Capillary Blood Beta-Hydroxybutyrate Measurement by Reagent Strip in Diagnosing Diabetic Ketoacidosis

Clinical Laboratory Science, Summer 2005 by Tantiwong, Puntip, Puavilai, Gobchai, Ongphiphadhanakul, Boonsong, Bunnag, Pongamorn, Ngarmukos, Chardpraorn

OBJECTIVE: To compare blood beta-hydroxybutyrate (β-OHB) levels measured by reagent strip with serum ketone levels assessed by the nitroprusside reaction in diagnosing diabetic ketoacidosis (DKA).

DESIGN: Prospective study of DKA in 19 Thai diabetic patients from September 2001 until August 2002.

SETTING: A university hospital.

PATIENTS: Nineteen patients with DKA and ten patients with metabolic acidosis from other causes.

INTERVENTIONS: Capillary blood β-OHB was measured by a blood ketone meter (MediSense® Optium(TM)). Concurrently, serum ketone was measured semiquantitatively by nitroprusside reaction (Ketostix®, N-multistix®, and Labstix®).

MAIN OUTCOME MEASURE: Sensitivity, specificity, and ROC curve of both methods in diagnosing DKA.

RESULT: Mean age ± SD of DKA patients was 45.6 ± 16.95 years. Plasma glucose was 675.25 ± 188.15 mg/dL, arterial blood pH 7.19 ± 0.12, anion gap and serum bicarbonate 29.93 ± 4.90 and 8 ± 3.35 mmol/L. Serum ketone was moderately to markedly positive in most cases. Capillary β-OHB ranged from 2.4 to >6 mmol/L. The sensitivity and specificity of serum ketone by nitroprusside reaction in diagnosing DKA were 95% and 100%. The sensitivity and specificity of capillary β-OHB was 90% and 100% respectively. The areas under ROC curves of serum ketone and capillary β-OHB were 0.975 and 0.950 (NS) respectively.

CONCLUSION: Serum ketone and blood β-OHB measurement are equally effective in diagnosing DKA among uncomplicated cases.

ABBREVIATIONS: β-OHB = beta-hydroxybutyrate; DKA = diabetic ketoacidosis; MA = metabolic acidosis; ROC = receiver operating characteristic.

INDEX TERMS: acetoacetate; beta-hydroxybutyrate; diabetic ketoacidosis; serum ketone measurement.

Clin Lab Sci 2005;18(3):139

Diabetic ketoacidosis (DKA) is a most serious acute complication of diabetes mellitus (DM). The mortality rate of patients with DKA varied between 4% and 10% in most published series.1,2,3 The higher mortality rate was found in elderly patients, complicated DKA, and in those with severe associated diseases.4,5 DKA results from absolute or relative insulin deficiency, in combination with increased levels of counter-regulatory hormones, e.g., glucagon, catecholamines, cortisol, and growth hormone. Lipolysis is enhanced resulting in the production of acetyl Co-A, which is a substrate for hepatic ketogenesis. Elevation of circulating ketone bodies, i.e., beta-hydroxybutyrate (β-OHB) and acetoacetate leads to metabolic acidosis. β-OHB is the main form of the ketone bodies with the concentration two to three times higher than that of acetoacetate.6

Because most hospital laboratories do not quantitatively measure acetoacetate or β-OHB levels routinely, clinical diagnosis of ketoacidosis depends on the semiquantitative assessment ofketones in plasma with the nitroprusside dipstick test. This test mainly measures acetoacetate, and is weakly reactive with acetone but does not detect β-OHB. As mentioned earlier, β-OHB is the predominant ketone body in DKA, and the ratio of β-OHB to acetoacetate higher than two to three may be found when the redox state of hepatocytes is changed, such as during hypoxia or shock. In these situations, semiquantitative tests for ketone bodies by the nitroprusside reaction may underestimate the total amount of ketones in the body. In addition, this assay may give a false positive result caused by drugs possessing a free sulfhydryl group.7 So the direct β-OHB measurement, although presently not recommended for diagnosing DKA seems to have more clinical benefits than the semiquantitative nitroprusside test for ketones currently used in most clinical settings. The quantitative estimation of the blood β-OHB levels can now be performed easily using reagent strips and a reflectance meter.8 The aim of this study was to compare the clinical utility of capillary blood β-OHB measurement by reagent strip and reflectance meter with the semiquantitative test for serum ketones by nitroprusside reaction in diagnosing DKA.

MATERIALS AND METHODS

We prospectively studied 29 diabetic patients with metabolic acidosis admitted in the Department of Medicine at Ramathibodi Hospital, from September 2001 to August 2002. Patients were selected based on the feasibility of performing blood β-OHB measurements and semiquantitative tests for serum acetoacetate concurrently. The first group was 19 patients with DKA (DKA group). The inclusion criteria in the DKA group were 1) plasma glucose levels greater than 250 mg/dL, 2) serum bicarbonate levels less than 18 mmol/L, 3) anion gap greater than 12 mmol/L, and 4) ketonemia as assessed by nitroprusside test.9 A second group included ten diabetic patients with metabolic acidosis due to causes other than DKA (MA group). Patients in this group had plasma glucose levels greater than 250 mg/dL, metabolic acidosis, and serum bicarbonate levels less than 18 mmol/L. Baseline capillary blood β-OHB levels, serum ketone levels measured by a nitroprusside reaction, plasma glucose concentration, serum electrolytes, and arterial blood gases were obtained in every subject.

BIOCHEMICAL MEASUREMENT

Capillary blood β-OHB measurement was performed using a reflectance hand-held meter with an electrochemical blood ketone sensor (MediSense® Optium(TM), MediSense/Abbott Laboratories, Abingdon, U.K.) whose accuracy in measuring blood β-OHB concentrations compared with an established laboratory enzymatic reference method was previously reported.8 Five microliters of capillary blood were applied to a ketone strip which was then inserted into the handheld sensor and the β-OHB concentrations were displayed in mmol/L after 30 seconds. Exact guidelines on the interpretation of blood β-OHB levels have not yet been established. The current consensus, however, suggests that levels below 0.6 mmol/L are regarded as normal, levels over 1 mmol/L represent hyperketonemia, and levels in excess of 3 mmol/L indicate ketoacidosis.10