Study selection and baseline characteristics
A total of 789 relevant pieces of literature were retrieved, including 483 in English and 306 in Chinese. After the removal of 293 duplicate articles, 496 articles remained. After reading the title and abstract, 339 articles were excluded, leaving 157. Subsequently, the investigators read the full texts and excluded 134 studies. Finally, 23 pieces of literature were included in the current study [10-12,16-35]. A total of 10335 individuals were enrolled, including 1165 with CRBSI and 9170 without CRBSI. The literature screening process is presented in Figure 1.
The clinical characteristics of the included studies are presented in Table 1. Twenty-four studies were included in the meta-analysis. These studies reported the association between CRBSI occurrence and various factors: seven studies reported the puncture times, four studies reported femoral vein catheterization, fifteen studies described the catheter indwelling time, twelve studies investigated the age, fifteen studies reported diabetes, three studies examined the dialysis time, three studies reported the length of hospital stay and four studies reported hypoproteinemia. Twenty studies originated from China, one was derived from the USA, one was completed in Turkey, one study was from France, and one study was carried out in Iran.
Table 1 Basic characteristics of the included studies
Author
|
Country
|
Design
|
Age
|
Number
(M/F)
|
CRBSI
(Yes/No)
|
Risk factors
|
NOS score
|
Ding
2020
|
China
|
Prospective cohort
|
49.72±2.18
|
NA
|
NA
|
Age; Dialysis time;
|
8
|
Jiang
2016
|
China
|
retrospective case-control
|
57.50±11.30
|
60(38/22)
|
60/1270
|
Catheter indwelling time;
|
6
|
Liu
2020
|
China
|
retrospective case-control
|
61.4±3.1
|
12(5/7)
|
12/58
|
Age; Diabetes; Length of hospital stay;
|
7
|
Chen
2020
|
China
|
retrospective case-control
|
58.64±10.40
|
57(32/25)
|
57/341
|
Puncture times; Catheter indwelling time; Age; Diabetes; Length of hospital stay; Hypoalbuminemia
|
7
|
Shi
2021
|
China
|
retrospective case-control
|
NA
|
32(18/14)
|
32/74
|
Catheter indwelling time; Age; Diabetes; Length of hospital stay
|
7
|
Zhang
2021
|
China
|
retrospective case-control
|
62.7±5.1
|
23(14/9)
|
23/177
|
Puncture times; Femoral vein catheterization; Age; Diabetes; Catheter indwelling time
|
6
|
Zhang
2021
|
China
|
prospective cohort
|
53.72±12.0
|
35(21/14)
|
35/180
|
Catheter indwelling time; Age;
|
7
|
Wan
2014
|
China
|
retrospective case-control
|
57.2±16.3
|
NA
|
34/330
|
Diabetes
|
8
|
Wang
2021
|
China
|
retrospective case-control
|
58.4±16.2
|
80(44/36)
|
80/750
|
Puncture times; Age; Diabetes
|
7
|
Li
2019
|
China
|
retrospective case-control
|
NA
|
70(40/30)
|
70/365
|
Puncture times; Femoral vein catheterization; Catheter indwelling time; Diabetes; Hypoproteinemia
|
7
|
Wang
2020
|
China
|
retrospective case-control
|
NA
|
60(40/20)
|
60/260
|
Catheter indwelling time; Dialysis time;
|
6
|
Shen
2020
|
China
|
retrospective case-control
|
NA
|
24(14/10)
|
24/1974
|
Catheter indwelling time; Age; Diabetes
|
6
|
Cao
2019
|
China
|
retrospective case-control
|
NA
|
78(36/42)
|
78/263
|
Diabetes; Hypoalbuminemia
|
7
|
Li
2018
|
China
|
prospective cohort
|
52.77±6.52
|
35(22/13)
|
35/45
|
Puncture times; Catheter indwelling time; Diabetes;
|
7
|
Lv
2021
|
China
|
prospective cohort
|
58.62±8.73
|
31(19/12)
|
32/66
|
Femoral vein catheterization; Hypoalbuminemia
|
7
|
Hua
2019
|
China
|
retrospective case-control
|
70.18±8.4
|
31(14/17)
|
31/125
|
Femoral vein catheterization; Age; Dialysis time
|
7
|
Chen
2019
|
China
|
prospective cohort
|
66.41±11.04
|
18(7/11)
|
18/81
|
Catheter indwelling time; Age; Diabetes
|
7
|
Samani
2014
|
Iran
|
Rcross-sectional
|
58.5±12
|
44(23/21)
|
44/176
|
Puncture times; Catheter indwelling time;
|
7
|
Martin
2020
|
USA
|
prospective cohort
|
NA
|
39(23/16)
|
39/188
|
Catheter indwelling time; Diabetes
|
7
|
Caylan
2010
|
Turkey
|
prospective cohort
|
NA
|
63(34/29)
|
63/185
|
Puncture times
|
8
|
Lemaire
2009
|
France
|
retrospective cohort study
|
NA
|
226(102/124)
|
226/1523
|
Catheter indwelling time; Diabetes
|
7
|
Li
2021
|
China
|
retrospective case-control
|
NA
|
36(25/11)
|
36/340
|
Catheter indwelling time; Age; Diabetes
|
6
|
Sun
2018
|
China
|
retrospective case-control
|
53.72±12.03
|
35(21/14)
|
35/180
|
Age; Diabetes; Catheter indwelling time;
|
6
|
Results of meta-analysis
The risk factors associated with CRBSI were integrated into the current study. The results showed that 9 factors, including puncture times, femoral vein catheterization, catheter indwelling time, age, diabetes, dialysis time, hypoproteinemia and hospitalization time were associated with CRBSI.
Puncture times
The relationship between puncture time and CRBSI was reported in seven studies. Significant heterogeneity was observed between studies, and sensitivity analysis was performed. Two articles were removed, and data were synthesized using a fixed-effects model (I2 = 0%, P<0.52). The pooled results indicated that puncture times were a risk factor for CRBSI (OR=3.22, 95%CI 2.42-4.27, P<0.00001)(Figure 2).
Femoral vein catheterization
Four studies reported the relationship between femoral vein catheterization and CRBSI. No significant heterogeneity was observed across studies, so a fixed-effects model was used to conduct the data synthesis (I2 = 38%, P = 0.18). The pooled results identified femoral vein catheterization as a risk factor for the occurrence of CRBSI (OR = 1.69, 95% CI 1.38-2.08, P < 0.0001) (Figure 3).
Catheter indwelling time
The relationship between catheter indwelling time and CRBSI was described in 14 studies. Since the research results demonstrated significant heterogeneity (I2 = 83%, P < 0.00001), a random-effects model was adopted to conduct the data synthesis. The pooled results indicated that catheter indwelling time is a risk factor for the occurrence of CRBSI (OR = 2.22, 95% CI 1.62-3.04, P<0.0001) (Figure 4). To identify the potential source of heterogeneity, we conducted a sensitivity analysis. No change in the results was noted after excluding each study, indicating that the results were stable. Subsequently, we conducted a subgroup analysis according to the type of study, revealing that catheter indwelling time is a risk factor for the occurrence of CRBSI in the retrospective case-control group (OR=1.59, 95%CI=1.41,1.80) and other study type groups (OR=1.72, 95%CI=1.54,1.93).
Age
Twelve studies reported the relationship between age and CRBSI. Significant heterogeneity was observed between studies, and sensitivity analysis was performed. Two articles were removed, and data were synthesized using a fixed-effects model (I2 = 47%, P =0.04).The pooled results indicated that age is a risk factor for the occurrence of CRBSI (OR = 1.69, 95% CI 1.49-1.91; P<0.00001) (Figure 5).
Diabetes
Fifteen studies reported the relationship between diabetes and CRBSI. No significant heterogeneity was observed across studies. Therefore, a fixed-effects model was used to conduct the data synthesis (I2 = 23%, P = 0.21). The pooled results demonstrated that diabetes is a risk factor for the occurrence of CRBSI (OR = 1.92, 95% CI 1.71-2.15, P<0.00001) (Figure 6).
Dialysis time
Three studies reported the relationship between dialysis time and CRBSI. There was no significant heterogeneity across studies, and a fixed-effects model was used to conduct the data synthesis (I2 = 0%, P = 0.74). The pooled results indicated that dialysis time is a risk factor for the occurrence of CRBSI (OR = 2.89, 95%CI 1.82-4.59, P < 0.00001) (Figure 7).
Length of hospital stay
The relationship between the length of hospital stay and the occurrence of CRBSI was described in three studies. No significant heterogeneity was observed across studies, so a fixed-effects model was used to conduct the data synthesis (I2 = 39%, P = 0.19). The pooled results indicated that the length of hospital stay is a risk factor for the occurrence of CRBSI (OR = 2.49, 95%CI 1.47-4.23; P = 0.0007) (Figure 8).
Hypoproteinemia
Four studies reported the relationship between hypoproteinemia and CRBSI. Significant heterogeneity was observed between studies (I2=58%, P =0.07). Therefore, sensitivity analysis was conducted, revealing that Chen's study[19] was the main source of heterogeneity. After eliminating this study, the data were synthesized using a fixed-effects model (I2 = 0%, P =0.79). The pooled results indicated that hypoproteinemia is a risk factor for the occurrence of CRBSI (OR = 3.19, 95%CI 2.08-4.89; P<0.000001) (Figure 9).
Publication bias
A funnel plot was applied to assess publication bias. In the current study, only the diabetes risk factor was assessed for publication bias. Overall, the results showed a roughly symmetrical figure, indicating no publication bias. (Figure 10).