diabetic complications. The present data offer a guide to the comparison of results from the studies and underscores the need for standardization of GHb measurements.

Rapid HbA1c testing in a community setting.

OBJECTIVE: To determine whether the DCA 2000 analyzer provides valid and reliable HbA1c results when used under field conditions and operated by nonmedical personnel. This study was part of a community diabetes education program, the Native American Diabetes Project, in which HbA1c was measured as an indicator of average glycemic control. RESEARCH DESIGN AND METHODS: Two study samples were taken, the first in the spring of 1994 and the second in the spring of 1995. Seven community members in 1994 and six new community members in 1995 were trained over 2 days, using standard protocol, to operate the DCA 2000 HbA1c analyzer and to collect two capillary blood samples from participants in the Native American Diabetes Project. Duplicate DCA 2000 HbA1c measurements performed by the community workers were compared with measurements from a high-performance liquid chromatography (HPLC) system. Validity and reliability measures were calculated. RESULTS: Of the participants, 43 were studied in 1994 and 14 in 1995. Comparison of the mean DCA 2000 results with those of HPLC showed high validity, with the absolute relative difference between the mean DCA 2000 and the external reference of HPLC (magnitude of mean DCA 2000-HPLC magnitude of /HPLC) as 4.0 and 2.0% for 1994 and 1995, respectively. The Pearson correlation coefficients (r) between these two measures were 0.968 and 0.996 for 1994 and 1995, respectively. While the 1994 data appeared to have less validity for values > 10%, they included only one value with a 60-min warm-up of the DCA analyzer. The 1995 data, all collected after a 60-min warm-up, had good correlation throughout the range of values. The within-run reliability was excellent, with an intraclass correlation coefficient of reliability of 0.959 and 0.975 for paired samples, for 1994 and 1995 respectively. The mean coefficient of variation for these paired measures was 3.0% in 1994 and 2.8% in 1995. Both validity and reliability were improved by changing the warm-up period of the DCA 2000 analyzer from 5 to 60 min. All correlation coefficients were statistically significant (P < 0.0001). CONCLUSIONS: The DCA 2000 gave valid and reliable HbA1c results when operated in a community setting by nonmedical personnel. Extending the warm-up period of the device to 60 min slightly improved the validity and reliability of the test.

Is glycohemoglobin testing useful in diabetes mellitus? Lessons from the diabetes control and complications trial.

To address the question, Do laboratory tests cost money or save money? we have used as a model for discussion a common chronic disease, diabetes mellitus, and a widely used laboratory test, that for glycohemoglobin, a measure of long-term glycemia used to manage diabetic patients. Diabetes mellitus is serious, highly prevalent, and costly. In 1992, $1 of every $7 spent on health in the US was for diabetes, predominantly for treatment of the chronic complications of the disease. The recently completed Diabetes Control and Complications Trial (DCCT) demonstrated that development and progression of the chronic complications of diabetes are related to the degree of altered glycemia as quantified by determinations of glycohemoglobin. Thus, use of glycohemoglobin testing for routine diabetes care provides an objective measure of a patient's risk for developing diabetic complications. Results of this test can alert patients and health providers to the need for change in the treatment plan. Optimal use of glycohemoglobin testing for diabetes care will require standardization of test results.

Glycated haemoglobin predicts progression to diabetes mellitus in Pima Indians with impaired glucose tolerance.

Glycated haemoglobin could offer several practical advantages over the OGTT for assessing glucose metabolism. Initial cross-sectional studies (1983-1985) on 381 subjects (mostly Pima Indians) described the relationship between HbA1c (a specific glycated Hb) and the OGTT. We performed follow-up OGTTs and HbA1c measurements on 257 of these same subjects 1.6-6.1 years later. Subjects were again grouped according to both the result of the OGTT (normal, IGT or diabetes, by WHO criteria) and HbA1c result (normal or elevated based on mean +/- 1.96 SD of normal). Of 66 subjects with IGT at baseline, 47 (71%) had normal HbA1c and 19 (29%) had elevated HbA1c. Twenty-six (39%) of these subjects had diabetes at follow-up. Of these subjects with IGT, a significantly greater percentage of subjects with elevated HbA1c at baseline (68%) showed worsening to diabetes than those with a normal HbA1c (28%); (chi-square = 7.8, df = 1, p < 0.01). Thus, in subjects with IGT, glycated Hb may be a useful predictor of progression to diabetes.

Glycemic control and development of retinopathy in youth-onset insulin-dependent diabetes mellitus. Results of a 12-year longitudinal study.

BACKGROUND: In 1979, the authors began a prospective study of the natural history of retinopathy in youth-onset insulin-dependent diabetes mellitus (IDDM). Their major goal was to determine if there was an association between glycemic control and the development and progression of retinopathy. METHODS: The study consisted of 420 individuals with IDDM (onset younger than 20 years of age) and no retinopathy at baseline. Study subjects were enrolled between 1979 and 1988. Stereo color fundus photographs were obtained annually. Two eye endpoints were recorded: duration when retinopathy was first detected, and when proliferative retinopathy was detected. Glycemic control was assessed by quarterly determinations of glycohemoglobin (GHb). Life-table analyses were performed relating duration of diabetes, sex, GHb, and age of diabetes onset to development of retinopathy. RESULTS: Retinopathy did not develop before 2 years' duration or before puberty. The prevalence of retinopathy was 50% by 9 years' duration and 100% by 20 years' duration. Retinopathy developed in females approximately 2 years sooner than in males, but plotting duration as postpubertal years resulted in nearly identical rates. Retinopathy developed significantly earlier in subjects with prepubertal onset of diabetes than in subjects with postpubertal onset if duration was plotted as postpubertal years. When separated into three groups based on GHb levels (< 7.5%, 7.5%-9%, > 9%), retinopathy developed approximately 2 years later in subjects in the less than 7.5% GHb group than those in the higher GHb groups. Proliferative retinopathy developed in 11 subjects. Their mean GHb level was higher than the mean GHb for those without proliferative retinopathy (10.9 versus 8.6%; P < 0.01). The higher the level of GHb, the sooner proliferative changes were detected. CONCLUSION: Long-term glycemic control is significantly related to both development and progression of retinopathy. Prepubertal duration of diabetes is a significant risk factor for the development of retinopathy.

Affinity binding assay of glycohemoglobin by two-dimensional centrifugation referenced to hemoglobin A1c.

We describe an automated assay of glycohemoglobin performed with the Abbott Vision analyzer. The assay is based on batch affinity-extraction with 3-aminophenylboronic acid-derivatized agarose beads. Reagents are packaged in a disposable test pack. Whole-blood specimens are hemolyzed with saponin within a glass capillary tube inserted into the test pack. The sample is automatically diluted with, mixed with, and separated from the solid-phase reagent. Bichromatic absorbance readings are used to calculate the percentage of hemoglobin bound. Based on the linear correlation between affinity-measured glycohemoglobin and HPLC-measured hemoglobin A1c, the percentage of hemoglobin bound is converted to a "standardized %HbA1c" result by use of regression parameters stored during a calibration run. The combination of affinity methodology with standardization by reference to HPLC produces values directly comparable with those obtained by methods specific for HbA1c. The method produces 10 results within 15 min. The assay operates with CVs < 5%, and the results correlate highly with those by ion-exchange and affinity minicolumn methods, and by ion-exchange HPLC.

Standardization of glycohemoglobin determinations in the clinical laboratory: three years of experience.

Measurement of glycohemoglobin has been recommended for the long-term assessment of glycemic control in diabetic patients. Because different analytical methods measure different glycohemoglobin species, it has been difficult to compare results between laboratories. Here we report 3 years of experience with calibration of an affinity chromatography method for measuring total glycohemoglobin (GHb). Calibration was achieved by including in each assay three hemolysate calibrators for which values for HbA1c and GHb had been determined by repeated analyses by high-performance liquid chromatography (HPLC) and affinity chromatography, respectively. Calibration improved interassay precision (CV = 3.20-7.90% and < 5.0% before and after the introduction of calibration, respectively) and eliminated lot-to-lot variability. In 91 samples, HbA1c was estimated by the calibrated affinity chromatography assay and measured by an ion-exchange HPLC method. Estimated and HPLC-measured HbA1c showed no clinically significant differences during 36 months. The high degree of long-term precision, the disappearance of lot-to-lot variability, and the excellent comparability between analytical methods measuring different species of glycated hemoglobins demonstrate the advantages of calibration.

Interlaboratory standardization of measurements of glycohemoglobins.

The diversity of methods used to measure glycohemoglobins (GHb) makes it difficult to compare patients' results among laboratories. We reported previously the feasibility of providing comparable results from different assays by use of common calibrators. We here compare results from seven different GHb methods calibrated by use of hemolysates assayed by a precise ion-exchange high-performance liquid-chromatographic (HPLC) method for hemoglobin A1c (HbA1c). Thus, regardless of the GHb species measured by the seven methods, results were referenced to the HbA1c content of the calibrators. Without this calibration, GHb values for single samples varied, e.g., from 4.0% to 8.1% and from 10% to 14.2% in the normal and high ranges, respectively. Calibration decreased between-method variability (single sample ranges of, e.g., 4.8% to 5.4% and 9.4% to 10.2% in the normal and high ranges, respectively) and improved interassay precision. We conclude that this approach to calibration of GHb measurements allows direct comparison of results obtained by different methods and improves precision.

Interlaboratory comparison of glycohemoglobin results: College of American Pathologists Survey data.

We describe recent changes in the College of American Pathologists Glycohemoglobin (gHb) Survey, made to improve the assessment of interlaboratory variability and the accuracy of results reported. The questionnaire portion of the survey was revised to include an updated list of current methods, and results for survey specimens were grouped according to the component measured (Hb A1, Hb A1c, or total gHb). The survey specimen material was changed to a material thought to give more reliable results with all available methods. After these changes, instituted in 1989, between-laboratory CVs decreased for some methods. Furthermore, gHb values between method types were more consistent with results obtained from fresh blood samples under very controlled laboratory conditions. However, these recent data also show that the interlaboratory variability is still quite high for some methods and that the variability within and between method types is still very great. We describe a pilot standardization program for gHb measurement.