Comparative effectiveness of parenteral anticoagulants (fondaparinux, argatroban, bivalirudin) in heparin-induced thrombocytopenia: A systematic review

Table 1 Characteristics of included studies

Ref.	Design	Setting	Patients (n)/mean age	HIT diagnosis	Interventions	Follow-up
Boyce et al[5] (2011)	RCT	Multi-center	50/62.5	Suspected HIT	Desirudin vs argatroban	35 days
Joseph et al[6] (2014)	Retrospective cohort	Single center	133/61.0	Confirmed/suspected	Bivalirudin	Hospital stay
Bain and Meyer[9] (2015)	Retrospective cohort	Single center	89/66.2	Confirmed/suspected	Bivalirudin vs lepirudin vs Argatroban	Hospital stay
Duewell et al[10] (2021)	Retrospective cohort	Multi-center	167/64.8	Suspected/confirmed	Argatroban vs bivalirudin	30 days
Al-Eidan et al[8] (2018)	Retrospective cohort	Single center	34/65.4	Isolated HIT	Argatroban vs fondaparinux	Hospital stay
Marchetti et al[11] (2022)	Retrospective cohort	Single center	32/68.5	Confirmed HIT	Argatroban	Treatment period
Davis et al[25] (2022)	Retrospective cohort	Multi-center	135/63.2	Confirmed/Suspected	DOACs	90 days
Albuloushi et al[26] (2022)	Retrospective cohort	Single center	48/66.8	HIT	DOACs	30 days
Snodgrass et al[13] (2016)	Retrospective cohort	Single center	43/61.7	Suspected HIT	Fondaparinux	Hospital stay
Pishko et al[15] (2019)	Retrospective cohort	Multi-center	1291/63.1	Suspected HIT	Various	30 days
Carré et al[27] (2021)	Retrospective case series	Single center	22/69.0	HIT	DOACs	6 months
Schindewolf et al[7] (2017)	Retrospective cohort	Multi-center	616/67.2	HIT	Fondaparinux vs others	Hospital stay
Kang et al[12] (2015)	Retrospective cohort	Single center	78/64.5	Suspected HIT	Fondaparinux	30 days
Tardy-Poncet et al[28] (2015)	Prospective cohort	Multi-center	62/66.8	Confirmed HIT	Argatroban	Treatment period
Nand et al[3] (1997)	Retrospective cohort	Single center	108/59.4	HIT with thrombosis	Various	Hospital stay
Dulicek et al[14] (2019)	Retrospective cohort	Single center	37/68.9	HIT	Fondaparinux	Treatment period

RCT: Randomized controlled trials; HIT: Heparin-induced thrombocytopenia; DOACs: Direct acting oral anticoagulants.

Table 2 Risk of bias assessment¹

Ref.	Study design	Selection	Comparability	Outcome assessment	Overall quality
Boyce et al[5] (2011)	RCT	Low risk	Low risk	Low risk	Good
Joseph et al[6] (2014)	Observational	Moderate risk	High risk	Moderate risk	Fair
Bain and Meyer[9] (2015)	Observational	Moderate risk	Moderate risk	Moderate risk	Fair
Duewell et al[10] (2021)	Observational	Moderate risk	Moderate risk	Low risk	Fair
Al-Eidan et al[8] (2018)	Observational	High risk	High risk	Moderate risk	Poor
Marchetti et al[11] (2022)	Observational	Moderate risk	High risk	Low risk	Fair
Davis et al[25] (2022)	Observational	Moderate risk	High risk	Moderate risk	Fair
Albuloushi et al[26] (2022)	Observational	Moderate risk	High risk	Moderate risk	Fair
Snodgrass et al[13] (2016)	Observational	High risk	High risk	Moderate risk	Poor
Pishko et al[15] (2019)	Observational	Low risk	Moderate risk	Low risk	Good
Carré et al[27] (2021)	Observational	High risk	High risk	Moderate risk	Poor
Schindewolf et al[7] (2017)	Observational	Low risk	Moderate risk	Low risk	Good
Kang et al[12] (2015)	Observational	Moderate risk	Moderate risk	Moderate risk	Fair
Tardy-Poncet et al[28] (2015)	Observational	Moderate risk	High risk	Low risk	Fair
Nand et al[3] (1997)	Observational	High risk	High risk	High risk	Poor
Dulicek et al[14] (2019)	Observational	High risk	High risk	Moderate risk	Poor

¹The inter-reviewer agreement for risk of bias assessment was assessed good (κ = 0.72).

RCT: Randomized controlled trials.

Table 3 Key clinical outcomes summary

Agent	Studies (n)	Thrombotic events (%)	Major bleeding (%)	Mortality (%)
Fondaparinux	4	0-15.4	5.1-11.1	5.9-9.3
Argatroban	6	5.9-21.7	5.9-14.5	0-21.7
Bivalirudin	3	7.5-16.9	7.8-15.8	12.4-20.0
Overall range	16	0-30	5-15	0-22

Table 4 GRADE evidence assessment¹

Comparison	Outcome	Studies	Design	Risk of bias	Inconsistency	Indirectness	Imprecision	Certainty
Any anticoagulant vs control	Thrombotic events	1 RCT, 8 Obs	Mixed	Serious	Not serious	Not serious	Serious	Low
Fondaparinux vs others	Thrombotic events	5 Obs	Observational	Serious	Serious	Not serious	Serious	Very low
Argatroban vs bivalirudin	Thrombotic events	3 Obs	Observational	Serious	Not serious	Not serious	Serious	Low
Any anticoagulant vs control	Major bleeding	1 RCT, 10 Obs	Mixed	Serious	Serious	Not serious	Serious	Very low
Any anticoagulant vs control	Mortality	1 RCT, 8 Obs	Mixed	Serious	Serious	Not serious	Serious	Very low

¹All major comparisons and outcomes provided a low to very low overall certainty of the evidence due to the dominance of observational study designs, risk of bias issues, and imprecision resulting from small sample sizes.

GRADE: Grading of Recommendations Assessment, Development and Evaluation; RCT: Randomized controlled trials; Obs: Observational.