General characteristic of the included studies
A total of 335 studies were identified from the electronic search, the manual search of the reference lists of relevant reviews did not yield new eligible studies. Of which 7 studies were excluded due to duplication. The full text was retrieved for the remaining 328 studies, 324 studies were excluded and 4 studies(5,11-13) were selected for the final analysis after detailed evaluations. The results of the study selection process are shown in Figure 1, and the general characteristics of the included studies are presented in Table 1.
Included studies published ranged from 2014 to 2015. Two studies perform general anesthesia and one study perform spinal anesthesia to conduct TKA surgery. Number of the included patients ranged from 16 to 26, and the mean number of the patients were 22. Age of the TKA patients ranged from 58.8 to 73.5. TFP administration dose were different in the included studies (12.5 μg to 50 μg). Follow-up duration from 2 days to 2 weeks.
Risk of bias
Figure. 2 and Figure. 3 shown the risk of bias summary and risk of bias graph respectively. Only one study was at low risk bias of random sequence generation. Three studies were with low risk of bias for allocation concealment, blinding of participant and personnel and outcomes assessment. Incomplete outcome data and selective reporting were all with low risk of bias in all of the included studies.
Total morphine consumption
A total of two studies compared the effect of TFP and control on total morphine consumption postoperatively. Compared with control group, administration with TFP was associated with a reduction of the total morphine consumption (WMD: -16.14; 95% CI: -25.82 to -6.46; P=0.001; Figure. 4), and significant heterogeneity was observed (I2 = 86.8%; P = 0.001).
VAS at 2 h, 4 h, 6 h, 12 h, 24 h, 48 h and 72 h
VAS at 2 h, 4 h, 6 h, 12 h, 24 h, 48 h and 72 h postoperatively can be seen in Table 2. Compared with control group, administration with TFP was associated with a significantly reduction of VAS at 2 h (WMD=-15.04; 95% CI: -22.48 to -7.60; P=0.000), 4 h (WMD=-16.15; 95% CI: -21.40 to -10.89; P=0.000), 6 h (WMD=-11.53; 95% CI: -17.27 to -5.79; P=0.000), 12 h (WMD=-18.03; 95% CI: -23.20 to -12.86; P=0.000), 24 h (WMD=-15.86; 95% CI: -20.95 to -10.76; P=0.000), 48 h (WMD=-17.08; 95% CI: -20.90 to -13.25; P=0.000) and 72 h (WMD=-11.62; 95% CI: -14.79 to -8.45; P=0.000).
Active flexion at day 1 and 3
The summary results for the effect of TCP on active flexion at day 1 and day 3 are shown in Table 2. Overall, TCP has significant effect on increasing active flexion at day 1 (WMD=10.49; 95% CI: 5.81 to 15.17; P=0.000) and day 3 (WMD=5.33; 95% CI: 2.95 to 7.71; P=0.000) compared with placebo.
Adverse effects
The summary results on occurrence of nausea, vomiting, hypertension, sweating, respiratory depression, pruritus and urine retention were shown in Table 2. Overall, no significant differences were noted in the incidence of nausea (RR: 1.09; 95% CI: 0.73 to 1.64, P = 0.675), vomiting (RR: 0.55; 95% CI: 0.23 to 1.27, P = 0.160), hypertension (RR: 0.50; 95% CI: 0.05 to 5.08, P = 0.558), sweating (RR: 0.50; 95% CI: 0.10 to 2.43, P = 0.390), respiratory depression (RR: 3.00; 95% CI: 0.13 to 69.52, P = 0.493), pruritus (RR: 1.67; 95% CI: 0.43 to 6.42, P = 0.458), urine retention (RR: 0.67; 95% CI: 0.12 to 3.57, P = 0.636) between TFP and control groups.