Simplified and improved fluid deprivation test for diagnosing diabetes insipidus

Abstract

Objective The challenge of finding patients with the rare conditon of diabetes insipidus in need of vasopressin treatment is demanding. The guidelines for performing the fluid deprivation test and interpreting the results are abundant. We evaluated the discriminative capacity of the fluid deprivation test in patients with polyuria to define a cut off for a more effective discrimination between diabetes insipidus and other polyuria syndromes. Research design and methods Retrospective review and data collection of all ambulatory fluid deprivation tests, of patients with mild polyuria and polydipsia (< 3 L/day), performed between 2000 and 2018. Serum osmolality, urine osmolality, urine volumes and clinical information of diagnosis were retrieved from the patient’s medical records. Results The study group consisted of 153 patients, 123 were diagnosed with non-diabetes insipidus and 30 with diabetes insipidus. After 12 h fasting (baseline) median duration of the fluid deprivation test was 5 h (fasting range: 12–21 h). At baseline, there was a significant difference between median serum and urine osmolality between the groups (P < 0.05). The best cut-off for the diagnosis of diabetes insipidus, was the combination of < 400 mosmol/kg in urine and > 302 mosmol/kg in serum. With this cut-off a sensitivity of 90% and specificity of 98% was achieved. Conclusion After 12 h fasting our proposed cut off clearly differentiated between diabetes insipidus, and non-diabetes insipidus suggesting a possibility to considerably reduce the duration of the fluid deprivation test.