Preoperative abnormalities in serum sodium concentrations are associated with higher in-hospital mortality in patients undergoing major surgery

BACKGROUND

Abnormal serum sodium concentrations are common in patients presenting for surgery. It remains unclear whether these abnormalities are independent risk factors for postoperative mortality.

METHODS

This is a secondary analysis of the European Surgical Outcome Study (EuSOS) that provided data describing 46 539 patients undergoing inpatient non-cardiac surgery. Patients were included in this study if they had a recorded value of preoperative serum sodium within the 28 days immediately before surgery. Data describing preoperative risk factors and serum sodium concentrations were analysed to investigate the relationship with in-hospital mortality using univariate and multivariate logistic regression techniques.

RESULTS

Of 35 816 (77.0%) patients from the EuSOS database, 21 943 (61.3%) had normal values of serum sodium (138-142 mmol litre(-1)) before surgery, 8538 (23.8%) had hyponatraemia (serum sodium ≤137 mmol litre(-1)) and 5335 (14.9%) had hypernatraemia (serum sodium ≥143 mmol litre(-1)). After adjustment for potential confounding factors, moderate to severe hypernatraemia (serum sodium concentration ≥150 mmol litre(-1)) was independently associated with mortality [odds ratio 3.4 (95% confidence interval 2.0-6.0), P<0.0001]. Hyponatraemia was not associated with mortality.

CONCLUSIONS

Preoperative abnormalities in serum sodium concentrations are common, and hypernatraemia is associated with increased mortality after surgery. Abnormalities of serum sodium concentration may be an important biomarker of perioperative risk resulting from co-morbid disease.

Epidemiology and outcome following post-surgical admission to critical care

PURPOSE

To describe the factors related to outcome in patients admitted to the intensive care unit (ICU) after major surgery at a national level (in Austria).

METHODS

Analysis of a prospectively collected database of ICU admissions over an 11-year period. Factors associated with mortality and how this changed with time were explored using logistic multilevel modelling.

RESULTS

A total of 88,504 surgical patients had a mean ICU length of stay of 6.5 days and total hospital stay of 31.3 days. They had an ICU mortality of 7.6% and a hospital mortality of 11.8%. Factors associated with hospital mortality included age (odds ratio (OR) 1.42 per 10 years of age), urgency of operation (2.02 for emergency when compared to elective), SAPS II score (OR 1.09), reason for admission being a medical cause and the specific nature of the surgery itself: thoracic (OR 1.81), cardiovascular (OR 1.25), trauma (OR 1.22) or gastrointestinal surgery (OR 1.71). In addition patients who had pre-existing chronic renal (OR 1.40), respiratory (OR 1.20) or cardiac failure (OR 1.29), cirrhosis (OR 2.50), alcoholism (OR 1.42), acute kidney injury (OR 1.88) and/or non-metastatic cancer (OR 1.20) were associated with higher hospital mortality than patients without this co-morbidity. There was a reduction in the OR for death over the whole 11-year period. This improved outcome remained valid even after adjusting for the identified risk factors for mortality (OR per year 0.96).

CONCLUSIONS

This study has shown the high level of demand for critical care for this patient group and an improving rate of survival.