Diabetes Over View

Methods For Measuring Substances In Blood And Urine

Measurement of glucose in blood

Reductiometric methods (the Somogyi–Nelson, the ferricyanide and neocuprine autoanalyser methods) are still in use for blood glucose measurement. The o–toluidine method also remains in use but enzyme–based methods are widely available, for both laboratory and near–patient use. Highly accurate and rapid (1–2min) devices are now available based on immobilized glucose oxidase electrodes. Hexokinase and glucose dehydrogenase methods are used for reference.

An initial rapid fall in glucose of up to 10 % at room temperature is seen when whole blood samples are preserved with fluoride, but subsequent decline is slow; centrifugation prevents the initial fall. Whole blood glucose values are 15 % lower than corresponding plasma values in patients with a normal haematocrit reading, and arterial values are about 7 % higher than corresponding venous values.

The use of reagent–strip glucose oxidase methods has made bedside estimation of blood glucose popular. However, the cost of the reagent–strips remains high. Some methods still require punctilious technique, accurate timing, and storage of strips in airtight containers. Reasonably quantitative results can be obtained even with visual color–matching techniques. Electrochemical and reflectance meters can give coefficients of variation of well under 5%. Reagent–strip methods have been validated under tropical conditions, but are sensitive to extreme climatic conditions. Diabetes may be strongly suspected from the results of reagent–strip glucose estimation, but the diagnosis cannot be confidently excluded by the use of this method. Confirmation of diagnosis requires estimation by laboratory methods.

Patients are asked to collect small blood samples (either in specially prepared plastic or glass capillary tubes or on filter–paper), and self–monitor using glucose reagent–strips with direct colour–matching or meters is now widely practiced. Patients should be properly trained in the appropriate techniques to avoid inaccurate or misleading results.

The insulin–treated patient is commonly requested to build up a “glycaemic profile” by self–measurement of blood glucose at specific times of the day (and night). A “7–point profile” is useful, with samples taken before and 90 min after breakfast, before and 90 min after lunch, before and 90 min after an evening meal, and just before going to bed. Occasionally patients may arrange to wake at 0300 h to collect and measure a nocturnal sample. The complete profile rarely needs to be collected within a single 24–hour period, and it may be compiled from samples collected at different times over several days.

Measurement of glucose in urine

Patients treated with Insulin who has no access to facilities for self–measurement of blood glucose should test urine samples passed after rising, before main meals, and before going to bed. Non–insulin–dependent patients do not need to monitor their urine so frequently. Urine tests are of somewhat limited value, however, because of the great variation in urine glucose concentration for given levels of blood glucose. The correlation between blood and urine glucose may be improved a little by collecting short–term fractions (15–30 min) of the urine output. Benedict’s quantitative solution or self–boiling, caustic soda/copper sulphate tablets may be used or the more convenient, but costly, semi–quantitative enzyme–based test–strips.

Ketone bodies in urine and blood

The patient should be advised to test for ketone bodies (acetone and aceto–acetic acid) when tests for glucose are repeatedly positive, or when there issubstantial disturbance of health, particularly with infections. The appearance of persistent ketonuria associated with hyperglycaemia or high levels of glycosuria in the diabetic patient points to an unacceptably severe level of metabolic disturbance and indicates an urgent need for corrective action. Rothera’s sodium nitroprusside test may be used or, alternatively, reagent–strips that are sensitive to ketones. In emergency situations such as diabetic ketoacidosis, a greatly raised concentration of plasma ketones can be detected with a reagent–strip and roughly quantified by serial 1 in 2 dilution of plasma with water.