Objective: To demonstrate that specificity to diagnose gestational diabetes increases when the end point diminishes from 180 to 170 mg/dL.
Material and methods: We made a transversal study based on the collection of 200 files randomly chosen, and whose diagnose was gestational diabetes identified by screening test. We established two end points: 180 and 170 mg/dL. The statistical analysis was based on the efficacy values for each one of them. Likelihood ratios for both tests were compared with the gold standard of the oral glucose tolerance test, 3 h with 100 g of glucose. A hundred patients were chosen for each group (the end point of the first and second group was greater than 180 and 170 mg/dL, respectively).
Results: The screening test results of the first group varied between 130 and 180 mg/dL. Fifty tolerance curves of four samples were negative and 50 were abnormal; 30 of these had diagnosis of gestational diabetes and 20 carbohydrate intolerance. In the second group the screening test results varied from 130 to 170 mg/dL. Sixty four tolerance curves of four samples were normal and 36 were abnormal; 24 of these had diagnosis of gestational diabetes and 12 carbohydrate intolerance.
Conclusions: Modifying the superior end point of the glucose screening test, from 180 to 170 mg/dL, maintains the test sensitivity and improves its specificity.