ESPE Abstracts

Introduction: It is challenging to differentiate central diabetes insipidus (CDI) from Nephrogenic Diabetes Insipidus (NDI) or Primary Polydipsia (PP) in patients with polyuria-polydipsia syndrome, especially in the pediatric population. Despite its limited accuracy and low tolerance in children, the water deprivation test (WDT) is still used as a reference.

Objectives: To review indications and outcomes of pediatric patients undergoing WDT; to describe the characteristics of these tests and their safety; to investigate the existence of baseline variables that predict outcomes; and to reevaluate criteria for suspending and reinterpreting tests.

Patients, materials and Methods: A retrospective analysis of 83 WDTs carried out in the investigation of 57 patients with polyuria-polydipsia syndrome, with a mean age of 8.2 years (0-17.9), in a referral University hospital over 23 years under the same protocol. All WDTs were independently reinterpreted by three experienced experts.

Results: Of the 83 tests performed, 47 (56,6%) were conclusive in establishing a definitive diagnosis. There were no serious adverse effects. The overall accuracy, including conclusive and inconclusive tests, was 53%. Simplified (2-hours) WDT (n = 21) were inconclusive in 47,6%. The mean duration of conclusive tests to confirm CDI or NDI was 2 hours, while PP and partial CDI needed 6 and 7 hours, respectively. Using ROC analysis, we found that a baseline urinary osmolality (uOsm) greater than 279 mOsm/kg excluded DI in all patients (Se= 100%; S P = 74%; P < 0.01). Moreover, a uOsm greater than 533 mOsm/kg at the time of test interruption also excluded the diagnosis of DI in all individuals (Se= 100%; S P = 100%; P < 0.01). Adopting the lower limit of 3% weight loss as a criterion for interrupting the WDT would preclude the diagnosis of nine (19%) conclusive tests.

Conclusion: WDT is safe and necessary in children. However, it is conclusive in only 53% of the tests and simplified WDT is even less effective. A basal uOsm greater than 279 mOsm/kg excludes DI, while a uOsm greater than 533 mOsm/kg at the time of test interruption is enough to exclude DI. The lower limit of 3% weight loss as a criterion for suspending WDT reduces the number of diagnoses. Considering the high rate of inconclusive WDT and its safety, we suggest revising the discontinuation and diagnostic criteria to increase its diagnostic performance. Baseline clinical and biochemical variables can help guide the indication but cannot replace the WDT.