. 2022 Aug 5;44(1):1263–1279. doi: 10.1080/0886022X.2022.2105233

Table 3.

Meta-analysis of the effect of anticoagulant interventions in patients under BPT.

Outcomes or subgroup analysis	Study reference number	Statistical method	Risk ratio or mean difference (95% CI)	P	I² value (%)	Certainty of the evidence (GRADE)
Main outcomes: the effect of CT versus NM on bleeding complication
Bleeding complication	[17,27–31,34–36]	M-H, fixed	RR = 1.45, CI: [1.08, 1.94]	0.01	44	⊕^⊝⊝⊝
Bleeding complication		M-H, random	RR = 1.70, CI: [0.91, 3.16]	0.09	44	Very low^a
Subgroup analysis: the relationship between conventional anticoagulant interventions classification and bleeding complication
NA	[17,27–29]	M-H, fixed	RR = 0.90, CI: [0.37, 2.22]	0.67	0	⊕^⊝⊝⊝; Very low^a
NA	[17,27–29]	M-H, random	RR = 0.97, CI: [0.37, 2.44]	0.95	0	⊕^⊝⊝⊝; Very low^a
UFH	[30,31,34,35]	M-H, fixed	RR = 3.78, CI: [1.63, 8.76]	0.002	72	⊕^⊝⊝⊝; Very low^a,b,c
UFH	[30,31,34,35]	M-H, random	RR = 4.04, CI: [0.54,30.52]	0.18	72	⊕^⊝⊝⊝; Very low^a,b,c
LMWH	[30,31]	M-H, fixed	RR = 4.58, CI: [1.32, 15.91]	0.02	0	⊕^⊝⊝⊝; Very low^b
LMWH	[30,31]	M-H, random	RR = 4.46, CI: [1.27, 15.69]	0.02	0	⊕^⊝⊝⊝; Very low^b
C	[30]	M-H, fixed	RR = 0.62, CI: [0.09, 4.26]	0.62	NR	⊕⊕^⊝⊝; Low
C	[30]	M-H, random	RR = 0.62, CI: [0.09, 4.26]	0.62	NR	⊕⊕^⊝⊝; Low
Unclassified	[36]	M-H, fixed	RR = 1.16, CI: [0.80, 1.68]	0.43	NR	⊕⊕^⊝⊝; Low
Unclassified	[36]	M-H, random	RR = 1.16, CI: [0.80, 1.68]	0.43	NR	⊕⊕^⊝⊝; Low
Main outcomes: the effect of CT versus NM on mortality
In-hospital mortality	[28,29,30–34,36]	M-H, fixed	RR = 1.25, CI: [1.10, 1.43]	0.0007	42	⊕^⊝⊝⊝; Very low^a,b,d
In-hospital mortality	[28,29,30–34,36]	M-H, random	RR = 1.19, CI: [1.00, 1.42]	0.05	42	⊕^⊝⊝⊝; Very low^a,b,d
Subgroup analysis: the relationship between conventional anticoagulant interventions classification and mortality
NA	[28,29,32–34]	M-H, fixed	RR = 1.31, CI: [1.31, 1.55]	0.002	48	⊕^⊝⊝⊝; Very low^b
NA	[28,29,32–34]	M-H, random	RR = 1.26, CI: [1.00, 1.59]	0.05	48	⊕^⊝⊝⊝; Very low^b
UFH	[33,34]	M-H, fixed	RR = 0.78, CI: [0.55, 1.13]	0.19	0	⊕^⊝⊝⊝; Very low^a
UFH	[33,34]	M-H, random	RR = 0.84, CI: [0.59, 1.19]	0.32	0	⊕^⊝⊝⊝; Very low^a
Unclassified	[36]	M-H, fixed	RR = 1.37, CI: [1.08, 1.73]	0.009	NR	⊕⊕^⊝⊝; Low
Unclassified	[36]	M-H, random	RR = 1.37, CI: [1.08, 1.73]	0.009	NR	⊕⊕^⊝⊝; Low
Secondary outcomes: the effect of CT versus NM on hemofilter lifespan
Hemofilter lifespan	[17,28,29]	M-H, fixed	MD = −10.59, CI: [−15.45, −5.72]	<0.0001	0	⊕⊕^⊝⊝; Low^a,b
Hemofilter lifespan	[17,28,29]	M-H, random	MD = −10.59, CI: [−15.45, −5.72]	<0.0001	0	⊕⊕^⊝⊝; Low^a,b

Abbreviation: NR, not reported; M-H, Mantel-Haenszel; IV, Inverse Variance. NM, nafamostat mesilate; UFH, unfractionated heparin; LMWH, low molecular weight heparin; C, citrate; NA, anticoagulant-free; CT, conventional anticoagulant therapy.

^aDowngraded one level due to imprecision (defined as wide confidence intervals including no effect and/or low overall sample size).

^bDowngraded one level due to high risk of bias (incomplete outcome data).

^cDowngraded one level due to large heterogeneity between studies.

^dDowngraded one level due to serious inconsistency: point estimates varied widely.