Skip to main content

Table 4 Multivariable logistic regression models for bridging use and overuse of bridging, adjusted for clustering at hospital level

From: Guideline compliance for bridging anticoagulation use in vitamin-K antagonist patients; practice variation and factors associated with non-compliance

  OR (95% CI)a
Model 1, All patients
 Bridging used (reference: no bridging used)
  ICU/CCU stay during admission 4.45 (1.72–11.51)
  Second surgery performed 3.21 (0.83–12.49)
  Mechanical heart valve 8.10 (2.38–27.50)
  Type of 1st surgery (reference category: urologic)  
   Orthopaedic 1.10 (0.42–2.91)
  Gastrointestinal 3.45 (1.21–9.87)
  Vascular 3.21 (1.01–10.21)
  Other 3.57 (1.14–11.21)
  Venous thromboembolism 3.91 (1.57–9.74)
  iCVA/TIA 2.49 (1.02–6.11)
  Previous bleedingb 3.59 (0.80–16.17)
Model 2, Low TE risk patients:
 Overuse of bridging (reference: compliant use)
  Type of 1st surgery (reference category: urologic)  
   Orthopaedic 3.18 (0.60–16.71)
   Gastrointestinal 14.85 (2.69–81.99)
   Vascular 13.01 (1.83–92.30)
   Other 57.30 (5.27–623.62)
  Non-elective surgery 8.67 (1.67–45.14)
  Lowest 25th percentile SES 0.33 (0.11–1.02)
  VKA reversal agent used 0.22 (0.04–1.16)
  1. CCU cardiac care unit, ICU intensive care unit, iCVA ischaemic cerebrovascular accident, TIA transient ischaemic attack, SES Socioeconomic status, VKA Vitamin-K antagonist
  2. a Adjusted for clustering at hospital level
  3. b Any previous bleeding event annotated in the medical record