Hospital-acquired hyponatraemia is a common and potentially serious condition. Risk factors for hospital-acquired hyponatraemia have not been studied in a controlled fashion.Methods
From 1501 patients in whom serum sodium (SNa) was determined, 50 cases with hospital-acquired hyponatraemia (in-hospital decrease in SNa≥ 7 mmol/l to < 136 mmol/l) were identified. They were matched by age, gender and department to 69 normonatraemic controls.Results
In the 50 cases, SNa fell from 141 ± 2 to 130 ± 4 mmol/l, while controls remained normonatraemic. During the development of hyponatraemia, C-reactive protein (CRP) increased in cases (median from 23 to 146 mg/l), whereas it decreased in controls (median from 31 to 24 mg/l, P = 0·008). Additional factors associated with hospital-acquired hyponatraemia included diabetes mellitus (16/50 vs. 10/69, P = 0·009) and the use of insulin (12/50 vs. 4/69, P = 0·007), antibiotics (41/50 vs. 38/69, P = 0·006) and opioids (32/50 vs. 27/69, P = 0·005). Multivariate conditional logistic regression showed that the use of insulin [odds ratio (OR) 10·5, 95% confidence interval (CI) 1·5–72·4], antibiotics (OR 4·5, 95% CI 1·4–14·6) and opioids (OR 2·9, 95% CI 1·1–7·8) was also independently associated with hospital-acquired hyponatraemia. Mortality (6/50 vs. 1/69, P = 0·04) and intensive care admission (15/50 vs. 7/69, P = 0·008) were higher in cases.Conclusions
An increase in CRP and the use of insulin, antibiotics and opioids are novel risk factors for hospital-acquired hyponatraemia. These factors represent interesting new clues regarding the pathophysiology of hospital-acquired hyponatraemia, suggesting that the acute-phase response, pain and/or direct drug effects could be involved in the release of antidiuretic hormone.