Hyponatraemia is one of the commonest biochemical abnormalities in hospitalised patients, and is generally associated with a poor outcome. Large controlled studies of severely hyponatraemic patients are however unusual. From our biochemistry laboratory database we prospectively collected all patients with a sodium (Na) <125 mmol/L over a 6 month period (n=104).For each hyponatraemic case, a normotraemic control was chosen as the next sequential patient in time on the laboratory database with a normal plasma Na. For each case and control, notes were examined and details extracted of age, sex, diagnosis, drug treatment, length of hospital stay and outcome. Hyponatraemia patients were older than controls - mean (+/-SD) age 69+/-14 vs 61+/-16,p<0.001.Chest infections (21% vs 11%, p<0.001), alcohol abuse (11% vs 3%, p<0.001), and diuretic use (14% vs 1%, p<0.001) were all over-represented in the hyponatraemic group. Mortality (27% vs 10%, p=0.009) and inpatient stay (16+/-12 vs 12+/-11, p=0.005) were both significantly higher in hyponatraemic patients. Thiazide diuretics were a particular risk factor for hyponatraemia (14% vs 1%, p<0.001).Degree of admission hyponatraemia did not influence mortality risk but the lowest plasma Na level did. Thus mortality in those with a lowest level from 120-124mmol/L was 23%, from 115-119mmol/L it was 30%, and <115mmol/L it was 50%.We conclude that hyponatraemia when severe (<125mmol/L) is associated with high mortality and prolonged hospitalisation. Hyponatraemia which deteriorates after admission appears particularly hazardous.
22 - 24 Mar 2004
British Endocrine Societies