Physician misdiagnosis of dehydration in older adults

Dehydration is a difficult clinical diagnosis in older adults because the physical signs of dehydration are often confusing. The clinical consequences of a diagnosis of dehydration are critical, since dehydration implies increased morbidity and mortality and aggressive rehydration can improve clinical outcome. The diagnosis is a sentinel event for nursing homes, and often is made at transfer to a hospital.

To define the accuracy of the clinical diagnosis of dehydration during hospital admission, and to observe persons admitted from long-term care.

A total of 102 consecutive medical admissions in persons older than 65 years with a diagnostic coding for dehydration either on admission or during the course of hospitalization over a 3-month period at a university teaching hospital were reviewed. The diagnosis of dehydration was considered confirmed if the calculated serum osmolarity was greater than 295 milliosmols (mOsmol). Subjects were considered to have intravascular volume depletion if the ratio of blood urea nitrogen (BUN) to serum creatinine was greater than 20 or the serum sodium was greater than 145 milligrams per deciliter. Subjects were considered to have hypovolemia if the serum osmolarity was greater than 295 and the BUN/creatinine ratio was greater than 20.

Among subjects with a clinical diagnosis of dehydration, only 17% had a serum osmolarity > 295 mOsm, and only 11% had a serum sodium greater than 145. A BUN/creatinine ratio greater than 20 was present in 68% of the subjects. Clinicians appear to be using the term dehydration synonymously with intravascular volume depletion. Even so, at least a third of the diagnoses of intravascular volume depletion in older adults were incorrect based on laboratory data.

Physicians who diagnose dehydration during hospital admission may be relying more on physical signs than laboratory data. Little change in laboratory markers for hydration status occurs from the time of diagnosis to hospital discharge, suggesting that the clinical diagnosis does not affect fluid management. The data suggest a need for improvement in the differential diagnosis and management of volume changes in older persons.