Search

The value of adding serum glucose assay to the 'biochemical profile' was assessed by a detailed follow-up of all inpatients with serum glucose concentrations of 10 mmol/l or more. After excluding patients who were on intravenous glucose infusions or were known to be diabetic. 1 in 10 was found to have unsuspected diabetes by further blood glucose measurements, glucose tolerance tests, and clinical assessment. The incidence was 0.8% of all inpatients receiving a biochemical profile, and in a full year 64 diabetics would be detected. The extra cost of laboratory tests was only 20 pounds for each diabetic detected.

We describe a new automated dye-binding method for serum albumin determination with bromcresol purple (BCP) that has several advantages over an existing bromcresol green (BCG) method. The continuous-flow method is sensitive, linear, and precise, with negligible sample interaction at an analytical rate of 60 samples per hour. Unlike BCG, BCP did not react with an albumin-free serum globulin preparation or pure human transferrin solutions. Reaction with serum was instantaneous; in contrast, BCG exhibits a slow nonspecific reaction with some specimens. The specificity of BCP was demonstrated by good agreement with results of "rocket" immunoelectrophoresis (EIA) where y(BCP) = 0.95X(EIA) + 1.72. The BCG method overestimated serum albumin concentration where y(BCG) = 1.01X(EIA) + 6.77. Precipitation, which affects the BCG method, was not observed with BCP. Blank corrections were negligible, salicylate did not interfere, and bilirubin affected the method only if present in very high concentration. The method offers a solution to the poor accuracy of existing BCG methods while retaining many of their desirable features.

Fasting glycosylated haemoglobin was analysed in 535 consecutive patients having an oral 75 g glucose tolerance test for diagnostic purposes. A reference range for fasting glycosylated haemoglobin was established from patients with a non-diabetic glucose tolerance test as defined by the World Health Organisation Expert Committee on Diabetes Mellitus. The predictive value of a glycosylated haemoglobin of over 10.0% (mean normal +/- 3 SD) for detecting a diabetic glucose tolerance test was 89% but sensitivity was only 43%. A raised glycosylated haemoglobin is useful for confirming the diagnosis of diabetes mellitus in patients with long-standing hyperglycaemia but glycosylated haemoglobin is within the reference range in many patients with newly developed diabetes or other minor abnormalities of glucose tolerance. The glucose tolerance test must remain the test of choice in these patients.

Carcinoembryonic antigen (CEA) has been measured in parallel with seven serum proteins and seromucoids in the sera of patients with malignant neoplasia and non-neoplastic disease. In the total group significant correlations were found between CEA and seromucoids and between CEA and several serum proteins. However, with two exceptions, when the individual disease groups were examined no correlation was seen. It is concluded that abnormal concentrations of the specific proteins measured do not consistently interfere in the CEA radioimmunoassay and do not explain the high CEA levels in patients with non-neoplastic diseases.