ABSTRACT
OBJECTIVE: To evaluate the accuracy of three portable glucoreflectometers (Reflolux II, Glucometer II and Glucoscot II), comparing them with the laboratory (Hitachi 704, exokinase method) and checking their reliability when measurements were repeated. DESIGN: A validity and reliability study. Three methods were evaluated and another used as reference. SETTING: The study was carried out at an urban Health Centre in Albacete. PATIENTS AND OTHER PARTICIPANTS: A sample composed of 102 types I and II diabetic patients included in an adult programme was used. The choice was made on the basis of a simple randomized sampling of the centre's records with no exclusion criteria. MEASUREMENTS AND MAIN RESULTS: The following correlations were obtained: Reflolux II = 0.94, Glucometer II = 0.85, Glucoscot II = 0.85. The difference of averages between machines and laboratory, using venous blood was: Reflolux II (2.5 +/- 2.9; t = -1.67: p > 0.01), Glucometer II (40.2 +/- 4.8; t = -16.42: p < 0.001), Glucoscot II (32.8 +/- 4.7: p < 0.001). Variation coefficients (n = 30): Reflolux II (8.1%); Glucometer II (8.5%); Glucoscot II (3.6%). CONCLUSIONS: Reflolux II was shown to be a more accurate machine than the other two under study. It is therefore the most suitable for use in Primary Care.
