This guidance was written prior to the February 27, 1997 implementation of FDA’s Good Guidance Practices, GGP’s. It does not create or confer rights for or on any person and does not operate to bind FDA or the public. An alternative approach may be used if such approach satisfies the requirements of the applicable statute, regulations, or both. This guidance will be updated in the next revision to include the standard elements of GGP’s.
REVIEW CRITERIA FOR ASSESSMENT OF GLYCOHEMOGLOBIN (GLYCATED or GLYCOSYLATED) HEMOGLOBIN IN VITRO DIAGNOSTIC DEVICES. This is a flexible document representing the current major concerns and suggestions regarding glycohemoglobin in vitro diagnostic devices. It is based on 1) current basic science, 2) clinical experience and 3) previous submissions by manufacturers to the FDA, and 4) the Safe Medical Devices Act of 1990 (SMDA) and FDA regulations in the Code of Federal Regulations (CFR). As advances are made in science and medicine and changes in implementation of Congressional legislation, these review criteria will be re-evaluated and revised as necessary. PURPOSE OF THE GUIDANCE DRAFT This document is an adjunct to the 21 CFR Parts 800-1299. While it is not to supersede the CFR, this document provides additional guidance and clarification on what information is necessary before the Food and Drug Administration (FDA) can clear a device for marketing. The FDA can make more informed decisions based on a uniform data base. We hope this will lead to more reliable, reproducible and standardized commercial tests. DEFINITION This document discusses all generic type devices intended for use in clinical laboratories as in vitro diagnostic tests for the quantitative determination of the fraction (expressed as per cent) of glycohemoglobin (glycated or glycosylated) hemoglobin (GHb%) in blood samples.1 PRODUCT CODE: LCP REGULATION NUMBER: 21 CFR 864.7470 CLASSIFICATION. Class II PANEL: Hematology (81) REVIEW REQUIRED: Premarket Notification (510(k). 1.0 CLINICAL INDICATION/SIGNIFICANCE/INTENDED USE OF GLYCOHEMOGLOBIN TESTS. 1. Clinical Indication/Significance/Intended Use Glycohemoglobin (GHb) is being used with increasing frequency to monitor long-term blood glucose control and compliance in patients with diabetes mellitus. The GHb test provides an index of the mean concentration of blood glucose during the preceding two to three months. It complements more traditional measures of glucose control, such as glucose quantitation in urine and blood.2 Diabetes mellitus is a metabolic disorder characterized by impairment of carbohydrate metabolism. Two major types have been identified. In type I insulin dependent diabetes) the pancreatic B cells fail to produce insulin, the protein hormone necessary for glucose oxidation to take place within cells of the body. Type II or Non-Insulin-Dependent-Diabetes Mellitus (NIDDM) patients are of two subtypes: nonobese and obese. Nonobese Type II diabetics have impaired insulin production. Type II obese NIDDM patients manifest ineffective insulin stimulation of target cells which is believed to be a post-receptor defect in target tissues. The different types of diabetics present with different signs and symp-toms, but all have characteristic fasting hyperglycemia or elevation of glucose at 1 to 2 hours. The goal of treatment of diabetes mellitus has been to control blood glucose levels to as near normal as possible in the hope that this will impede, if not prevent, progression of vascular disease, e.g., stroke, blindness, renal failure, heart attack, circulatory insufficiency to lower extremities, etc. This may be accomplished through diet, weight loss, oral hypoglycemic drugs or by regular injections of insulin, depending on the type of diabetes and individual patient differences.3,4 The term "glycohemoglobin" (GHb) refers to a series of minor hemoglobin components that are stable adducts formed by hemoglobin with various sugars. The reaction between glucose and hemoglobin is an example of a nonenzymatic condensation of glucose with the free amino groups on the globin (protein) component of hemoglobin1. This process is slow, continuous and irreversible. The human erythrocyte is freely permeable to glucose. Within each erythrocyte, GHb is formed from hemoglobin at a rate dependent on the ambient concentration of glucose.5 The higher the prevailing ambient levels of blood glucose, the higher will be the levels of GHb. Per cent GHb is increased within the erythrocytes of patients with diabetes mellitus.6 Tests for glycohemoglobin are unreliable for the diagnosis of diabetes mellitus. Such use yields too many false negative and false positive results.2 It may be useful, however, to screen for those at greatest risk for diabetes complications. It is not yet clear if patients diagnosed as having diabetes based on oral glucose tolerance test results with persistent normal per cent glycohemoglobin GHb%) values develop typical diabetic complications. A claim for GHb tests to diagnose diabetes mellitus then would require premarket approval. False results could suggest incorrect long term blood glucose control. When used with the more immediate testing of glucose in blood and urine, however, false GHb% test results can be evaluated before disastrous measures are taken, because there could be short term low or high glucose levels in blood and/or urine with high or low GHb% levels (long term effect). Specimen types: List all specimen types/matrix(ices) claimed by the manufacturer for use in the test. Matrix is defined as the milieu containing the analyte in the patient sample submitted for analysis (hereinafter "specimen type" includes considerations for matrix or milieu)15. It can be whole blood, packed erythrocytes and/or lysed washed packed red cells, etc. 2. DEVICE DESCRIPTION: Discuss the principles of the device methodology and whether it is well-established or new and unproven. 3. NONCLINICAL LABORATORY STUDIES: SPECIFIC PERFORMANCE CHARACTERISTICS FDA requests different types and amounts of data and statistical analyses in applications to market in-vitro diagnostic devices. The amount and type of data requested depends on: 1) the test analyte, 2) the intended use (which determines whether the application is a 510(k), an original Premarket Approval application (PMA), 3) whether the test is quantitative or qualitative, 4) whether the data design is independent or paired, and 5) certain claims made by the manufacturer. The performance of the device can be established by comparison of the device to a predicate device (any legally marketed device). The National Committee for Clinical Laboratory Standards (NCCLS) is an example of a source for reference methods. Prove all claims for substantial equivalence and specific parameters for using the device. Include data to support use of the test with all claimed specimen types/matrices. A. Analytical/Laboratory/In Vitro Studies. 1. Statistical Describe statistical methods, assumptions used, statistical analyses and corresponding computer outputs and references well enough so a knowledgeable reader with access to the original data can verify the reported results. Provide data and statistical analyses determined with the device to support performance parameters specific to and important for operating the device, e.g., reproducibility. Discuss the statistical methods for the type of data submitted, e.g., quantitative continuous data, qualitative discrete data, etc. The distribution of data (normal vs. Non-normal type of data (paired vs. independent). Use references for study design and statistical methods that are from standard texts and/or refereed biomedical journals. Test data: 2. Performance Characteristics Establish the linearity of the test in relation to the hemoglobin concentration of the sample. Using varying concentrations of hemoglobin, document the range of hemoglobin in which accurate GHb% results are obtained. In the package insert state the hemoglobin levels which may cause the test to become over- or under-loaded and give aberrant results. Give instructions in the package insert for the action to be taken when samples are above or below linearity ranges. 3. Specificity/Cross-Reactivity/Interference Studies If the manufacturer makes a claim for restricted assay specificity, e.g., it is claimed that an immunologically-based test, electrophoresis or cation exchange methodology detects HbA1c rather than HbA1 only, the specificity of that test for the claimed restricted species should be demonstrated. For Tests Employing Cation Exchange, HPLC and Electrophoresis Methodologies, present chromatograms from gas-liquid and liquid chromatography or electrophoregrams from electrophoretic techniques so that users can see the efficiency of the separation of the claimed glycohemoglobin peak from other peaks and observe the resolution from interfering substances in the matrix. Express migration by Rf (or retention times for columns). State temperature conditions. Describe in detail the solvents or carriers and their order of use.17 Include similar types of information for electrophoresis as for chromatography as applicable.17 Present chromatograms or electrophoregrams of samples containing Hemoglobin F to determine whether of not there is interference. Present chromatograms or electerophoregrams of samples from patients with various hemoglobinopathies (such as HbS, HbG, HbH, Hb Wayne, HbC, thalassemia, etc., to determine whether or not there is interference. Describe the exact source of columns and packing materials, and their handling. Give the solvent sources and solvent compositions, and describe the history of experience with each column. Data should be provided to demonstrate that the specimen does not exhibit test interference if it is hemolyzed, lipemic (affinity chromatography only) or bilirubinemic at specified levels24. Any of these conditions found to interfere with the test should be declared as interfering conditions in the package insert. It is already documented that electrophoresis tests are unaffected by hypertriglyceridemia25 and that tests employing cation exchange methodology are affected by severe lipemia26 and bilirubinemia27. HPLC is also affected by bilirubinemia27. If these limitations are declared and referenced in the respective package insert, the respective data does not have to be provided. Labile GHb is an acutely generated, non-enzymatic, reversibly linked glucose intermediary product present in blood after a heavy meal which can cause a falsely elevated GHb% test. Submit data to support a claim that the presence of labile GHb does not cause a statistically significant elevation of the true GHb% value. Several test protocols are acceptable for this purpose. Show that normal and diabetic specimens incubated for 3 to 4 hours at 37oC with at least 55 Mm/L or 1400mg/Dl of glucose28,29 give the same GHb% values as the original untreated specimen. If the test cannot meet this stringent, non-physiological test, it will be acceptable to show that for 21 samples in the high GHb range there is not a clinical statistically significant difference between GHb% values obtained with the new test vs. GHb% values obtained using an accepted (published) labile removal method analyzed by paired student's t test, and that the "labile-removed" value is lower than that with no labile removal, i.e., labile GHb was present in the sample. For example, take 21 samples with HbA1c values between 10 and 15%. Divide each sample into 3 aliquots: a) aliquot #1 - no treatment, b) aliquot #2 remove labile by the method being tested, c) aliquot #3 - remove labile by incubation of washed erythrocytes in saline at 37oC for 5 hours or by another accepted (published) method. Run aliquots #1, #2 and #3 in the same assay. Aliquot #2 and #3 should give the same results and be lower than #1 (or equal to #1 if the assay itself is not affected by labile GHb). If an interference is not examined, then add a statement to the Limitations Section of the package insert that the device has not been tested for cross-reactivity or interference of one or more substances. Analysis of Variance of Reproducibility17,19,20,22,23 The National Committee for Clinical Laboratory Standards (NCCLS) recommends23 an analysis of variance experiment testing two clinically significant levels near medical decision limits (subnormal, normal, or elevated) of an analyte, in this case normal and elevated. Use controls simulating patient samples or actual patient specimens 2 times in the same run and in two different runs each day for 20 days. This permits separate estimation of between-day, between-run and within-day standard deviations (SD), as well as within-run and total SDs. Acceptable alternatives that include only one run per day are also discussed in the cited document.23 The three important assumptions (homogeneity of error variances, additivity, and normality) of analysis of variance should be used to demonstrate the validity of the above results. Calculate total, between- and within-day, and between- and within-run means and coefficients of variation of imprecision for each set of values. 5. Comparison Studies. Compare the device to a FDA-cleared device. In addition the device may be compared to a reference method. GHb A1c by HPLC or an affinity chromatography method is recommended for use as a reference method14. HPLC products are commercially available. HPLC shows excellent assay precision and permits rapid separation of HbA1c from the other minor components.3 The A1c peak suffers less variability from sample storage does the HbA1 peak.4 Affinity chromatography has far fewer interferences including sample degradation problems than do any of the ion-exchange or electrophoresis methods. The reference method chosen should be used for all comparisons throughout the study. Compare results obtained using packed red cells (if claimed) and/or whole blood samples free from interfering substances from 40 to 100 persons covering the whole assay range (from normal to clinically relevant high levels of GHb%) to results obtained with another test already on the market or to a reference method.17,18 Analyze the data using linear regression methods18. (The X axis is the independent variable or comparison test. The Y axis is the dependent variable or new test.)19,20 Linear-regression analysis is often most useful for estimating the differences or errors between two analytical methods, because the errors can be calculated at any medically important concentration within the range studied; furthermore, the slope and intercept may give some indication of the type of systematic error, which may aid in reducing the analytical errors. Because the reliability of the estimates of slope and intercept can be affected by nonlinearity in the data set, outliers, a narrow range of data, and variability of the comparison method, samples preferably should cover the complete range of concentrations that might be encountered.21 The slope, intercept, and their estimated standard errors, correlation coefficient, the standard error of the estimate, the assay range, and nature and size of samples tested should be reported in the Performance Characteristics section of the package insert. B. Clinical Data/Reference Ranges. The device results may correlate well using linear regression (slope close to 1.0 and intercept close to zero)17,18 with a method that has a published reference range for healthy individuals. If the device results correlate well, 40-60 subjects are sufficient to confirm agreement.30,31 If the device results do not correlate well, establish a reference range with samples from 120 to 200 normal persons characterized by age, sex, geographic location, any symptoms of disease and any other factors that would influence the values obtained, e.g., pregnancy.31 It is suggested that the statistics used to characterize the population and the confidence limits used be stated in the package insert. The populations studied should be characterized according to age and disease status. The number of persons tested should be stated. Investigate all sample type(s) claimed in the intended use unless other data proves that there is no difference between them. A range of analyte values for samples from specified patient groups, e.g., diabetics, may also be provided. 4. LABELING CONSIDERATIONS Do not make unproven claims for clinical significance in the package insert. The following are additional details for labeling. A. The Intended Use Statement [ 809.10(b)(2)] Whether it is for use in clinical laboratories, doctors' offices, or over- the-counter (OTC). (The Limitations section should include any specific training required for test performance or use.) Whether it is for screening, monitoring, confirmation or exclusion and/or to aid in the diagnosis as an adjunct to other procedures. Clinical significance, if it can be stated in a few words. (If the clinical significance statement is lengthy or complicated, create a separate heading entitled "Clinical Significance".) A typical intended use statements: "ABC's *** test is a laboratory test intended for the quantitative determination of percent glycated hemoglobin in whole blood by [methodology] using the ABC automated system (if applicable) to monitor long term blood glucose control in individuals with diabetes mellitus." Conditions for Use Describe any special applications of the device or specific contraindications or indications for use not addressed in the Intended Use Statement, e.g., "Despite serious consideration the measurement of glycohemoglobin has not been shown to be reliable in the diagnosis of diabetes mellitus."2,5 For example, one study shows that even though a cutoff of three standard deviations above the mean of a "normal" population has a specificity of 99% for diabetes, the sensitivity is only 48%49. B. Summary and Explanation of the Test [ 809.10(b)(3)] It is suggested that this section should discuss the following merits and limitations of this test: 1. For Affinity Chromatographic Methods Only; This method detects all glycohemoglobins, not just HbA132, This method is not affected by the presence of abnormal hemoglobins. This method is unaffected by carbamlyated hemoglobins in uremic patients. In uremic patients there is condensation of urea-derived cyanate with the N-terminal amino groups on the beta chains of HbA. This urea bound hemoglobin elutes with HbA1 on cation-exchange columns producing falsely elevated results33,34. 2. Labile Glycohemoglobin (All methods, if data so demonstrates.) This method is unaffected by the presence of "labile" glycosylated hemoglobin. C. Specimen collection and preparation for analysis [ 809.10(b)(7)] Data or literature references should be provided in the 510(k) to back all statements made. Include a description of: The type of specimen to be collected, e.g., whole blood, packed cells, washed packed cells, and acceptable anticoagulants. Acceptable anticoagulants or other additives, preservatives, etc., to maintain specimen. Collection Precautions: Special conditions for patient preparation, e.g., fasting, timing of collection, intervals of collection, etc. For routine clinical use, testing every 3 to 4 months is generally sufficient. In certain clinical situations such as diabetic pregnancy, or after a major change in therapy, it may be useful to obtain GHb% values in 2 to 4 week intervals5. Electrophoresis Methodology A statement that hypertriglyceridemia does not interfere electrophoresis methodology25. For Tests Employing Cation Exchange Methodology, severely lipemic samples may show elevated results26. If a sample appears lipemic, it is recommended that washed packed red cells be used as the sample. For Tests Employing Cation-Exchange or HPLC Test Methodology, it has been reported that samples with elevated levels of bilirubin may exhibit falsely elevated levels of glycohemoglobin27. State in the package insert in range of oC (low to high), collection, transport, handling and storage conditions to maintain stability of the specimen. Do not use terms such as "room temperature" without qualification. Provide data or appropriate literature references in the 510(k) to back any claims made. Due to instability of the samples for Cation-Exchange and Electrophoresis methodologies consistent storage conditions and time intervals between samples should be recommended. For example, all samples should be prepared into hemolysates on the same day of collection and tested on the morning of the fourth day after collection. A statement that each laboratory must develop and evaluate sample handling procedures based on their specific situation should be made. The laboratory should be responsible for educating clinicians concerning the large analytical errors that can be introduced by inappropriate sample handling. [Storage conditions may contribute significantly to both cation-exchange and electrophoresis assay imprecision6 The temperature variation for some cation exchange minicolumns may be 1% HbA1 per 1oC46. This temperature problem must be addressed in any one of several ways, for example: However, reports indicate that a temperature vs. GHb concentration conversion chart47 may not provide accurate results6. Such charts do not compensate for temperature fluctuations during assay11. Use three level calibrators to compensate for temperature variations and recommendations for methods of strict temperature control. Various substances other than sugars can form adducts with hemoglobin, thereby altering its charge characteristics. Falsely elevated results can occur if these adducts comigrate with GHb. Examples include individuals with opiate addiction44, lead poisoning, uremia, and alcoholism,5 as well as those receiving large doses of aspirin (acetylated hemoglobin).45 Clinically the most important of these interfering adducts occurs in uremia.5 This increase correlates with the BUN (blood urea nitrogen) and appears to result at least in part from the carbamylation of hemoglobin by urea-derived cyanate (carbamylated hemoglobin).33 D. Quality Control [ 809.10(b)(8)(vi)] Describe specimens or commercially available products that should be used for positive and negative control including recommended levels of analyte, if materials are not provided in the kit. Recommendations for frequency and placement of quality control samples within run and from run to run. Directions for interpretation of the results of quality control samples (satisfactory limits of performance). Conclude with a statement similar to the following: "If controls do not behave as stated (above), test results are invalid." E. Limitations of the Procedure [ 809.10(b)(10)] The following are examples of the types of limitation statements for GHb% that are suggested for inclusion: 1. Limitation applying to all GHb% test methods This test is unreliable for the diagnosis of diabetes mellitus2,5. There are too many false positive and/or false negative results, depending upon where the test cutoff is set. For example, one study shows that even though a cutoff of three standard deviations above the mean of a "normal" population has a specificity of 99% for diabetes, the sensitivity is only 48%49. This test is not useful in judging day-to-day glucose control, and should not be used to replace daily home testing of urine and blood glucose5. As with any other laboratory procedure, a large discrepancy between clinical impression and test results usually warrants investigation. Some of the following test limitations should be considered. Any cause of shortened red cell survival will reduce exposure of red cells to glucose with a consequent decrease in %Ghb values, e.g., hemolytic anemia or other hemolytic diseases, pregnancy, recent significant blood loss, etc. Per cent Ghb results are not reliable in any patients with chronic blood loss and consequent variable erythrocyte lifespan6,11,36,37,38. 2. All tests except those with data demonstrating lack of interference by labile Ghb. [If the manufacturer claims that the methodology of the test removes or is unaffected by labile glycohemoglobin, data should be presented in the 510(k) to back the claims.] Falsely elevated %Ghb test results can be caused by "labile Ghb", an acutely generated, reversible, non-enzymatically linked glucose intermediary product present after a heavy meal. The higher the prevailing ambient levels of blood glucose, the higher will be the potential for falsely elevated results caused by labile Ghb. The package insert should advise users that labile Ghb may be removed by any one of the following methods, depending on the test methodology employed: 1) Incubation of a twenty-fold sample dilution overnight in isotonic saline at room temperature39. 2) Dialysis overnight at 4oC vs. two changes of volumes of buffer containing 4.59 g NaH2PO4.H2O, 1.18 g Na2HPO4 and 0.65 g KCN per liter (pH 7.0)40. 3) Incubation at 37oC for 5 hours in 0.9% saline41. 4) Incubation for 30 minutes at 38oC in 30 mM semicarbazide and 12 mM aniline, pH 5. Note: These chemicals are toxic28. 5) Lysis of erythrocytes for 15 minutes in 50 volumes of 0.05M potassium biphthalate lysing buffer, Ph 5 at 37oC. Care must be taken to avoid denaturation of hemoglobin29,42. 3. For all cation exchange (including HPLC), and electrophoresis methodologies Note that elevated levels of Hemoglobin F (HbF) usually found in infants and some pregnant women result in falsely elevated results because HbF comigrates with HbA1c43. Patients with various hemoglobinopathies (such as HbS, HbG, HbH, Hb Wayne, HbC, thalassemia, etc.) may have incorrect results depending on charge characteristics of these variants6. 4. For assays quantifying HbA1 only, note that in most of the following situations, including uremia, the HbA1a and b fraction is more affected than the HbA1c fraction. Thus assays that quantify HbA1c specifically show only slight alterations (rarely greater than 1 GHb%) from these interferences.5 F.Interpretation of Results/Expected Results [ 809.10(b)(11)] Actions to be taken if a samples' test result is above or below linearity ranges according to hemoglobin concentration, e.g., dilution procedures. Ranges for healthy persons Ranges for defined disease groups. Reference ranges from the scientific literature Alternatively, reference range studies may also be reported from literature references, especially if the new test was used by the authors, or if the performance characteristics demonstrate results equivalent to those obtained with a test used by the author, e.g., regression equation slope = 1.0, r = approximately 0.95. Otherwise all expected values should be determined with the submitted product. In normal healthy individuals, GHb comprises approximately 4 to 8% of the total hemoglobin. This level may be doubled in the diabetic patient. Because glucose binding to hemoglobin occurs slowly and depends on the circulating level of blood glucose, the GHb% level represents a time- averaged blood glucose level. There is a time lag of approximately two to four weeks before GHb% reflects changes in the blood glucose level. Known insulin-dependent diabetic patients usually have elevated GHb% levels48. Percent GHb quantitations on these patients may be in the 9-17% range depending on the degree of hyperglycemia. Diabetic patients in good control may have GHb% values in the normal range. To date, there is no specific GHb% value that is accepted as indicating "good" or "poor" control. When using GHb% to monitor the diabetic patient, results must be interpreted individually; that is, the patient should be monitored against him- or herself and values compared to the normal range for that particular method. For Electrophoresis and High Performance Liquid Chromatography Test Methodology. Provide Graphics Provide examples of electrophoregrams from common hemoglobinopathies, e.g., HbS and C. Provide a detailed discussion with examples of how to calculate GHb% in the presence of genetically abnormal hemoglobins. G. Performance Characteristics [ 809.10(b)(12)] Provide the statistical results of the comparison study. Scattergrams are recommended but optional. Report the mean, standard deviation and/or per cent coefficient of variation for the various levels of %GHb in the within-run, within-day, between-day and total assay precision studies. Mean values should be arranged from lowest to highest (or vice-versa) to illustrate recognizable trends. Summarize studies performed to establish assay linearity depending on hemoglobin concentration, and state conclusions. 5. BIBLIOGRAPHY 1. Fuentes-Arderiu X. "Glycohemoglobin," not "glycated hemoglobin" or "glycosylated hemoglobin" Clin Chem 1990;36:1254. 2. Lester E. The clinical value of glycated hemoglobin and glycated plasma proteins. Ann Clin Biochem 1989;26:213-9. 3. Karam JH. Diabetes mellitus, hypoglycemia, & lipoprotein disorders. In Schroeder SA, Krupp MA, Tierney LM Jr, McPhee SJ, editors. Current medical diagnosis and treatment 1991. Norwalk and San Mateo: Appleton & Lange, 1991:852-893. 4. The National Diabetes Data Group. Diabetes 1979;28:1039 5. Goldstein DE, Little RR, Wiedmeyer HM, England JD, and McKenzie EM. Glycated hemoglobin: methodologies and clinical applications. Clin Chem 1986;32:B64-B70. 6. Goldstein DE, Wiedmeyer HM, England JD, Little RR, Parker KM. Recent advances in glycosylated hemoglobin measurements. CRC Crit Rev Clin Lab Sci 1984;21:187-228. 7. Larsen ML, Hþrder M, Mogensen EF. Effect of long-term monitoring of glycosylated hemoglobin levels in insulin-dependent diabetes mellitus. New Engl J Med 1990;323:1021-1025. 8. The DCCT Research Group. Diabetes control and complications trial (DCCT); Update. Diabetes Care 1990;13:427-33. 9. Kunkel HG and Wallenuis G. New hemoglobins in normal adult blood. Science 1955;122:288. 10. Bookchin RM, Gallop PM. Struycture of hemoglobin A1c: nature of the N- terminal beta chain blocking group. Biochem Biophys Res Commun 1968;32:86-93 11. Peacock I. Glycosylated haemoglobin: measurement and clinical use. J Clin Pathol 1984;37:841-51. 12. Mallia AK, Hermanson GT, Krohn RI, Fujimoto EK, Smith PK. Anal Lett 1981;14:649-61. 13. The DCCT Research Group. Feasibility of centralized measurements of glycated hemoglobin in the diabetes control and complications trial: a multicenter study. Clin Chem 1987;33:2267-71. 14. Little RR, England JD, Wiedmeyer HM, McKenzie EM, Mitra R, Erhart PM, Durham JB, and Goldstein DE. Interlaboratory standardization of glycated hemoglobin determinations. Clin Chem 1986;32:358-60. 15. Nomenclature and definitions for use in the national reference system for the clinical laboratory. 5(21):561. Order code NRSCL8-P, ISBN 0273- 3099 16. International Committee of Medical Journal Editors. Special report. Uniform requirements for manuscripts submitted to biomedical journals. NEJM 1991;324:424-428. 17. Information for authors. Clin Chem 1991;37:1-3. 18. National Committee for Clinical Laboratory standards. User Comparison of quantitative clinical laboratory methods using patient samples, proposed guideline. 1985;6(1). Order code EP9-P. 19. Peters T, Westgard JO. Evaluation of methods, Chapter 7 in: Tietz NW, editor. Fundamentals of clinical chemistry, Third Edition, Philadelphia: Saunders. 1987: 225-37. 20. Westgard JO, de Vos DJ, Hunt MR, Quam EF, Carey RN, Garber CC. Method evaluation. American Society of Medical Technology, Bellaire, TX, 1978. 21. Information for authors. Clin Chem 1990;36:1-4. 22. Vadlamudi SK, Stewart WD, Fugate KJ, and Tsakeris TM. Performance characteristics for an immunoassay. Scand J Clin Lab Invest 1991;51:134- 138. 23. National Committee for Clinical Laboratory Standards. Evaluation of precision performance of clinical chemistry devices - second edition; tentative guidelines. 1991:1-56. Order Code EP5-T2. 24. National Committee for Clinical Laboratory Standards. Interference Testing in clinical chemistry; proposed guideline. 1986 Order code EP7-P 25. Aleyassine H, Gardiner RJ, Blankstein LA, Dempsey ME. Agar gel electrophoretic determination of glycosylated hemoglobin: effect of variant hemoglobins, hyperlipidemia, and temperature. Clin Chem 1981;27:472. 26. Dix D, Cohen P, Kingsley S, Lea MJ, Senkbeil J, Sexton K. Interference by lactesence in glycohemoglobin analysis. Clin Chem 1979;25: 494-5. 27. Simon M, Eissler J. Critical factors in the chromatographic measurement of glycohemoglobin (HbA1). Diabetes 1980;29: 467-74. 28. Nathan DM, Avezzano ES, and Palmer JL. A rapid chemical means for removing labile glycohemoglobin. Diabetes 1981;30:700-1. 29. Biss‚ E, Berger W, Flckiger R. Quantitation of glycosylated hemoglobin, elimination of labile glycohemoglobin during sample hemolysis at pH 5. Diabetes 1982;31:630-3. 30. Ash, KO. Reference intervals (normal ranges): a challenge to laboratorians. Am J Med Tech 1980;46:504-11. 31. National Committee for Clinical Laboratory Standards. How to define, determine, and utilize reference intervals in the clinical laboratory; proposed guideline. Villanova, PA. 1991. Order code C28-P. 32. Klenk DC, Hermanson GT, Krohn RI, Fujimoto EK, Mallia AK, Smith PK, England JD, Wiedmeyer HM, Little RR, Goldstein DE. Determination of glycosylated hemoglobin by affinity chromatography: comparison with colorimetric and ion-exchange methods, and effects of common interferences. Clin Chem 1982;28: 2088-94. 33. Fluckiger R, Harmon W, Meier W, Loo S, Gabbay KH. Hemoglobin carbamylation in uremia. N Eng J Med 1981;304: 823-7. 34. Bruns DE, Lobo PI, Savory J, Wills MR. Specific affinity- chromatographic measurement of glycated hemoglobins in uremic patients. Clin Chem 1984;30: 569-71. 35. National Committee for Clinical Laboratory Standards. Internal Quality Control Testing: Principles and Definitions; approved guideline. Villanova, PA. 1991. Order code C24-A:4 36. Bunn HF, Gabbay KH, Gallop PM. The glycosylation of hemoglobin: relevance to diabetes mellitus. Science 1987;200:21-7. 37. Horton BF, Huisman THJ. Studies on the heterogeneity of haemoglobin, VII; minor haemoglobin components in haematological disease. Brit J Haemat 1965;11:296-304. 38. Lind T, Cheyne GA. Effect of normal pregnancy upon the glycosylated hemoglobin. Br J Obstet Gyn 1979;86:210-3. 39. Shenouda FS, Cockram CS, Baron MD, Tsatsoulis A, Wen Han L, Sonksen PH. Importance of short-term changes in glycosylated haemoglobin. Br Med J 1982;284:1084-5. 40. Trevilli LA, Ranney HM, Lai H. Hemoglobin components in patients with diabetes mellitus. New Engl J Med 1971;284:353-7. 41. Goldstein DE, Peth SB, England JD, Hess RL, Da Costa J. Effects of acute changes in blood glucose on HbA1c Diabetes 1980;29:623-8. 42. Bannon P. Effect of pH on the elimination of the labile fraction of glycosylated hemoglobin. Clin Chem 1982;28:2183. 43. Menard L, Dempsey ME, Blankstein LA, Aleyassine H, Wacks M, Soeldner JS. Quantitative determination of glycosylated hemoglobin a1 by agar gel electrophoresis. Clin Chem 1980;26:1598. 44. Coriello A, Giugliano D, Dello Russo P, Sgambato S, D'Onotrio F. Increased glycosylated hemoglobin A1 in opiate addicts. Evidence for hyperglycemic effect of morphine. Diabetologia 1962;22:379. 45. Nathan DM, Francis TB, Palmer JL. Effect of aspirin on determinations of glycosylated hemoglobin. Clin Chem 1983;29:466-9. 46. Schellekens APM, Sanders GTB, Thornton W, von Groenestein T. Sources of variation in the column-chromatographic determination of glycohemoglobin (HbA1). Clin Chem 1981;27: 94-9. 47. Hankins WD, Holladay L. A temperature conversion nomogram for glycosylated hemoglobin analysis. Clin Chem Acta 1980;104: 251. 48. Rahbar S. Biochem Biophys Res Commun 1969;36:838-43. 49. Forrest RD, Jackson CA, Yudkin JS. The glycohaemoglobin assay as a screening test for diabetes mellitus: the Islington Diabetes Survey. Diabetic Med. 1987;4: 254-9. HFK-440 NChace/chron 2/24/91 Version 9/27/91
This document was still considered current as of July 1997.
It will be reviewed again in July 1998.
Submit comments on this guidance document electronically via docket ID: FDA-2013-S-0610 - Specific Electronic Submissions Intended For FDA's Dockets Management Staff (i.e., Citizen Petitions, Draft Proposed Guidance Documents, Variances, and other administrative record submissions)
If unable to submit comments online, please mail written comments to:
Food and Drug Administration
5630 Fishers Lane, Rm 1061
Rockville, MD 20852
All comments should be identified with the title of the guidance.