where Hgbi is the Hgb concentration (g/dL) before surgery, Hgbe is the lowest Hgb concentration (g/dL) 24 hours after surgery, Hgbt is the total amount (g) of allogeneic Hgb transfused, and BV is the estimated total body blood volume (L).17 As Hgb concentrations after blood transfusions were compared in this study, the Hgbt variable was removed from the equation. Based on Hgb decrease data in a study that compared IV and topical TXA in TKA,14 we determined that a sample size of least 140 patients (70 in each cohort) was needed in order to have 80% power to detect a difference in Hgb decrease of 0.36 g/dL in IV and topical TXA.
All data were reported with descriptive statistics. Frequencies and percentages were reported for categorical variables. Means and standard deviations were reported for continuous variables. The groups of continuous data were compared with unpaired Student t tests and 1-way analysis of variance. Comparisons among groups of categorical data were analyzed with Fisher exact tests. Statistical significance was set at P < .05.
Results
Data were collected on 291 patients (156 THA, 135 TKA). There was a significant (P = .044) sex difference in the THA group: more men in the topical TXA subgroup and more women in the IV TXA subgroup. Other patient demographics were similarly matched with respect to age, height, weight, and body mass index (Table 1).
The secondary outcomes (differences in complications and LOS) are listed in Table 3.
Discussion
TXA, an analog of the amino acid lysine, is an antifibrinolytic agent that has been used for many years to inhibit fibrin degradation.3,18 TXA works by competitively inhibiting tissue plasminogen activation, which is elevated by the trauma of surgery, and blocking plasmin binding to fibrin.3,19 The mechanism of action is not procoagulant, as TXA prevents fibrin breakdown and supports coagulation that is underway rather than increasing clot formation. These characteristics make the drug attractive for orthopedic joint surgery—TXA reduces postoperative blood loss in patients who need fibrinolysis suppressed in order to maintain homeostasis without increasing the risk of venous thromboembolism. IV TXA has been well studied, which supports its efficacy profile for reducing blood loss and transfusions; there are no reports of increased risk of thromboembolic events.20-22 Despite these studies, the risk of adverse events is still a major concern, especially in patients with medical conditions that predispose them to venothrombotic events. Topical TXA has become a viable option, especially in high-risk patients, as studies have shown 70% lower systemic absorption relative to IV TXA plasma concentration.23 Still, too few studies have compared the efficacy, safety, and cost of IV and topical TXA in both THA and TKA.
Topical TXA costs an average of $2100 per case, primarily because standard dosing is 3 g per case. Despite repeat dosing for IV TXA (first dose at incision, second dose 3 hours after first), IV TXA costs were much lower on average: $939 less for THA and $829 less for TKA. As numerous studies have outlined results similar to ours, cost-effectiveness should be considered in decisions about treatment options.
Patel and colleagues14 reported that the efficacy of topical TXA was similar to that of IV TXA and that there were no significant differences in Hgb decrease, wound drainage, or need for transfusions after TKA. Their report conflicts with our finding significant differences favoring topical TXA for Hgb change (P = .015) and reduced calculated blood loss (P = .019) in TKA. A potential reason for these differing results is that the topical TXA doses were different (2 g in the study by Patel and colleagues,14 3 g in our study). Martin and colleagues24 compared the effects of topical TXA and placebo and found a nonsignificant difference in reduced blood loss and postoperative transfusions when the drug was dosed at 2 g. Konig and colleagues3 found that topical TXA dosed at 3 g (vs placebo) could reduce blood loss and transfusions after THA and TKA. These studies support our 3-g dose protocol for topical TXA rather than the 2-g protocol used in the study by Patel and colleagues.14 Our results are congruent with those of Seo and colleagues,25 who found topical TXA superior in decreasing blood loss in TKA. Furthermore, our study is unique in that it compared costs and found topical TXA to be more expensive by almost $1000 on average.
Wei and Wei9 concluded that IV TXA 3 g and topical TXA 3 g were equally effective in reducing total blood loss, change in hematocrit, and need for transfusion after THA. In contrast, we found a significant (P = .031) difference favoring topical TXA for Hgb change. The 2 studies differed in their dosing protocols: Wei and Wei9 infused a 3-g dose, whereas we gave a maximum of two 1-g IV doses. The higher IV dose used by Wei and Wei9 could explain why they found no difference between IV and topical TXA, whereas we did find a difference. Our study was unique in that it measured Hgb change, blood loss, and cost.
Our study included an in-depth analysis of blood loss: estimated blood loss, drain outputs, calculated blood loss, and Hgb change. The equation we used for calculated blood loss is well established and has been used in multiple studies.3,16,17 To thoroughly assess the safety of TXA, we reviewed and documented complications that occurred within 90 days after surgery and that could be attributed to TXA. This study was adequately powered and exceeded the required sample size to detect a difference in one primary outcome measure, perioperative Hgb change, as calculated by the prestudy statistical power analysis.
Our study had several limitations. First, it was a retrospective chart review; documentation could have been incomplete or missing. Second, the study was not randomized and thus subject to drug selection bias. Third, patients were selected for topical TXA on the basis of perceived risk factors, such as prior or family history of DVT, PE, cardiac events, or cerebrovascular events. It was thought that, given the decrease in systemic absorption with topical TXA, these high-risk patients would be less likely to have a thromboembolic event. Their complex past medical histories may explain why the topical TXA group had more cardiac events. Furthermore, 1 orthopedic surgeon used topical TXA exclusively, and the other 3 used it selectively, according to risk factors. In addition, unlike TKA patients, not all THA patients received drains. This study was powered to measure a difference in perioperative Hgb change but may not have been powered to detect the statistically significant difference favoring topical TXA for calculated blood loss in TKA. In the THA group, a statistically significant difference was found for reduced Hgb decrease but not for estimated or calculated blood loss. This finding reinforces some of the disparities in measurements of the effects of blood conservation strategies. The study also lacked a placebo or control group. However, several other studies have found that both IV TXA and topical TXA are superior to placebo in decreasing blood loss, Hgb change, and transfusion requirements.10,12,20,22 In addition, the effects of TXA are based on estimates of blood conservation and are not without their disparities.