Identification of studies
A total of 1145 studies were identified by the search strategies (including searching the databases, Google Scholar, and articles retrieved in the references). After automatic duplicate removal, 639 articles were saved and 24 studies were identified[17-28,16,29,15,30-38] (Fig 1. PRISMA study flow diagram).
Risk of bias and publication bias
The risk of bias is shown in Fig 2. (Fig 2. Risk of bias graph)Funnel plots were used to investigate the risk of publication bias based on the results of outcomes enrolled more than 10 studies. (Fig 3. Publication bias).
We identified 24 RCTs, with 20 in English and 4 in Chinese. All selected studies were published between 1982 and 2020. The details of enrolled studies were shown in Table 1. (Table 1. Characteristics of the included studies)And we provide the synthesized meta-analysis outcomes in Table 2.( Table 2.Outcomes of meta-analysis)
Altogether, 21 studies provided information regarding the HHS in different stages. In our meta-analysis, CHA has higher HHS results within 6 weeks (WMD=9.097, P<0.003), 3 months (WMD=3.347, P<0.001), 12 months (WMD=2.183, P=0.034) (Fig 4. Forest plot analysis of HHS).
Nine RCTs were pooled together and indicated that CHA resulted in fewer refractures than UCHA (RR= 0.227, P<0.0001) (Fig 5. Forest plot analysis of Refracture).
Subsidence or loosening
Our study find that CHA has lower rates of subsidence or loosening compared with UCHA (RR= 0.295; P<0.014)
The pooled results showed that CHA contributed more to reducing the reoperation rate (RR=0.468, P=0.004)
The incidence of dislocation was measured in 7 trials enrolling 1296 patients (CHA, 656; UCHA, 640). There was no significant difference between the CHA and UCHA groups.
The pooled analysis found no heterogeneity, and there was no difference within groups.
The results showed no difference between the two groups.
Data regarding wound infection were reported by 13 studies in 2193 patients (1104 vs 1089). No significant difference was detected (RR=1.068, P=0.781).
Eleven studies reported pneumonia in 1935 patients (975 vs 960). The results showed no significant difference between the 2 groups (RR=0.799, P=0.282).
We combined the mentioned symptoms, such as tachycardia, arrhythmia, myocardial infarction together as cardiovascular complications. And there was no significant difference between the two groups (RR=1.278, P=0.324).
Urinary Tract Infection
Seven articles considered urinary tract infections and investigated 812 patients (409 vs 403). No significant difference was found between the groups (RR=1.087, P=0.683).
Pain (Visual analogue scale, VAS)
Six studies showed that CHA has lower VAS(WMD = -0.568, P=0.001).
Deep Vein Thrombosis
Five included articles reported deep vein thrombosis in 1174 patients (598 vs 576). There was no significant difference between the two groups (RR=1.050, P=0.910).
Five trials reported pressure injury in 1069 patients (530 vs 529). The results showed that UCHA carried a higher risk of pressure injury event (RR=0.432, P=0.014).
Five trials reported pulmonary embolism in 1160 patients (590 vs 570). The rate of pulmonary embolism was higher in CHA group but the difference was not significant (RR=2.556, P=0.063).
Renal failure was measured in 4 enrolled studies. The result suggested that there was no difference (RR= 0.972, P=0.954).
Four studies provided information on gastric bleeding. The limited data showed no significant difference (RR=1.382, P=0.614) .
No significant difference was found regarding hematoma (RR=1.836, P=0.287).
Mental Status Change
No significant difference was found (RR=0.862, P=0.476).
Nineteen studies reported the operative duration. It was longer in the CHA group (WMD=8.739, P<0.00001).
This parameter was measured by 15 studies in 2277 people (CHA, 1142; UCHA, 1135). There was no significant difference(WMD=26.218mL, P=0.085), but the results should be considered carefully due to the high heterogeneity (P<0.0001, I2=80.4%) .
The length of hospital stay was described by 10 studies including 1450 patients. CHA has shorter overall length of hospital stay, although there was no significant difference(WMD=-0.330, P=0.279).
The pooled effect of 6 RCTs (n= 1682) did not favour either of the two groups in terms of blood transfusion (RR=1.000, P=0.999).
The results showed no significant differences between the CHA and UCHA groups for mortality within, within 6 months, and after 2 years. However, the mortality was slightly higher at one year in UCHA group.