Table 1.

Probabilities used in the decision analysis

Variables	Baseline probability	Range of plausible probabilities	Threshold value within published range	Threshold from 0 to 1
Probability of DVT	0.32 [12]	0.18 to 0.50 [[9] to [11,12]]	No	Yes (> 0.6*)
Probability of PE following development of DVT	0.1[42,43]	0.01 to 0.18 [22,24,44,45]	No	Yes (> 0.3*)
Probability of death from PE	0.17 [46]	0 to 0.50 [47]	No	No
Probability of a CNS bleed on LMWH	0.108 [15]	0.03 to 0.23 [24,44]	No	Yes (> 0.017+)
Effectiveness of LMWH in preventing DVT (e)	0.47 [12]	0.33 to 0.82 [12,42]	Yes (> 0.8*)	Yes (> 0.8*)
Probability of death from a CNS bleed	0.105 [48]	0.08 to 0.3 [49,50]	No	No
Probability of disabling neurological deficit after CNS bleed	0.17 [51]	0.16 to 0.33 [52,53]	No	No
Probability of an ICU-related systemic bleed	0.035 [23]	0.027 to 0.046 [23]	No	No
Probability of death from ICU-related systemic bleed	0.02 [23]	0.001 to 0.035 [23,46]	No	No
Effectiveness of not receiving LMWH in reducing CNS bleeds	0.315 [4,15]	0.001-0.99 [4-6,23,47,54]	Yes (> 0.05+)	Yes (> 0.05+)
Effectiveness of not receiving LMWH in reducing ICU-related systemic bleeds	0.146	0 to 0.66 [8,12,55]	No	No

Table of probabilities and plausible ranges used for the decision analysis. The last two columns indicate variables for which a threshold value was identified in one-way sensitivity analysis. Values with an asterisk (*) indicate a threshold value above which providing anticoagulant prophylaxis becomes the preferred strategy. Values with a plus (+) indicate a threshold value above which withholding anticoagulant prophylaxis becomes the preferred strategy.

CNS: central nervous system; DVT: deep vein thrombosis; ICU: intensive care unit; PE: pulmonary embolism