Post-thyroidectomy bleeding: analysis of risk factors from a national registry

Abstract

Background: Post-thyroidectomy haemorrhage occurs in 1-2 per cent of patients, one-quarter requiring bedside clot evacuation. Owing to the risk of life-threatening haemorrhage, previous British Association of Endocrine and Thyroid Surgeons (BAETS) guidance has been that day-case thyroidectomy could not be endorsed. This study aimed to review the best currently available UK data to evaluate a recent change in this recommendation.

Methods: The UK Registry of Endocrine and Thyroid Surgery was analysed to determine the incidence of and risk factors for post-thyroidectomy haemorrhage from 2004 to 2018.

Results: Reoperation for bleeding occurred in 1.2 per cent (449 of 39 014) of all thyroidectomies. In multivariable analysis male sex, increasing age, redo surgery, retrosternal goitre and total thyroidectomy were significantly correlated with an increased risk of reoperation for bleeding, and surgeon monthly thyroidectomy rate correlated with a decreased risk. Estimation of variation in bleeding risk from these predictors gave low pseudo-R2 values, suggesting that bleeding is unpredictable. Reoperation for bleeding occurred in 0.9 per cent (217 of 24 700) of hemithyroidectomies, with male sex, increasing age, decreasing surgeon volume and redo surgery being risk factors. The mortality rate following thyroidectomy was 0.1 per cent (23 of 38 740). In a multivariable model including reoperation for bleeding node dissection and age were significant risk factors for mortality.

Conclusion: The highest risk for bleeding occurred following total thyroidectomy in men, but overall bleeding was unpredictable. In hemithyroidectomy increasing surgeon thyroidectomy volume reduces bleeding risk. This analysis supports the revised BAETS recommendation to restrict day-case thyroid surgery to hemithyroidectomy performed by high-volume surgeons, with caution in the elderly, men, patients with retrosternal goitres, and those undergoing redo surgery.