Patient Demographics
All 90 patients included in the study were followed up postoperatively for 12 weeks, so no patients were lost to follow-up. There were no significant differences in the demographic data and preoperative blood test results among the three groups of patients (Table 1).
Table 1. Patient demographic data and preoperative blood test results
Variable
|
Control group
(30 patients)
|
i.v. group
(30 patients)
|
Topical group
(30 patients)
|
P value
(between groups)
|
Patient characteristics
|
|
|
|
|
Age (yr)
|
44.6±6.7
|
43.8±5.5
|
45.1±6.3
|
0.715a
|
Gender (male/female)
|
24/6
|
25/5
|
28/2
|
0.311b
|
BMI
|
23.1±1.4
|
22.9±1.5
|
23.2±1.7
|
0.745a
|
Medical history
|
|
|
|
|
Diabetes mellitus
|
1
|
1
|
1
|
1.000b
|
Hypertension
|
3
|
1
|
2
|
0.585b
|
Arrhythmia
|
1
|
2
|
2
|
0.809b
|
ASA score
|
|
|
|
|
Ⅰ
|
24
|
24
|
23
|
0.966b
|
Ⅱ
|
4
|
5
|
5
|
Ⅲ
|
2
|
1
|
2
|
Schatzker type
|
|
|
|
|
Ⅴ
|
14
|
17
|
11
|
0.300b
|
Ⅵ
|
16
|
13
|
19
|
Preoperative blood tests
|
|
|
|
|
Hb (g/dL)
|
12.7±1.0
|
12.8±0.9
|
13.1±1.4
|
0.360a
|
Hct (%)
|
39.2±2.7
|
39.1±2.1
|
40.1±3.1
|
0.283 a
|
D-dimer (mg/L)
|
6.3±1.2
|
6.7±1.7
|
6.1±0.8
|
0.191 a
|
FIB (g/L)
|
4.6±1.1
|
5.2±1.5
|
4.8±1.2
|
0.172 a
|
Abbreviations: BMI, Body mass index; ASA, American Society of Anesthesiologists; Hb, Hemoglobin; Hct, Hematocrit; FIB, fibrinogen.
Intergroup comparisons performed using ANOVA or Chi-square test (aANOVA; bChi-square test).
Operation Time, IBL, and TPD
The mean operation time in the control, i.v., and topical groups was 117.7±19.4, 109.1±17.5, and 115.3±20.5 min, respectively. There was no significant difference between these values (P = 0.206), although the time in the i.v. group was shorter than in the other groups. The average values of IBL in the control, i.v., and topical groups were 144.5±21.1, 116.5±15.2, and 137.5±19.2 mL, respectively, which were significantly different (P < 0.001); the best effect occurred in the i.v. group (P < 0.001). The topical group showed the greatest reduction in TPD among the three groups, with a significant difference compared to the i.v. group (64.1 mL vs 76.5 mL, P = 0.004) (Table 2).
Table 2. Postoperative data and postoperative blood test results
Variable
|
Control group
(30 patients)
|
i.v. group
(30 patients)
|
Topical group
(30 patients)
|
P value
(between groups)
|
Intergroup comparison
|
P1 P2 P3
|
Surgical data
|
|
|
|
|
|
|
|
Duration of surgery (min)
|
117.7±19.4
|
109.1±17.5
|
115.3±20.5
|
0.206a
|
-
|
-
|
-
|
Duration of tourniquet (min)
|
97.3±8.5
|
95.4±10.1
|
101.6±11.4
|
0.056a
|
-
|
-
|
-
|
IBL (mL)
|
144.5±21.1
|
116.5±15.2
|
137.5±19.2
|
< 0.001a
|
< 0.001
|
0.184
|
< 0.001
|
TPD (mL)
|
156.3±27.6
|
76.5±15.2
|
64.1±16.3
|
< 0.001a
|
< 0.001
|
< 0.001
|
0.004
|
Postoperative blood tests
|
|
|
|
|
|
|
|
HB (g/dL)
|
|
|
|
|
|
|
|
24 h
|
11.7±1.1
|
12.2±2.2
|
11.9±0.8
|
0.430a
|
-
|
-
|
-
|
POD#3
|
11.2±2.4
|
12.5±3.4
|
11.6±1.2
|
0.125a
|
-
|
-
|
-
|
Hct (%)
|
|
|
|
|
|
|
|
24 h
|
33.5±4.1
|
34.1±3.9
|
33.7±4.4
|
0.849a
|
-
|
-
|
-
|
POD#3
|
33.7±2.9
|
34.6±2.4
|
34.1±3.1
|
0.466a
|
-
|
-
|
-
|
D-dimer (mg/L)
|
|
|
|
|
|
|
|
24 h
|
9.6±3.4
|
8.1±2.9
|
8.4±3.3
|
0.165a
|
-
|
-
|
-
|
POD#3
|
11.2±4.1
|
9.8±3.1
|
10.3±3.8
|
0.335a
|
-
|
-
|
-
|
FIB (g/L)
|
|
|
|
|
|
|
|
24 h
|
5.7±2.1
|
5.1±1.9
|
5.4±2.3
|
0.546a
|
-
|
-
|
-
|
POD#3
|
7.4±2.6
|
6.4±1.3
|
6.8±2.9
|
0.264
|
-
|
-
|
-
|
Real Hb decrease (g/dL)
|
1.5±0.6
|
0.9±0.3
|
1.2±0.4
|
< 0.001a
|
< 0.001
|
0.026
|
0.002
|
Transfusion rate (%)
|
6.7%
|
0
|
0
|
0.129b
|
-
|
-
|
-
|
Abbreviations: POD#3, the third postoperative day.
P1 represents the P value obtained by comparison between control group and i.v. group;
P2 represents the P value obtained by comparison between control group and topical group;
P3 represents the P value obtained by comparison between i.v. group and topical group;
Intergroup comparisons performed using AVONA or Chi-square test (aANOVA; bChi-square test).
Postoperative Blood Tests
The results of the routine blood tests and blood coagulation tests in the three groups are summarized in Table 2. The mean values of postoperative Hb at 24 hours in the control, i.v., and topical groups were 11.7±1.1, 12.2±2.2, and 11.9±0.8 g/dL, respectively, with no significant difference between the groups (P = 0.430). The mean values of POD3 Hb in the control, i.v., and topical groups were 11.2±2.4, 12.5±3.4, and 11.6±1.2 g/dL respectively, and there was also no significant difference between the groups (P = 0.125). Similarly, there were no significant differences in the mean Hct values between groups for the above two time points. However, there was a statistically significant difference in the average real Hb reduction among the three groups (P < 0.001): the values were 1.5±0.6 g/dL in the control group, 0.9±0.3 g/dL in the i.v. group, and 1.2±0.4 g/dL in the topical group. The best effect was observed for the i.v. group (P = 0.002, versus the topical group).
The mean values of postoperative D-dimer after 24 h in the control, i.v., and topical groups were 9.6±3.4, 8.1±2.9, and 8.4±3.3 mg/L, respectively, with no significant difference between groups (P = 0.165). The mean values of POD3 D-dimer in the control, i.v., and topical groups were 11.2±4.1, 9.8±3.1, 10.3±3.8 mg/L, respectively, with no significant difference between groups (P = 0.335). Similarly, there were no significant differences for mean FIB values among the three groups at the above two time points (all p values >0.05).
Blood Loss and Transfusion
After careful calculation and verification, we found that there were significant differences in TBL among the three groups (P < 0.001). The pairwise comparisons and statistical analyses between groups, namely, control group vs i.v. group, i.v. group vs topical group, and control group vs topical group, all showed statistically significant differences (Fig. 1). Similar results were also obtained for HBL; the HBL values in the control, i.v., and topical groups were 341.1±43.7, 219.3±33.4, and 224.5±33.6 mL, respectively (all intergroup P values < 0.001) (Fig. 2).
The final analysis revealed that two patients in the control group were transfused with 2 units of PRBCs owing to symptoms of postoperative anemia, whereas no patients in the i.v. group and the topical group required blood transfusion. No statistical differences were found in the transfusion rate among the three groups (P = 0.129) (Table 2).
Vascular Events, Wound Complications, and Adverse Reactions
In this study, the venous plexus of the calf muscle was the most common site for DVT, followed by the popliteal vein. No severe complications, such as pulmonary embolism, myocardial infarction, or cerebral infarction, occurred to any patient enrolled in this study. There were no significant differences in wound complications and adverse reactions among the three groups. The specific data are described in Table 3.
Table 3. Vascular events, wound complications, and adverse reactions resulting from TXA
Variable
|
Control group
(30 patients)
|
i.v. group
(30 patients)
|
Topical group
(30 patients)
|
P value
(between groups)
|
Vascular events
|
|
|
|
|
DVT
|
3
|
5
|
3
|
0.661
|
MI
|
0
|
0
|
0
|
-
|
CI
|
0
|
0
|
0
|
-
|
PE
|
0
|
0
|
0
|
-
|
GIH
|
0
|
1
|
0
|
0.364
|
Total
|
3
|
6
|
3
|
0.421
|
Wound complications
|
|
|
|
|
Dehiscence
|
0
|
0
|
0
|
-
|
Hematoma
|
0
|
0
|
0
|
-
|
Edge necrosis
|
0
|
0
|
0
|
-
|
Infection
|
1
|
0
|
0
|
0.364
|
Total
|
1
|
0
|
0
|
0.364
|
Adverse reactions
|
|
|
|
|
Epilepsy
|
0
|
0
|
0
|
-
|
Rash
|
0
|
1
|
0
|
0.364
|
Headache
|
2
|
4
|
1
|
0.338
|
Nausea and Vomiting
|
1
|
1
|
1
|
1.000
|
Total
|
3
|
6
|
2
|
0.260
|
Abbreviations: TXA, tranexamic acid; DVT, deep vein thrombosis; MI, myocardial infarction; CI, cerebral infarction; PE, pulmonary embolism; GIH, gastrointestinal hemorrhage.
Chi-square test was performed for intra-group comparisons
Pain, Functional, and QoL assessment (Figures. 3,4,5)
According to our results, topical TXA had the best effect on pain control in the early postoperative period (POD2 and POD4, P = 0.002 and 0.025, respectively), as well as in the later periods (PO6W, P <0.001).
The knee function HSS scores were used to assess the knee function in patients both before and after surgery. Although there were no significant differences in postoperative follow-up among the three groups at PO 6W and PO 12W (P values: 0.052 and 0.432), it is worth mentioning that the topical group benefitted the most, followed by the i.v. group.
To assess the overall knee function, we used both the PCS and MCS of the SF-12. In PCS, the topical group showed a sustainable development advantage from the 6th week to the 12th week after surgery (P = 0.237 and 0.051, respectively). This advantage is likely to become more significant as the follow-up time increases. The same conclusion also applies to the comparison of MCS among the three groups.