What Should Define Preoperative Anemia in Primary THA?

Abstract

Background: The use of tranexamic acid (TXA) in THA decreases the risk of transfusion after surgery. However, nearly 10% of patients still undergo a transfusion, which has been independently associated with an increased risk of complications. Preoperative anemia has been proven to be a strong predictor of transfusion after THA, but the ideal "cutoff" values in today's population that maximize sensitivity and specificity to predict transfusion have yet to be established.

Questions/purposes: (1) Which preoperative factors are associated with postoperative transfusion in the setting of TXA use? (2) If preoperative hemoglobin (Hgb) remains associated with transfusion, what are the best-supported preoperative Hgb cutoff values associated with increased transfusion after THA?

Methods: A retrospective chart analysis was performed from January 1, 2013, to January 1, 2015, on 558 primary THAs that met prespecified inclusion criteria. A multivariable logistic regression analysis model was used to identify independent factors associated with transfusion. Area under the receiver-operator curve (AUC) was used to determine the best-supported preoperative Hgb cut point across all participants, as well as adjusted by sex and TXA use. Overall, 60 patients with a blood transfusion were included and compared with 498 control subjects (11% risk of transfusion).

Results: After controlling for potential confounding variables such as age, sex, American Society of Anesthesiologist score, intravenous TXA (IV TXA) use, and preoperative Hgb, we found that patients with lower preoperative Hgb (g/dL per 1-unit decrease, odds ratio [OR], 2.6; 95% CI, 2.0-3.5; p < 0.001), female sex (vs male, OR, 4.2; 95% CI, 1.7-10.3; p = 0.002), and those unable to receive IV TXA (topical TXA/no TXA, OR, 13.5; 95% CI, 6.3-28.6; p < 0.001) were more likely to receive a transfusion. Of these, preoperative Hgb was found to be the variable most highly associated with transfusion (AUC, 0.876). A preoperative Hgb cutoff value of 12.6 g/dL maximized the AUC (0.876) for predicting transfusion across all patients unadjusted for baseline characteristics (sensitivity = 83, specificity = 84) with values of 12.5 g/dL (sensitivity = 85, specificity = 77) and 13.5 g/dL (sensitivity = 92, specificity = 77) for women and men, respectively.

Conclusions: The 1968 WHO definitions of anemia (preoperative Hgb < 13 g/dL for men and < 12 g/dL for women) used currently may underestimate patients at risk of transfusion after THA today. Further studies are needed to see if blood conservation referral decreases the risk of transfusion with preoperative treatment of anemia.

Level of evidence: Level III, therapeutic study.

Fig. 1

The receiver operator curve (ROC) for predicting transfusion using preoperative Hgb, sex (male or female), and TXA use (yes or no) across all patients is shown. An area of 0.5 indicates no discriminability, whereas an area of 1 indicates perfect ability to predict transfusion with a given covariate. The blue solid line represents preoperative hemoglobin (AUC = 0.8762), the red dash line represents sex (AUC = 0.6661), and the green dash-dot line represents TXA (AUC = 0.7740).

Fig. 2

The predictive probability of blood transfusion by preoperative Hgb adjusted for sex is shown.