Epidemiology and prognosis of hyperosmolar state in the elderly

OBJECTIVE: We studied elderly patients admitted for hyperosmolar state (HS) to evaluate current outcome of HS and identify prognosis factors associated with mortality. STUDY DESIGN: A clinical retrospective study in an eight bed ICU. PATIENTS AND METHODS: Eighteen over 65-year-old patients admitted with a serum osmolality greater than 325 mOsm/kg were reviewed. Age, sex, diabetes mellitus, underlying medical condition, presence of an acute precipitating factor, Apache II and Glasgow scores, systolic arterial pressure, state of hydration, core temperature, heart rate, serum osmolality, creatininemia, lactatemia, plasma urea and bicarbonate, and protidemia were collected at the admission. Amount of fluid, time course of osmolality correction, length of hospitalization and mortality were recorded. All data were analyzed to identify possible correlations with patient outcome. RESULTS: Mean age: 75 +/- 11 years; sex ratio 1/2; hyperosmolar hyperglycemic states: 13 patients; hyperosmolar hypernatremic states: five patients; mean Apache II score: 18 +/- 7; Glasgow coma score: 11 +/- 3; mean osmolality: 370 +/- 25 mOsm/kg. In nine patients, infection was the precipitating factor. Five patients died (28%). At the admission, low blood pressure and high heart rate were related to mortality. During hospitalization, the occurrence of an acute cardiocirculatory failure and/or the need of mechanical ventilation significantly worsens the outcome. CONCLUSION: Our results showed that ICU mortality of HS in the elderly was at 28%. Haemodynamic state was the only factor of prognosis at the admission. Deaths were mostly related to acute respiratory and circulatory failure.