Fibrinogen/Albumin Ratio and Outcome in Adult Cardiac Surgery With Cardiopulmonary Bypass
Abstract
BACKGROUND:
Fibrinogen/albumin ratio (F/A ratio) has been proposed as a prognostic factor for different outcomes in different clinical settings, including cardiovascular interventions. However, few studies address the F/A ratio as a possible predictor of postoperative outcome in adult cardiac surgery with cardiopulmonary bypass (CPB). The hypothesis of the present study is that by including the F/A ratio in the existing risk models for major morbidity after adult cardiac surgery may result in better discrimination, calibration, and risk reclassification of the model.
METHODS:
This is a retrospective study including 836 adult patients who underwent cardiac surgery with CPB. Every patient received an F/A ratio calculation, and other preoperative factors were collected (demographics, comorbidities, operation details, and EuroSCORE II). The primary outcome measure was major morbidity, defined as 1 or more of the following: surgical revision for any cardiac reason; stroke; acute kidney injury requiring renal replacement therapy; deep sternal wound infection; prolonged mechanical ventilation. The F/A ratio alone and EuroSCORE II-based predictive models with or without the F/A ratio were tested for major morbidity predictive properties.
RESULTS:
F/A ratio was univariately associated with major morbidity (odds ratio 1.025, 95% confidence interval 1.061–1.091, P = .004), but carried a poor discrimination for major morbidity with an area under the curve (AUC) for receiver operating characteristics (ROC) analysis of 0.606. A multivariable risk model based on the EuroSCORE II yielded a significantly (P < .001) better AUC (0.731); the inclusion of F/A ratio within this model did not significantly improve this value, which remained at 0.730. The models with or without F/A ratio had similar calibration for major morbidity prediction, and the model inclusive of F/A ratio did not significantly change the risk classification of the patients.
CONCLUSIONS:
The F/A ratio alone has poor predictive properties for major morbidity, and there is no added value in including the F/A ratio into the existing risk models. In the setting of adult cardiac surgery with CPB, we did not confirm the usefulness of the F/A ratio described in other settings, such as cardiac surgery without CPB. This may be due to both the consumption and sequestration of these proteins during CPB.