Study objective

To determine the sex-specific associations between postoperative haemoglobin and mortality or complications reflecting ischaemia or inadequate oxygen supply after major noncardiac surgery.

Design

A retrospective cohort study with prospective validation.

Setting

A large university hospital health system in China.

Patients

Men and women undergoing elective major noncardiac surgery.

Interventions and measurements

The primary exposure was nadir haemoglobin within 48 h after surgery. The outcome of interest was a composite of postoperative mortality or ischaemic events including myocardial injury, acute kidney injury and stroke within hospitalisation.

Main results

The study included 26,049 patients (15,757 men and 10,292 women). Low postoperative haemoglobin was a strong predictor of the composite outcome in both sexes, with the risk progressively increasing as the nadir haemoglobin concentration dropped below 130 g l^-1 in men and 120 g l^-1 in women (adjusted odds ratio [OR] 1.43, 95% CI 1.37-1.50 in men, and OR 1.45, 95% CI 1.35-1.55 in women, per 10 g l^-1 decrease in postoperative nadir haemoglobin). Above these sex-specific thresholds, the change of nadir haemoglobin was no longer associated with odds of the composite outcome in either men or women. There was no significant interaction between patient sex and the association between postoperative haemoglobin and the composite outcome (Pinteraction = 0.673). Validation in an external prospective cohort (n = 2120) with systematic postoperative troponin and creatinine measurement confirmed our findings.

Conclusions

Postoperative haemoglobin levels following major noncardiac surgery were nonlinearly associated with ischaemic complications or mortality, without any clinically important interaction with patient sex.