A comparison of analytical goals for haemoglobin A1c assays derived using different strategies

Analytical goals for the performance characteristics of assays of haemoglobin A1c (HbA1c) have been investigated using different assumptions for generation of estimates, these being based on strategies using data on biological variation and on the clinical use of results. The derived goals are highly dependent on the assumptions made. In general, in monitoring of patients (using results from the same laboratory), the analytical imprecision is the most demanding, whereas bias (inaccuracy) is the most important characteristic when strategies for several centres (laboratories) to achieve similar results are invoked. Goals for analytical quality should be given in a form in which both analytical imprecision and bias (or systematic error) are specified. When several goals are to be considered (for different relevant assumptions), the most demanding should be used.

Influence of radiographic contrast media on granulocyte enzymes and complement during uncomplicated urographies

Four different radiographic contrast media (RCM) were used for i.v. urography in 40 patients, none of whom had complications. No rise in C3d was observed for any of the RCM, indicating that complement was not activated. However, significantly decreased values for CH50 were detected when the non-ionic RCM iopamidol and iohexol were used, and this may be due to interaction between the RCM and the complement molecules. Significantly increased numbers of neutrophils were observed in patients receiving ioxaglate, iohexol and diatrizoate, which may be due to inhibition of granulocyte adherence. No rise in the concentration of elastase and lactoferrin was observed. On the other hand, significantly decreased values of elastase were seen after injection of diatrizoate, which may be due to inhibition of the degranulation process by this media.

Involvement of the extrinsic pathway in the activities of low molecular weight heparins

Anticoagulant effects of the three LMW heparins (LMWHs) Enoxaparine, Fragmin, Logiparin and of unfractionated heparin (UFH) were compared. The heparins were added to plasma to nominal concentration of 0.2 and 0.5 anti XaU/ml plasma. Dilute tissue thromboplastin (TTP) and CaCl2 were added to platelet poor plasma (PPP), platelet rich plasma (PRP) and citrated blood. Thrombin activity was recorded with chromogenic substrate. In PPP, UFH was definitely more inhibitory than LMWH. In PRP, 0.2 U/ml of LMWHs were about as effective as UFH. At 0.5 U/ml PRP, UFH and Logiparin were more effective than Enoxaparine and Fragmin. Factor XII deficient plasma was very sensitive to heparin, and UFH and Logiparin were again more inhibitory. In whole blood, fibrinopeptide A determinations showed that UFH was more inhibitory than LMWH. We conclude that the net anticoagulant effects of these heparins result from interactions with platelets in addition to accelerated inactivation of clotting factors. The in vivo anticoagulant effect of these drugs can therefore not be predicted from their nominal anti Xa and anti IIa effects alone.

Extrinsic pathway inhibitor (EPI) and the post-heparin anticoagulant effect in tissue thromboplastin induced coagulation

It is known that the anticoagulant effect of blood or plasma is greater when heparin is given in vivo than when added in similar heparin concentrations in vitro. In this study, we neutralized heparin in citrated blood with polybrene, and then triggered coagulation with dilute tissue thromboplastin (TTP) and CaCl2. The clotting time was longer and the release of fibrinopeptide A (FPA) was retarded in the post injection samples compared to samples spiked with heparin in vitro. We have earlier reported that the extrinsic pathway inhibitor (EPI) is released to the blood after heparin injection. This was demonstrated here also for LMW heparin Enoxaparine both after intravenous and subcutaneous administration. Polyclonal blocking antibodies to EPI were added to blood or plasma heparinized in vivo or in vitro, and the direct heparin effect was neutralized with polybrene. When TTP and CaCl2 now were added and clotting time and the release of FPA recorded, the postheparin effect was greatly reduced by the antibodies. Addition of EPI antibodies to post-heparin plasma samples from cancer patients caused a marked reduction in the thromboplastin clotting times. We conclude that the release of EPI to the blood contributes significantly to the anticoagulant effect of heparin ex vivo.

The influence of induced hyperglycaemia on gastric emptying rate in healthy humans

The effect on gastric emptying rate (GER) of elevated plasma glucose was investigated in eight healthy non-diabetics. They received intravenous infusions of 1000 ml 10% glucose (555 mmol, 1720 kJ) in 2 h: one-half before and the rest during the measure of GER. A control group was established with infusion of hypertonic sodium chloride and in a third group the GER was measured twice without infusion. GER was measured after 6 h of fasting, and following ingestion of a 100 g omelette (1400 kJ) tagged with 40 MBq 99mTc-sulphur colloid and 150 ml water with 8 MBq 111In-DPTA. Anterior and posterior recordings were made on gamma camera every 10th min during 1 h. Time-activity curves from the gastric area were generated for solid and liquid phases, respectively, using geometric means. The GER of solids was delayed, at least partly, by prolongation of the lag phase, and the GER of liquids was delayed following the intravenous infusion of glucose. The GER of solids was delayed following hypertonic saline infusion but not to the same extent as followed glucose in spite of the double osmotic load of saline. The percentage delay of GER of solids following glucose infusion was related to the increase in plasma glucose.

Effect of long-term monitoring of glycosylated hemoglobin levels in insulin-dependent diabetes mellitus

BACKGROUND

The value of routine measurements of glycosylated hemoglobin (hemoglobin A1c) in the care of patients with diabetes mellitus is uncertain. We undertook this study to determine whether knowledge of hemoglobin A1c values would result in improved metabolic control in a group of patients with insulin-dependent diabetes mellitus (IDDM).

METHODS

We randomly assigned 240 patients with IDDM to one of two groups that were comparable in age, sex, duration of diabetes, and initial hemoglobin A1c levels. The patients were followed for a year, and the hemoglobin A1c concentration was measured at three-month intervals. The hemoglobin A1c values were used in assessing glycemic control and modifying therapy in one of the two groups. In the other, care givers were not aware of the hemoglobin A1c levels and relied on blood or urine glucose measurements to monitor treatment.

RESULTS

Among the 222 patients still being followed after one year, the mean hemoglobin A1c value decreased significantly--from 10.1 to 9.5 percent (P less than 0.005)--in the group whose hemoglobin A1c level was monitored (n = 115), whereas the initial and one-year values in the control group (n = 107) were 10.0 and 10.1 percent, respectively. The proportion of patients with poor control, defined as those having a hemoglobin A1c value above 10.0 percent, decreased from 46 to 30 percent (P less than 0.01) in the group whose hemoglobin A1c level was monitored but did not change significantly (45 to 50 percent) in the control group. The patients in the group whose hemoglobin A1c level was monitored were seen and their insulin regimens changed more often, but they were hospitalized for acute care of their diabetes less often than those in the control group. A similar decrease in hemoglobin A1c values occurred in the control group in the following year, when their care givers knew their hemoglobin A1c values.

CONCLUSIONS

Regular measurements of hemoglobin A1c lead to changes in diabetes treatment and improvement of metabolic control, indicated by a lowering of hemoglobin A1c values.

Setting analytical goals for random analytical error in specific clinical monitoring situations

Strategies abound for the setting of analytical goals in clinical chemistry. Many, especially those more recently proposed for particular clinical situations, are concerned with tests used in diagnosis. We suggest a general theory for the setting of goals in situations that specifically involve the monitoring of individuals. Goals are calculated from the formula CVA less than [(delta c 2/2Z2)-CVB2]1/2, where CVA is the analytical imprecision (as coefficient of variation, CV); delta c is the percentage change in serial results that is considered clinically significant; Z is the Z-statistic, which depends only on the probability selected for statistical significance; and CVB is the average inherent within-subject biological variation (as CV). Examples given show applications in hematology and in monitoring diabetes mellitus, chronic renal failure, and hepatitis. The derived goals are for total random analytical error (imprecision and intermittent systematic variation), and provide objective criteria that should be achieved in practice. The effect of analytical variability on both variability in test results and the probability that a stated change can be considered significant should be calculated whether or not the goals are attained.

Setting analytical goals for random analytical error in specific clinical monitoring situations

Strategies abound for the setting of analytical goals in clinical chemistry. Many, especially those more recently proposed for particular clinical situations, are concerned with tests used in diagnosis. We suggest a general theory for the setting of goals in situations that specifically involve the monitoring of individuals. Goals are calculated from the formula CVA less than [(delta c 2/2Z2)-CVB2]1/2, where CVA is the analytical imprecision (as coefficient of variation, CV); delta c is the percentage change in serial results that is considered clinically significant; Z is the Z-statistic, which depends only on the probability selected for statistical significance; and CVB is the average inherent within-subject biological variation (as CV). Examples given show applications in hematology and in monitoring diabetes mellitus, chronic renal failure, and hepatitis. The derived goals are for total random analytical error (imprecision and intermittent systematic variation), and provide objective criteria that should be achieved in practice. The effect of analytical variability on both variability in test results and the probability that a stated change can be considered significant should be calculated whether or not the goals are attained.

Discrepancies in assessment of metabolic regulation in insulin-dependent diabetes mellitus--can we do without measurements of glycated haemoglobin?

Our purpose was to compare physicians' assessment of metabolic control in insulin-dependent diabetes mellitus (IDDM) with measurements of glycated haemoglobin HbA1c. One hundred and twenty diabetics were evaluated by physicians without access to HbA1c. Simultaneously HbA1c was measured by a highly precise isoelectric focusing method (interval of non-diabetics 5.2-6.8%). A comparison between physicians' classification into good, acceptable or poor control, and a biochemical classification by HbA1c showed a considerable discrepancy. Identity between classifications was seen in 52%, but 30% of the diabetics were considered in good or acceptable control despite HbA1c values above 10.0%. Mean HbA1c was significantly higher in women than in men (p less than 0.01). Among patients, overrated by the physicians, were relatively more women, whereas no other differences could be demonstrated. Home blood glucose monitoring did not lead to better agreement between classifications. We conclude that measurements of HbA1c is presently the most accurate single assay of metabolic control in insulin-dependent diabetics.

Association of IDDM and attenuated response of 2',5'-oligoadenylate synthetase to yellow fever vaccine

Basal and yellow fever vaccination-induced 2',5'-oligoadenylate synthetase (2',5'A) activity was determined in blood mononuclear cells (peripheral blood lymphocytes [PBLs]) from insulin-dependent diabetes mellitus (IDDM) and matched control subjects. The live attenuated yellow fever vaccine represented a primary stimulus in all subjects. First, basal 2',5'A activity increased severalfold in response to yellow fever vaccination. In IDDM subjects, this increase was significantly lower (P = .025). Second, the 2',5'A activity increased proportionately to the higher basal 2',5'A activity in IDDM subjects. In control subjects, the increase in 2',5'A activity was not dependent on the basal activity. There was no relationship between basal or stimulated 2',5'A activity and age, sex, duration of IDDM, age at onset of IDDM, metabolic control, or HLA-DQ beta-chain gene polymorphism. There is a direct relationship between 2',5'A activity and latent viral infections associated with the presence of double-stranded RNA and with cellular interferons (IFNs) formed in response to viral infections. The higher basal 2',5'A activity (P = .05) in relation to the stimulated activity may therefore signify a latent infection or the presence of double-stranded RNA in PBLs of IDDM subjects. In vitro stimulation of PBLs showed increased IFN sensitivity in IDDM subjects. Analysis of 2',5'A activity is proposed to be a sensitive measure of the activation of the IFN system and the level of latent infectivity.

Metabolic evaluation of non-insulin-dependent diabetes in primary health care--the clinical usefulness of glycated haemoglobin measurements

To evaluate the clinical usefulness of measurements of glycated haemoglobin in the assessment of metabolic regulation in non-insulin-dependent diabetes mellitus (NIDDM) in primary health care, measurements of glycated haemoglobin, HbA1c, were compared with postprandial blood glucose measurements. 120 consecutive patients with NIDDM, following the routine control by their general practitioners, were offered a measurement of HbA1c. Approximately two-thirds of the patients had HbA1c values above the acceptable level, despite postprandial blood glucose values within the recommended intervals of good and acceptable metabolic control. Blood glucose values above 10.0 mmol/l, indicating poor metabolic regulation, were reflected in high HbA1c values. Measurements of HbA1c provide an index of metabolic regulation over time, which is otherwise not obtainable in the usual clinical setting, and HbA1c should be measured regularly to evaluate long-term metabolic control and the need for intensified therapy in patients with NIDDM.

The utility of glycated hemoglobin in identification of impaired glucose tolerance

The purpose of this study was to investigate the ability of measurements of glycated hemoglobin HbAlc by a highly precise method to characterize impaired glucose tolerance (IGT). HbAlc was measured by an isoelectric focusing method (CV = 2.1% using mean of duplicates) in 100 consecutive persons referred to the laboratory by their general practitioners for an oral glucose tolerance test (OGTT) because of random blood glucose measurements within the uncertain range or positive urine tests for glucose. The patients were classified according to the most recent World Health Organization criteria in normals (n = 64), persons with IGT (n = 27) and diabetics (n = 9). HbAlc in persons with IGT was significantly higher (p less than 0.001) than in normals. It was possible to calculate a reference interval of IGT from 5.6-7.5% (mean +/- 2SD), but a considerable overlap with the reference interval of non-diabetics (5.2-6.8%) make measurements of HbAlc insufficient for diagnosis of IGT. Poor correlation was seen between plasma glucose measurements and HbAlc in persons with normal OGTT and IGT (r = 0.3). The question is, whether measurements of HbAlc provides additional information in IGT, identifying those with highest mean blood glucose concentrations in daily life, thereby being of value in itself in evaluation of glucose tolerance.

Heparin cofactor II, antithrombin and protein C in plasma from term and preterm infants

The coagulation inhibitors heparin cofactor II (HC II), antithrombin (AT) and protein C (PC) were measured in healthy term and preterm infants in order to establish reference standards. The mean value for HC II in term infants was found to be about half of the adult values. Values below 25% in healthy infants may suggest hereditary deficiency states. One girl with congenital HC II deficiency was detected. Mean AT and PC levels were somewhat higher than HC II. Healthy preterm infants have significantly lower HC II and AT values than healthy term infants. Serial AT measurements have been used in monitoring seriously ill infants and used as a prognostic indicator. In a small number of unhealthy neonates HC II was reduced to an even greater extent than AT, and on recovery normalized more rapidly than AT.

Heparin induces release of extrinsic coagulation pathway inhibitor (EPI)

Extrinsic pathway inhibitor (EPI) is a potent inhibitor of the factor VIIa-tissue thromboplastin complex. The effect of heparin on EPI activity was studied using a chromogenic substrate assay. Addition of heparin to test plasma or whole blood in vitro increased EPI activity. This increase in EPI activity was reduced by the addition of polybrene and/or antibodies to antithrombin (anti-AT). Polybrene was therefore added to the assay system. However, part of this effect (up to 20% of baseline value) was not abolished. After intravenous injection, EPI activity increased dose-dependently. The increase was about 200% of baseline value after 7500 U heparin, and was not reduced by the addition of polybrene and/or anti-AT. A slower and prolonged increase in EPI activity occurred after subcutaneous injections of unfractionated heparin and low molecular weight heparin (LMWH). Venous occlusion failed to increase EPI activity levels. Normal EPI values were observed in patients with severe liver cirrhosis and during warfarin treatment. Gel filtration of human endothelial cell culture supernatant revealed one inhibitory fraction with molecular weight about 43.000. We conclude that EPI probably is produced in endothelial cells and may be released by heparin.

Liver transplantation in a 48-year-old female with primary biliary cirrhosis

In 1983, a Danish female with primary biliary cirrhosis underwent orthotopic liver transplantation (OLT). The transplantation took place in Groningen under a Danish-Dutch cooperation, at this writing, and more than four years after transplantation, the patient is still alive. The quality of her life has been dramatically improved, making possible her return to work. Liver tests, liver function, and biopsies are normal and without signs of development of primary biliary cirrhosis in the transplanted liver. Liver transplantation is now frequently performed at several centres throughout the world with an increasing success rate. In Scandinavia, centres have been established in Norway, Finland, and Sweden, whereas in Denmark the organisation of a programme for liver transplantation is still being discussed. Because of the extremely high costs, liver transplantation has to be approved medically and politically, and the brain death criteria have to be accepted before it can be considered as a generally accepted modality of treatment.

ELISA analysis of IgA subclass antibodies to dietary antigens. Elevated IgA1 antibodies in children with coeliac disease

Enzyme immunoassays for the quantitation of IgA1 and IgA2 antibodies to dietary antigens were developed. Serum IgA1 antibodies to bovine serum albumin (BSA) were detectable in 2/30 healthy adults, in 3/26 healthy children, and at high levels in 8/11 children with coeliac disease, without relation to gluten exposure. IgA1 antibodies to ovalbumin (OA) and beta-lactoglobulin (BLG) at high titers were seen in one coeliac child but were otherwise low or absent. IgA2 antibodies to BSA were detectable in 28/48 healthy subjects and in 8/11 coeliac children. IgA2 antibodies to OA and BLG were measurable in a few samples from each group. IgA1 antibodies to the gluten component glycgli were found at low levels in 15/56 normal sera, and anti-glycgli antibodies of the IgA2 subclass in 14/48 sera from healthy persons, also at low levels. IgA1 anti-glycgli antibodies were measurable in 5/11 sera from CD patients on a gluten-free diet. Elevated levels of IgA1 anti-glycgli antibodies were detected in all sera from CD patients challenged with gluten, except in 1 patient with a markedly reduced serum IgA level. In contrast, the IgA2 anti-glycgli antibody levels were unaffected. Thus, increased levels of IgA antibodies to dietary protein antigens in childhood coeliac disease were observed only within the IgA1 isotype.

Acyclovir crystalluria

Acyclovir crystals in the urine of a 28-year-old woman are described. The patient received intravenous acyclovir for herpes encephalitis. The crystalluria was accompanied by impairment of the renal tubules as evidenced by increased urinary beta-2-microglobulin excretion. Consecutive series of 60 patients on oral acyclovir and 20 patients on intravenous acyclovir were studied in order to define the frequency of acyclovir crystalluria. None of the patients studied had crystalluria, and none elevated serum creatinine levels. It is suggested that acyclovir crystalluria might be an early sign of acyclovir crystal nephropathy.

Comparison of overnight, morning and 24-hour urine collections in the assessment of diabetic microalbuminuria

With the aim of comparing different urine collection periods in the assessment of micro-albuminuria, urinary albumin excretion rates (AERs) were measured in samples from 24 h, overnight, and morning urine collections in 54 patients aged 17 to 62 years with insulin-dependent diabetes mellitus with a mean duration of 15 years. The AER in overnight urine was found to be reduced by 25% compared to the rate in 24 h and morning urine. Assessing the ability to predict a 24 h AER within the microalbuminuric range (20-200 micrograms/min) we found a sensitivity of 90% and a specificity of 88% for both overnight and morning urine samples. These values were slightly improved by relating AER to the excretion of creatinine and it is concluded that overnight as well as morning urine collections can be used when diagnosing microalbuminuria in insulin-dependent diabetics. Furthermore the results show that the albumin to creatinine ratio in morning urine is a reliable estimate of 24 h AER and better than measurement of the albumin concentration alone.

Monitoring of heparin therapy: should heparin assays also reflect the patient's antithrombin concentration?

Chromogenic substrate (CS) assay of heparin may be performed with or without addition of antithrombin (AT) to the test plasma. Both types of assay are used for monitoring of heparin therapy, reflecting either heparin activity (heparin act), or heparin concentration (heparin conc) when AT is added. In plasma samples from 43 patients treated with intravenous heparin for DVT, the ratio between heparin act and heparin conc varied from 0.36 in patients with AT plasma concentration below 0.50 U/ml, to 0.85 in patients with AT above 1.00 U/ml (mean ratio 0.61). A formula expressing heparin act as a function of AT and heparin concentration in the test plasmas of the patients was used to calculate heparin act of the total material comprising 280 patients. Mean heparin act and heparin conc were both significantly correlated to clinical outcomes (bleeding complications, pulmonary embolism and phlebography score). For monitoring heparin therapy, guidelines for plasma heparin activity or concentration ("therapeutic ranges") are requested. When using a heparin act assay, the heparin dose needed in patients with low plasma AT concentration to reach a fixed therapeutic range, may imply undue risk of bleeding. On the other hand, when a heparin conc assay indicate plasma heparin conc within therapeutic range, antithrombotic activity may still be inadequate in patients with low plasma AT concentration.

Plasma heparin cofactor II, protein C and antithrombin in elective surgery

The coagulation inhibitors heparin cofactor II (HC II), protein C (PC) and antithrombin (AT) were measured in 14 patients undergoing uncomplicated abdominal, orthopedic or vascular surgery. Marked transient decrease of all the inhibitors was found after all surgical procedures, most prominently in vascular surgery. The greatest mean reduction was in HC II. The fall in PC and AT showed a pattern similar to that of albumin. In contrast to some earlier suggestions our results indicated that, except for HC II, the reduction probably was not due to consumption. Redistribution, hemodilution and, probably, reduced synthesis were the most important contributing factors. Serial measurements of the three coagulation inhibitors may be of prognostic value, but the present study's observed mean decrease of c. 25% in abdominal and 50-62% in vascular surgery must be borne in mind. Still lower inhibitor levels suggest consumption and a complicated postoperative course.