Study of the Reliability of a Device (HumanSens Plus) Allowing Immediate Measurement Uric Acid in the Blood Obtained by Finger Tip Puncture.

The aim of this study is

• to assess the validity of the capillary measure of uric acid level by the HumaSens plus system by comparing the measurement of total blood uric acid level by the HumaSens plus system with the routine measurement of serum uric acid by a standard and validated automated colorimetric method (primary objective).
• to look for a systematic bias and confounding factors (secondary objectives)

Methods

Design : This is a mono-centric prospective study. Patients will be recruited from the diabetology day hospital of hospital Lariboisière, Paris

Data collection : For each patient included at the Lariboisière Diabetology day hospital and included in this study, serum uric acid levels will be measured once from venous blood samples in the biochemistry laboratory using standard automated colorimetric methods, as presently routinely done and by liquid chromatography coupled with mass spectroscopy. In, addition, uric acid levels will also be simultaneously measured from capillary blood with the HumaSensplus system.

Hematocrit, glucose, cholesterol and triglycerides will be measured as well from venous blood samples and taken medications will be listed.

Planned analysis :

Planned sample size:

To obtain an estimated intraclass correlation of 0,800, with a two-sided 95% confidence interval with a width of 0,100, if the uric acid level is measured twice for each subject, we calculated that we needed a random sample of 204 subjects.

Statistical analyses:

Primary aim This study aims at studying the reliability of measurements of acid uric level by HumanSensplus against laboratory and measurements by liquid chromatography coupled with mass spectroscopy on serum from venous blood. There is no structure to the replicate measurements in the sense that nothing ties the first or second measurement on one patient to the first or second on another. The study thus conforms to a one-way random effect model. The data will be analyzed using a one-way random-effects ANOVA model.

The estimator of the intraclass correlation coefficient will be

With BMS=between-subject mean square, WMS=within-subject mean square and k0 = number of replicates. In this study k0=2.

The intraclass correlation coefficient will be calculated and the two-sided 95%CI will be provided for each comparison.

In addition differences between measurements will be plotted against mean of measurements in accordance with Bland and Altman.

Secondary aims The search for systematic bias and effect of confounding factors (hematocrit, glycemia, lipid concentrations and concomitant medications) will use different models of multiple regression

Logistic aspects

Ethical aspects:

Agreement will be sought from the local Ethics committee. All patients will be fully informed and will be provided with an information form. The data will be entered into an excel database and will be analysed anonymously. Fingertip glycaemia is presently systematically obtained in all these day hospital patients, and the addition of fingertip uricaemia in view of substituting this measurement to the systematic serum uric acid measurement presently done from a venous system in the biochemistry laboratory can be considered as a mere attempt to improve routine patient care.

Inclusion procedure:

The study will target 200 patients seen in diabetology day hospitalisation and fulfilling the inclusion criteria. As 15 patients are seen in the diabetology day hospitalisation ward every day, we estimate that the necessary patients sample size could be obtained within 6 weeks.